A COMPREHENSIVE TREATISE INORGANIC AND THEORETICAL CHEMISTRY Vol II - J. W. MELLOR

A COMPREHENSIVE TREATISE INORGANIC AND THEORETICAL CHEMISTRY Vol II - J. W. MELLOR

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: A C O M P R E H E N S I V E TREATISE O N INORGANIC A N D T H E O R E T I C A L CHEMISTRY V OLUME I1 F, C1, Br, I , Li, Na, K , Rb, Cs W orks by J. W. M IILLOR, D Sc., F.R.S. A COMPREHENSIVE TREATISE O N INORGANIC A N D THEORETICAL CHEMISTRY. I n Thirteen Volumes. Royal 8vo. V ol. I. 1 1, 0 . W ith 274 L)iagrains. E 3 3s. net. J'oI. 11. F', C1, Br, I , I,i, Na, I<, ICb, Cs. With I 70 Diagrams. L 3 3s. net. Vol. 111. Cu, A g, A u, Ca, Sr, Ba. With 158 D iagrams. £ 3 3.7. n et. VoI. IV. Ra and Ac Families, Be, Mg, Zn, Cd, Hg. With 2 32 D iagrams. A 3 3s. net. Vol. V . 13, Al, Ga, I n, TI, Sc, Ce, and Rare Earth Metals, C ( Part I.) W ith 206 D iagrams. £ 3 3s. n et. \'ol. V I. C (Part II.), Si, Silicates. With 2 2 1 D iagrams. A 3 3s. net. V ol. V I I . Ti, Zr, Elf, T h, G e, Sn, Pb, I nert Gases With 255 D iagrams. 3 s. net. I N T R O D U C T I O N T O M O D E R N INORG-4KIC C REMISTRY. With 2 32 I jiagrams. Crown avo, 9s.net. MODERN INORGANIC CHEMISTRY. With 3 69 1)iagrams. Crown 8v0, 12s.6d. n et. I-IIGHER M A T H E M A T I C S F O R S T U D E N T S O F C H E M I S T R Y A N D PEIYSICS. With special reference to Practical W ork. W ith 1 89 D iagrams. Svo, z r s . n et. TI-IE C R Y S T A L L I S A T I O N O F I R O N A N D S T E E L : an Introduction to the Study of Metallography. W ith 65 I llustrations. Crown 8v0, 8s. 6d. n et. . L O N G M A S S , G R E E N A N D CO., LTD. LONDON, N E W YORK, TORONTO, C ALCUTTA, B OMBAY, A N D MADRAS. A COMPREHENSIVE TREATISE I NORGANIC A N D T H E O R E T I C A L CHEMISTRY BY J. W . M ELLOR, I3.Sc., F.1Z.S. V OLUME II L O N G M A N S , G R E E N A N D C O. 39 PA'l'EKNOSTER ROW, L ONDOX, E.C.4 N EW Y ORK, TORONTO C ALCU? TA, E OMBAY A N D k 1ADRAS 1927 L?'.D. C ONTENTS CIIAPTER X VII T H E HALOGEN8 8 I. The Occnrrence of Fluorine (1); 9 2. T he History of Fluorine (3) ; 5 3. T he Prepara- tion. of Fluorine (7) ; 9 4. T he Properties of Fluorine (9) ; $ '5. T he Occurrence of chlorine, Bromine, and Iodine (15) ; § 6. T he History of Chlorine, Bromine, and Iodine (20) ; 5 7. T he Preparation of Chlorine (25); 5 8. T he Preparation of 0 Bromine (38) ; 5 9. T he Preparation of Iodine (41) ; 5 1 . T he Physical Properties of Chlorine, Bromine, and Iodine (46) ; f 11. S olutions of Chlorine, Bromine, and Iodine in Water, etc. ( 71); 5 12. C hemical Reactions of Chlorine, Bromine, and Ioaine (90) ; $ 13. Colloidal Iodine and Iodized Starch (98) ; 9 14. T he Atonlic Weights of Chlorine, Bromine, and Iodine (101) ; 9 1 5. T he Colour of Solutions of bdine (110) ; 5 16. B inary Compounds of the Halogens with One Another (118). CHAPTER X VIII T HE C OMPOUNDS O F T H E H ALOGENS W ITH H YDROGEN $ 1. The Preparation of Hydrogen Fluoride and Hydrofluorio Acid (127) ; 5 2. T he Properties of Hydrogen Fluoride and Hydrofluoric Acid (129) ; 9 3. The, Fluorides (137) ; § 4. E quilibrium, and the Kinetic Theory of Chemical Action (141) ; § 5. T he Union of Hydrogen and Chlorine in Light (148) ; 3 6 . T he Preparation of Hydrogen Chloride and Hydrochloric Acid (158) ; $ 7. T he Preparation of Hydrogen Brbmide and H-ydrobromic Acid (167) ; 5 8.. T he Preparation of Hydrogen Iodide and Hydriodic Acid (170) ; 9 9. T he Physical Properties of the Hydrogen Chloride, Bromide, and Iodide (173) ; 9 10. Properties of Hydrochloric, Hydrobromic, and Hydriodic Acids (182) ; § 11. T he Chemical Propertics of the Hydrogen Halides and the Corresponding Acids (200) ; $ 12. T he Chlorides, Bromides, and Iodides (214); 5 13. Colour Changes on Heating Elements and Compounde (221) ; 5 14. Double a d Complex Salts (223) ; § 15. Double Halides (228) ; 9 16. P erhalides or P olyhdides (233). V CONTENTS CHAPTER T HE O XIDES A KD OXYACIDS O F X IX CHLORINE, B R O M I N E , A N D I ODINE § 1. Chloride Monoxide (240) ; $ 2 . T he Preparation of Hypochlorous, Hypobrornons, and Hypoiodous Acids (243); 6 3. T he Properties of the Hypohalous Acids and '258); $ 5 T he Hypochlorites, Hypo. their Salts (250) ; $ 4. Bleaching Powder ( bromites, and Hypoiodites (267); $ 6. E lectrolytic Processes for the Preparation . of Hypochlorites, Hypobromites, and Hypoiodites (276) ; 5 7 Chlorine, Bromine, and Iodine Trioxides; and the Corresponding Acids (281); 5 8. Chlorine Di- or . ; 0. Per-oxide (286); $ 9 I odine Di- or Tetra-oxide (291) 5 1 T he Halogen Pentoxides (293); $ 11. T he Preparation of Chloric, Bromic, and Iodic Acids, and of their Salts (296); $ 12. T he Properties of Chloric, Bromic, and Iodic Acids i n d their the Salts (305); 5 13. T he Halogenates-Chlorates, Bromates, and Iodates-of Metals (324); 5 14. P erchloric Acid and the Perchlorates (370) ; 5 15. P erbromic Acid and the Perbromates (384); $ 16. Periodic Acid and the Periodates (386) ; Q 17. T he Perchlorates (395) ; $ 18. Periodates (406). CHAPTER XX T HE ALKALI METAL8 9 1 T he History of the Alkali Metals (419); $ 2. T he Occurrence of the Allrali Metals . (423); § 3. T he Potash Salt Beds (427);$ 4. T he Extraction of Potassium Salts (436); 5 5 T he Extraction of Lithium, Rubidium, and Casium Salts (442); 5 6. . T he Preparation of the Alkali Metals (445) ; 5 7 T he Properties of the Alkali . Metals (451) ; 5 8. T he Binary Alloys of the Alkali Metals (478); Ej 9. T he Hydrides of the Alkali Metals (481); $ 10. T he Oxides of the Alkali Metals (484) ; 5 11. Hydroxides of the Alkali Metals (495); $ 12. T he Alkali Fluorides (512); $ 13. A mmoninm Fluoride (519); 5 14. T he Alkali Chlorides (521); 5 15. T he Properties of the Alkali Chlorides (529); 5 16. A nlmoniuln Chloride (561) ; 5 17. T he Alkali Bromides (577); 5 18. A mmonium Bromide (590); § 19. T he Alkali Iodides (596) ; $ 20. A mmonium Iodide (615) ; $ 21. T he Alkali Monosulphides (621); 5 22. T he Alkali Polysulphides (629) ; 5 23. T he Alkali Hydrosnlphides (641); $ 2 . A mmonium Sulphides (645); 5 25. T he Alkali Sulphates (656); 4 5 26. Alkali Acid Sulphates ; Alkali Hydrosulphates (677); $ 27. d mmonimn Sulphstes (694); $ 28. T he Occurrence and Preparation of the Allrali Carbonates (710); 4 29 T he Manufacture of Soda by N. L eblanc's Process (728) ; § 30. T he Ammonia-Soda or E. Solvay's Process (737); $ 31. T he Properties of the Alkali Carbonates (747); § 32. T he Alkali Hydroca,rbonates, Bicarbonates, or Acid Carbonates (772); 5 33. T he Ammonium Carbonates (780); $ 34. C arbamic Acid and the Carbarnates (792); $ 35. Commercial Ammonium Carbonate " (797) ; 5 36. T he Alkali Nitrates (802); 5 37. Gunpowder (825); 5 39. Ammonium Nitrate (829); 5 39. N ormal or Tertiary Alkali Orthophosphates (847); § 40. Secondary Alkali Ortbophosphates (851) ; $ 41. P rimary Alkali Orthophosphates (858); $ 42. Alkali Pyrophosphates or Diphosphates (862) ; 5 43. Alkali Metal phosphates (867); 9 44. A mmonium Phosphates (871) ; $ 45. T he Relation 87) between the Alkali Metals ( ' 9 . I NDEX . ..... ... , - - . . . . . . . , . . . . 881 , , A BBREVIATIONS A q. = aqueous atm. = atmospheric or atmosphere(s) at. vol. = a tomic volurne(s) at. wt. = a tomic weight(s) T 3 or O = absolute degrees of temperature K b.p. = boiling point(s) "0 = centigrade degrees of temperature coeff. = coefficient conc. = concentrated or concentration dil. = d ilute eq. = equivalent(s) f.p. = freezing point(s) 4 . p . = melting point(s) gram-molecule(s) 401(s) = gram-molecular mol. ht. = molecular heat@) 401. VOI. = molecular volume(s) mol. wt. = molecular weight(s) press. = prcssure(s) sat. = s aturated soln. = solution(s) sp gr. = specific gravity (gravities) sp. ht = specific heat(s) sp. vol. = specific voIume(s) temp. = t empcrature(s) vap. = vapour CHAPTER XVII $ 1. The Occurrence of Fluorine T HEf our elements fluorine, chlorine, bromine, and iodine together form a remarkable family, and they are grouped under the name halogens or salt-formers-;As, seasalt ; y rvvdw, I produce. J . S . C. Schweigger used this term in 1811, and it was also employed by J. J. Berzelius 1 f or the non-oxygenated negative radicles-simple or compound-which combine with the metals to form salts. J. J. Berzelius was inclined to restrict the term more particularly t o the simple radicles F, Cl, Br, I, a nd the compound radicle CN. J. J . Berzelius' term halogen has been retained for the four elements, and cyanogen dropped from the list. The binary salts-fluorides, chlorides, bromides, and iodides-are called halides, halide salts, or haloid salts. This term was also employed by J. J. Berzelius for the salts formed by the union of thc metals with fluorine, chlorine, bromine, iodine, and cyanogen ; a s before, cyanogen has again been dropped from the list. The first member of the family of halogens, fluorine, is the most chemically active element known ; t he chemical activity of the other members decreases with increasing at. wt. Fluorine can scarcely be said to occur free in nature, although C. A. Kenngott (1853) and F. Wohler (1861) suggested that the violet f e l ~ a r of Wolsendorf, and H . Becquerel and H. Moissan (18pO) 2 t hat the violet fluorspar from Quinci6 (ViUefranche), probably contain free fluorine as an occluded gas. These varieties of fluorspar were designated hepatischer Flussspath and Stinkjhssspath by K. C. von L e ~ n h a r d (1821) and J. F . L. H ausmann (1847).3 When these minerals are powdered they emit a peculiar odour recalling ozone, and this has been attributed by various observers to the presence of various substances-4.g. hypochlorous acid (M. Schafhautel), ozone (C. F. Schonbein), free fluorine, or of fluorine from the dissociation of an unstable fluoride or perfluoride (0. Loew).4 The chemical reactions of t he gas, however, were found by H. Becquerel and H. Moissan to correspond with fluorine which must be present either as occluded free fluorine, or else as an unstable perfluoride. The evidence is not decisive though the former is the more probable explanation of the reactions. P. Lebeau 5 obtained similar indications of fluorine in emeralds obtained from the vicinity of Limoges. Combined fluorine is fairly widely distributed in rocks. According to F. W. Clarke,c it is about half as abundant as chlorine, since he estimates that the terrestrial matter in the half-mile crust--land and sea-contains 0.2 per cent. of chlorine, and 0.1 per cent. of fluorine. F. W. Clarke places fluorine the 20th and chlorine the 12th in the list of elements arranged in the order of their estimated abundance in the half-mile crust of the earth. Small quantities of fluorine are commonly present in igneous rocks. J. K L . Vogt estimated that fluorine is the more abundant in the acidic rocks ; chlorine, in the basic rocks. The most characteristic minerals hontaining fluorine are Jluorspar, Jluor, or Jlu~rite-calcium fluoride -andcryolit2-a double fluoride of aluminium and sodium; the less important or rarer fluoriferous minerals are : Jluellite, AZF"3H20 chiolite, 5NaF.3A1F3 ; sellaite, MgF2 ; ; tysonite, (Ce, L a, D i)F3 ; pachnolite and thomsenolite, NaF.CaF2.AlF3.H20; ralstonite, 2NaF.MgF2.6A1(F,0H)3.4Hz0rosopite, CaF2.2A1(F,0H)3. Fluorine is also con;p tained in some phosphates--e.g.,fluorapatite,phosphorite, sombrerite, coprolites, VOL. 11. 1 B INORGANIC AND THEORETICAL CHEMISTRY and staffelite ; a nd in some si1icates-e.g. topaz, tourmaline, herderite, yttrocerite, amphibole, nocerine, kodolite, melinophane, hieratite, lepidolite, and in many other silicate minerals. Several mineral waters have been reported to contain minute quantities of soluble fluorides. The spring a t Gerez (Portugal) is one of the richest, for, according t o C. liepierre,' it cont'ains 0 '296 t o 0'310 g r l . of solid matter per litre, and of this, 0 .022 t o 0'027 g rm. is an alkali fluoride ; a nd of the 9 3 spring waters examined by P. Carles, 87 contained soluble fluorides. R . Parmentier has denied the existence of fluorine in many waters in which it is supposed to exist ; b ut according to A. Gautier and P. Clausmann, all mineral waters contain fluorine, and the proportion is greatest in waters of volcanic origin. Thermal alkali bicarbonate waters are particularly rich in the element, although the proportion does not appear to depend upon the temp. As a general rule, mineral waters of the same kind show a n increase of fluorine accompanying a rise in the total salts. I n the case of calcium sulphate waters, whatever their origin, the amount of fluorine is about 2 mgrms. per litre. I n 1849, G. Wilson reported on the occurrence of fluorine in the Clyde waters, and in the North Sea ; a nd generally it has been found that sea water contains about three milligrammes per litre ; t he proportion varies slightly in different places and a t different depths. - A. Gautier 8 f ound about 0.11 mgrrn. of combined fluorine per litre of gas collected from a fumerole fissure in the crater of Vesuvius ; a nd 3 .72 mgrms. per litre in the condensed water from the boric acid fumerole of a spring a t Larderello (Tuscany). At the beginning of the nineteenth century L. J. P roust and M. de la MBthhrie 9 first noticed the presence of fluorine in bones, and the fact has since been confirmed by numerous others. A. Carnot found 0 .20 t o 0 .65 per cent. of calcium fluoride in fresh bones, while old fossil bones contained much more-4-88 to 6.21 per cent. This fact was first noticed by J. Stocklasa in 1889. Modern bones were found by A. Carnot to contain a minimum proportion of fluorine ; t ertiary bones contained more ; mesozoic bones still more; and in Silurian and devonian bones, the proportion of fluorine was nearly the same as in apatite. A. Carnot attributes the progressive enrichment of bones to the action of percolating waters containing a small proportion of fluorides in soh--e.g. the waters of the Atlantic contain 0 .822 g rm.per cubic met're. According to F. Hoppe, the enamel of the teeth contains u t o 2 per cent. of calcium fluoride ; a nd according to W. Hempel and W. Schefler, t e t eeth of horses contain 0 .20 t o 0 -39 per cent. of fluorine, and the teeth of man, 0 .33 t o 0'59 per cent.10unsound teeth had but 0'19 per cent. of fluorine. P . Carles 11 found 0 .012 per cent. of fluorine in the shells of oysters and mussels living in sea water, while fossil oyster shells contained 0 '015 per cent. He also reported about one-fourth as much fluorine in fresh-water mussel shells as is present in the shells of sea-water mussels. The brain (E. N. Horsford),lz blood (G. Wilson, and G. 0. Rees), and the milk of animals (3'. 5. H orstmar) have some fluorine. The brain of man contains about 3 mgrrns. of fluorine, and although the ~ 6 l of fluorine in the animal and vegetable : organism has not been clearlv defined, some physiologists believe that the presence of fluorine is necessary, in aome subtle way, to enable the animal organism to assimilate phosphorus. G. Tammann found that least fluorine was contained in the shells of eggs, and most in the yolks. About 0.1 per cent. of fluorine occurs in the ash of vegetable matter-particularly the grasses.13 A. G. Woodman and H. P. Talbot reported that fluorine is common in malt liquors ; most malted beers contain not less than 0 .2 mgrm. per litre. T. L . Phipson has reported 3 .9 per cent. of fluorine, and 3 2.45 of phosphoric acid in fossil wood from the Isle of Wight, thus showing that the wood had been "fossilized by phosphate of lime and fluorspar." % J. J. Berzelius, Lehrbuch der Chemie, Dreaden, 1 .266,1843 ; J. S . C. Schweigger, Schweigger'a bourn., 3. 249, 1811. T HE HALOGENS 3 C. A. K enngott, Sitzber. Akad. Wien, 10. 286, 1863 ; 11. 16, 1853 ; A. W . von Hofmann, and B. WBhier's Briefwechsel in dem Jahren 1829-73, Braunschweig, 2. 107, 1888 ; H,Becquerel and H. Moissan, Compt. Rend., 111. 669, 1890. a K. C . v on Leonhard, H a d u c h der Orykfogno.sie, Heidelberg, 565, 182 1 ; J . F. L. H ausmann, Handbuch der Mimrabgie, Gottingen, 1441, 1847.. M,Schafhautel, Liebig's Ann., 46. 344, 1843 ; C. F . Schonbein, Journ. prakt. Chem., ( I ) , 74. 326, 1868; ( l ) ,88. 95, 1861 ; G . W y r o u b o f f ,BuZZ. 8 02. Chirn., ( 2 ) , 5. 334, 1856 ; G . Meissner, Unlersuchungen iiber den Sauerstoff, Hannover, 1863; A. S chrotter, Sitzber. Alcad. Wian, 41. 726, 1860 ; Chem. Ztg., 25. 355, 1901 ; J . Garnier, ib., 25. 89, 1901 ; T . Z ettel, ib., 25. 385, 1901 ; H. Moiasan, ib., 25. 480, 1901 ; 0. h e w , Rer., 14. 1144, 2441, 1881. P. Lebeau, Compt. Rend., 121. 601, 1895. 6 F . W . Clarke, The Data of ffeochemistm~, Washington, 34, 1916 ;J . H. L. V ogt, Zeil. paid. Qeol., 225, 314, 377, 413, 1898 ; 10, 1899. C. Lepierre, Compt. R c E . , 128. 1289, 1899 ; P. Carles, ib., 144. 3 7,201, 437, 1907 ; F . Pasmentier, ib., 128. 1100, 1899 ; A. Gautier and P. Clausmann, ib., 158. 1389, 1631, 1914; G. W ilson, B. A. Rep., 47, 1849 ; Chemist, I . 5 3, 1850. A. Gautier, Compt. Rend., 157. 820, 1913 ; V . R. M atkucci, ib., 129. 6 5, 1899 ; J. S tocklam, Cbm. Ztg., 30. 740, 1906 ; A. B run, Recherche8 sur Fexhalui8on wlcuni(2.ue,GenBve, 191 1. 0 J . L. Proust, Jour?~. Phy8., 4 . 224, 1806 ; M . de la MkthBrie, ib., 43.225, 1806 ; A. Carnot, 2 Contpt, Rend., 114. 1189, 1892 ; 115. 246, 1892 ; J . S tocklam, Biedermann'a Ccntrb., 18. 4 44, 1889. 16 F. Hoppe, Arch. p ath. A nnt., 2 . 13, 1862 ;W . H empel and W . Scheffler, Zeit. anorg. Chem., 4 PO. 1 , 1899 ; E . W rampermeyer, Zeit. anal. Chern., 32. 342, 1893 ; T . G assmann, Zeit. physioL Ckenh., 55. 465, 1808. P. Casles, Compt. Rend., 144. 4 37, 1240, 1907. l z E. N . H orsford, Liebig's Ann., 1 49. 202, 1869; G. 0 R ees, Phil. Mag., ( 3), 15. 558, 1839 ; G. Tmmann, Zed. p hy~iol. Chem., 12. 322, 1 M 8 ; Journ. Pltarm. Chim., ( 5 ) , 18. 109, 1888 ; F J.NicklL, C m p t . BemA, 43.885,1856 ; F . S . Horstmar, Pogq. Bnn., i l 339,1860 ; G. W ilson, l. B. A. Rep., 67, 1850 ; E din. Phil. Jozwn., 49. 227, 1850 ; Proc. Roy. Soc. Ed.in,, 3. 463, 1857. l a H. W iLon, Journ. prakt. Chem., ( l ) 5 7.246,1852 ; H . Osj, Ber., 26. 151, 1895 ; P. J . NicklBs, , Ann. Chim. Phys., ( 3), 53. 433, 1858 ; T . L. P hipson, Chem. &ws, 66. 181, 1892 ; Compt. Rend., ii5.473,1892 ; A. G. Woodman and H. P. T albot, Journ. Amer. Chem. Sm., 20. 13W2, 1898. a J. ma Liebig's $ 2. The &tory of Fluorine The mineral now known as fluorspar or fluorite was mentioned in 1529 by G. Agricola, in his Berman~us,sive de re metaWica dialogus (Basiliae, 1529), .and deaignatedJluwes, which, in a later work 1 b y the same writer, was translated into PEikse. A. J. Cronstedt,"n 1758, us6d the terms Pluss, Plusspat, and GEasspat, aynonymously. C. A. Napione (1797) called the mineral Jluorite ; P. 5. B eudant (1832),Jluorine ; a nd M. Sage (1777), spath fusible. These terms are derived from the Latin Jluo, I flow, in reference t o t he fluxing action and the ready fusibility of the mineral ; consequently, Jluor lapis, sputum vitreum, and Glasspath mean the fluxing stone. J. G . JVallerius 3 refers to the luminescence of the mineral when warmed, and this phenomenon led to its being called Zithophosphorus and phosphoric spar. The variety which gives a greenish phosphorescence is called chlorophaneI appear-and also pyro-emerald. xXwpds, green ; IE. Kopp reports4 that H. Schwanhardt in 1670 etched glass by the action of fluorspar and sulphuric acid, and that in 1725, M. P auli made a liquid for etching glass by mixing nitric acid and powdered fluorspar. I n 1764, A . S. Marggraff 6 distilled the mixture of sulphuric acid and fluorspar in a glass retort, and found a white powde~ be suspended in the water of t he receiver. He therefore concluded to that the sulphuric acid separates a volatile earth from the fluorspar. C. W. Scheele.0 repeated A. S. Marggraff 's experiment, and, in his E x a m n chemicum jhoris mineralis ejwque acidi (1771), concluded that the sulphuric acid liberates a peculiar acid which is united with lime in fluorspar. The acid was called Plusssaure-fluor acid-and fluorspar was designated Jlusssaurer KaEL. After the expulsion of the fluor acid from the lime by sulphuric acid, selenite-calcium sulphate-remained in the retort. He found that hydrochloric, nitric, or phosphoric acid could also be -used in place of sulphuric acid with analogous results. M . Boullanger 7 t ook the .INORGANIC AND THEORETICAL CHEMISTRY view that Scheele's fluor acid was nothing but muriatic acid combined with some earthy substance, and A. G . Monnet that it was a volatile compound of sulphuric acid and fluor. C. W. S cheele,~owover,refuted both hypotheses in 1780 ; a nd concluded : I h ope that I h ave now demonstrated that the acid of fluor is and remains entirely a mineral acid eui g enerie. C. W. Scheele generally used glass retorts for the preparation of the acid, and he was much perplexed by the deposit of silica obtained in the receiver. C. W. Scheele thought that the new acid had the property of forming silica when in contact with water, and it was therefore regarded as containing combined silica. The source of tho silica was subsequently traced by J. C. F. Meyer and J . C. MTiegleb9 t o the glass of the retorts, and was not formed when the distillation was effected in metal vessels, and the acid va-pours dissolved in water contained in leaden vessels. The gas obtained when the fluorspar is treated with sulphuric acid in metal vessels is hydrofluoric acid, and if i n glass ressels, some hydrofluosilicic acid is mixed with the hydrofluoric acid. I n Lavoisier7ssystem,lO Scheele's acid of fluor became Z'acide Jluorique-a combination of oxygen with an unknown radicle,$uorium ; a nd in 1789, A. L. Lavoisier wrote : It remains to-day to determine the nature of the fluoric radicle, but since the acid has not yet been decomposed, we cannot form any conception of the radicle. I n 1809, J. I;. Gay Lussac and L. J. Thenard l1 a ttempted t o prepare pure hydrofluoric acid, and although they did not succeed in making the anhydrous acid, they did elucidate the relation of silica and the silicates t o this acid. H. n avy's work on t he elementary nature of chlorine was published about this time ; a nd he received two letters-dated Nov. lst, 1810, and Aug. 25th, 1812 12-from A. Ampere suggest-, <I mg many ingenious and original arguments " in favour of the analogy between hydrochloric and hydrofluoric acids. I n the first letter, A. Ampere said : It remains to be seen whether electricity would not decompose liquid hydrofluoric acid i f w ater were removed as far as possible, hydrogen going to one side and oxyfluoric acid to the other, just as when water and hydromuriatic acid are decomposed by the same agent. The only difficulty to be feared is the combination of the oxyfluoric acid set free with the conductor with which i t would be brought into contact i n the nascent state. Perhaps there is no metal with which i t would not combine, but supposing that oxyfluoric acid should, like oxymuriatic acid, be incapable of combining with carbon, this latter body might be a sufficiently good conductor for i t to be used with success as such in this experiment. I n the second letter, A. AmpBre suggested that the supposed element be called le Jluor-Jluorine-in agreement with the then recently adopted name chlorineFrench, E chlore. A. AmpPre's suggestion has been adopted universally. No one e doubted the existence of t he unknown element fluorine although it successfully resisted every attempt to bring it into the world of known facts. Belief in its existence rested on the many analogies of its compounds with the other three members of the halogen family. For over seventy years it was neither seen nor handled. During this time, many unsuccessful experiments were made to isolate the element. H . Davy 1 3 t hus describes his attempts : I u ndertook the experiment of elect'rizing pure liquid fluoric acid with considerable interest, as i t seumed to offur the most probable muthod of ascertaining its real nature, but considerable difficulties occurred in executing the process. The liquid fluoric acid immediately destroys glass and all animal and vegetable substances, i t acts on all bodies containing metallic oxides, and I know of no substances which are not rapidly dissolved or decomposed by it, except metals, charcoal, phosphorus, sulphur, and certaiu combinations of chlorine. I a ttempted to make tubes of sulphur, of muriates of lead, and of copper containing metallic mires, by which i t might be electrized, but without success. I succeeded, T HE HALOGENS however, in boring a piece of horn silver in such a manner that I was able to cement a platina wire into it, by means of a s irit lamp, and by inverting this in a tray of platina fiIled with liquid fluoric acid I c ontrive to submit the fluid to the agency of electricity in such a manner that in successive experiments i t was possible to collect any elastic fluid that might be produced. B Having failed t o isolate the element by the electrolysis of hydrofluoric acid and the fluorides, H. n a v y tried if t he element could be driven from its combination by double decomposition. H e attempted to drive the " fluoric principle " from the dry fluates of mercury, silver, potassium, and sodium by means of chlorine. H e said : The dry s d t s were introduced in small quantities into glass retorts, which were exhausted and then flUed with pure chlorine ; t he part of the retort i n contact with the salt was heated gradually till i t became red. There was soon a strong action, the fluate of mercury was rapidly converted into corrosive sublimate, and the fluate of silver more slowly became horn silver. I n both experiments there was a violent action upon the whole of the interior of the retort. On examining the results, i t was found that in both instances there had been a considerable absorption of chlorine, and a production of silicated fluoric acid gas and oxygen gas. I t ried similar experiments with similar results upon dry f luate of potassa and soda. By the action of a red-heat they were slowly converted into muriates with the absorption of chlorine, and the production of oxygen, and silicated fluoric acid gas, the retort being corroded even to its neck. H. Davy assumed that his failure t,o obtain the unknown element was due to the potency of its reactions. H. D avy tried vessels of sulphur, carbon, gold, horn silver, and platinum, but none appeared to be capable of resisting its action, and " i ts strong affinities and high decomposing agencies " led t o its being regarded as a kind of alcahest or universal solvent. G. Aimit (1833) employed a vessel of caoutchouc, with no better result. The brothers C. J. a nd T. Knox (1836) 14 sagaciously tried to elude this difficulty by treating silver or mercury fluoride with chlorine in an apparatus made of fluorspar itself. E . Fritmy believed that the failure in this as well as in P. Louyet's analogous attempt with fluorspar or cryolite vessels, in 1846, was due to the fact that the two fluorides do not decompose when moisture is rigorously excluded; and, if moisture be present, they form hydrofluoric acid. E. Fritmy also did not succeed in decomposing calcium fluoride by means of oxygen, when heated to a high temp. in a platinum tube. E. Fritmy electrolyzed fused fluorides -calcium, potassium, and other metal fluorides-in a platinum crucible with a platinum rod as anode. The platinum wire electrode was much corroded, and a gas was evolved which E. Frbmy believed t o be fluorine because it decomposed water forming hydrofluoric acid, and displaced iodine from iodides. He was able to decompose calcium fluoride a t a high temp. by means of chlorine, and when the fluoride is mixed with carbon. E . FrBmy, however, made no further progress in isolating the elusive element, although he did show how anhydrous hydrofluoric acid could be prepared. G. Gore l 6 m ade some experiments on the electrolysi~ silver fluoride and on the action of of chlorine or bromine on silver fluoride a t 1 5 ~ 5or 38 days, and at 110' for 6 days, in f~ vessela of various kinds-with vessels of carbon, a volatile carbon fluoride was formed. 8. Kammerer lqfailed to prepare the gas by the action of iodine on silver fluoride io sealed ghss tubes ; according to L. PfaundIer, the product of the action is a mixture of silicon fluoride and oxygen. 0. Loew heated cerium tetrafluoride, CeF,.H,O, or the double salt, 3KZ'.2CeF4.2H,0, and obtained a g as, which he considered to be fluorine, when the tetrafluoride decomposed forming the trifluoride, CeP,. B. Brauner also obtained a gas resembling chlorine by heating lead tetrafluoride, or double ammonium lead tetrafluoride, or potassium hydrogen lead fluoride, K,HPbF,. I n the latter case a mixture of potassium fluoride, KF, and lead difluoride, PbF,, remained. 0. Ruff claims to have made a l ittle fluorine by heating the compound HKPbF,. As H. Moissan has said, i t is possible thak fluorine might be obtained by a chemical process in which a higher fluoride decomposes into a lower fluoride with the liberation of fluorine-say, 2CeF4=2CeF,f P ,. 0: Ruff has failed to confirm B. Brauner's observations with the fluorides in question. With lead tetrafluoride in a p latinum vessel, lead difluoride and platinum tetrafluoride are formed; liquid or gaseous silicon tetrafluoride is practically .without action on the salt although a small quantity of a gas which acts on potassium iodide is formed without altering the INORGANIC AND THEORETICAL CHEMISTRY composition of the gas. Antimony pentafluoride acts similarly. With sulphur and iodine t he corresponding higher fluorides are formed. Other suggestions have also been made to prepare fluorine by chemical processes-0. T. Christensen I 7 proposed heating the higher double fluorides of manganese ; A . C. Oudemans, potassium fluochromate ; a nd H. Moissan, platinum fluophosphates. About 1883, H. B . Dixon and 8 . B . Baker made an attempt t o displace fluorine by oxygen from uranium pentafluoride, UP,. A . B audrimont tried the action of boron trifluoride on lead oxide without success. Abortive attempts have been made by L. Varenne, d . P. P rat, 9.Cillis, and T. L . P hipson 18 t o prepare the gas by wet processes analogous t o those employed for chlorine by the oxidation of s o h . containing hydrofluoric acid. We now know that this is altogether a wrong line of attack. Some of t he dry processes indicated above may have furnished some fluorine ; for example, in H. B. Dixon and H. B . Baker's experiment, ~ i l v e foil in the vicinity of the uranium fluoride r was spotted with white silver fluoride; gold foil, with yellm aurio fluoride; and platinum foil, with chocolate platinio fluoride. I n 1834, M. F araday 19 t hought that he had obtained fluorine state " b y electrolyzing fused fluorides, but later, he added : " i n a separate I h ave not obtained fluorine; m y expectations, amounting t o conviction, passed away one by one when subject to rigorous examination. This was virtually the position of the fluorine que~tion about 1883, when H.Moissan,2O a pupil of E. FrBmv, commenced systematic work on the subject, and the reports of the various stages" of his work have been collected in his important monograph Le Jluor et ses compose's (Paris, 1900). H e first tried (1) The decomposition of gaseous fluorides by sparking-e.g. the fluorides of silicon, SiF4 ; phosphorus, PF5 ; boron, BF3 ; a nd arsenic, AsF3. The silicon and boron fluorides are stable. Phosphorus trifluoride forms the pentafluoride. The fluorine derived from phosphorue pentafluoride reacts with the material of which the vessel is made ; similarly with arsenicfluoride. (2) The action of platinum a t a red heat on the fluorides of phofiphorus and silicon. Phosphorus pentafluoride furnishes some fluorine which unites with the platinum of the apparatus used ; phosphorus trifluoride formed the pentafluoride and fluo-phosphides of platinum ; silicon tetrafluoride gave no signs of free fluorine ; H. Moissan came to the concIusion that no reaction carried out a t a high temp. was likely to be fruitful. (3) The electrolysis of arsenic trifluoride t o which some potassium hydrogen fluoride was added to make the liquid conducting ; a ny fluoride given off a t the anode was absorbed by the electrolyte forming arsenic pentafluoride. H. Moissan then tried the electrolysis of highly purified anhydrous hydrofluoric acid, but he found, consonant with G. Gore's and M. F araday's observations,21 that anhydrous hydrofluoric acid is a non-conductor of electricity. I a small quantity f of water be present, this alone is decomposed, and a large quantity of ozone is formed. As the water is broken up, the acid becomes less and lesa conducting, and, when the whole has disappeared, the anhydrous acid no longer allows a current to pass. He obtained an acid so free from water that " a current of 35 ampBres furnished by fifty Bunsen cells was totally stopped." The current passed readily when fragments of dry potassium hydrogen fluoride KF.HF, were dissolved in the acid, and a gaseous product was liberated a t each electrode. Success ! T he element -fluorine was isolated by Henri Moissan on June 26th, 1886, during the electrolysis of a soh. of potassium fluoride in anhydrous hydrofluoric acid, in an apparatus made wholly of I n this way, H. -Moissan solved what H. E. Roscoe called one of the most difficult problems in modern chemistry. While the new element possessed special properties which gave i t an individuality of its own, a nd a few surprises occurred during the study of some of its combinationa ; y et the harmonious analogy between the members of the halogen family-fluorine, chlorine, bromine, and iodine-was fully vindicated. With fluorine in the world of reality, chemists were unanimous in placing the newly discovered element a t the head of the halogen family, and in t hat very position which had been so long assigned to it by presentiment or faith. T HE HALOGENS @. Agricola, lnlerpretatio Germanica mcum ~ e mietallicce, Basil, 464, 1540. A. J. C romtedt, Minerdogie, Stockholm, 93, 1758 ; C. A. N apione, Elenzenti di XineraEogia, Turin,373, 1797 ; F. S . B eudant, Traitk &!rnentai~e min~ralogie, de Paris, 2. 517, 1832 ; M. Sage, E ~ h e n s e. minhalogie docimtique, Paris, 155, 1777. d J. G. l@allerius, Mineralogie, Berlin, 87, 1750. * H . K opp, Geschichte der Chemie, Braunschweig, 3. 363, 1845. A. S. Marggraff, M&m.Acad. Berlin, 3, 1768. C. W . Scheele, Mkm. Acad. Stockholm, ( l ) , 33. 120, 1.771 ; Opuscula chemica et physica, Lipss, 2. 1, 1789. M. BouUanger, Expdriences et obserrmtiom sur le spath vitreux, ouJEuorapathiqzre, Paris, 1773 ; -4. G. Monnet, Rozier's obsermtions aur In pk.ysique, 10. 106, 1777 ; A nn. Chim. Phys., ( I ) , 10. 42. 1791. C. W . Scheele, Mdm. Acad. Stockholm, ( 2 ) ,I 1 , 1780 ; OpuscuEa chemica et physica, Lipsae, . 2. 92, 1789 ; Chemical Essays, L ondon, 1-51, 1901. J. C. F. Meyer, Schr. Berlin. Ges. Naturforu., 2. 319, 1781 ; J . C. Wiegleb, CreEE's Die neuesten fintdeckungen i n der Chemie, 1. 3, 178.1 ; C. F B ucholz, zb., 3. 50, 1781 ; L. B. , G. de Morveau, . Journ. Phys., 17, 216, 1781 ; M . H . K laproth, CreU's Ann., 5. 397, 1784; F . C . A chard, ib., 6. 145, 1785 ; M . P uymaurin, ib., 3. 467, 1783. A. L Lavoisier, Traitd klhentaire de chimie, Paris, 1 263, 1789. . l 1 J. L. G ay Lussac and L. J . T h h a r d , Ann. Chim. Ph.ys., ( I ) , 6 9. 204, 1809. l a A. A m p h e , reprinted Ann. Chim. Phys., (6),4. 8,1885 ; F D. C hattaway, Chem. News, 107. . 25, 37, 1913. l 9 H . D avy, Phil. Tram., 103. 263, I813 ; 1 M . 62, 1814 ; A nn. Chim. Phys., ( I ) , 88. 2 71,1813. l 8 G. A im& Ann. Chim. Phys., ( 2 ) ,55. 443, 1834 ;C. J. a nd T . K nox, Proc. Roy. Irk% A d . , 1. 54,1841 ; Phil. Hag., ( 3 ) ,9.107,1836 ; C. 6.K n o ~ , i b . ,( 3), 16. 190,1840; P. Louyet, C mpt. Rend., 23. 960, 1846 ; M.434, 1847 ; 3 Frkny, ilt., 38. 3 93, 1854 ; A nn. Chim. Phys., ( 3 ) ,47. 6 , 1856. . 1 6 G. Gore, Phil. T r a m , , 160. 227, 1870 ; 161. 321, 1871 ; Chem. News, 50. 150, 1884. ."1 K ammerer, Journ. prakt. Chem., ( I), 85. 452, I862 ; ( 11, 90. 191,1863 ; A. B audrimont, ih., ( I ) , 7 . 447, 1836 ; L. P faundlec, Sitzber. Akad. V i e n , 4 . 258, 1863 ; 0. L oew, Rer., 14. 1144, 6 2441, 1881 ; B. Brauncr, ib., 14. 1944, 1881 ; Journ. Chem. SOL, 65. 393, 1894 ; Zeit. anorg. C hm., 98. 38, 1916; 0 . R uff,ib., 98. 27, 1916; Zed. angew. Chem., 20. 2217, 1907. l 7 0 T . C hristensen, Journ. p a k t . Chem., (2), 34. 4 1, 1886 ; A . Baudrimont, ib., ( I), 7. 4 47, . 1836; A. C. O udemans, Rec. Trav. Chim. Pays-Bas, 5. 111, 1886; 8. Moissan, Bull. Soc. Chim., (8), 5. 454, 1891 ; H . B . D ixon and H . B . Baker, Private c ommunication. L. P hipson, Chem. News, 4 . 21.5, 1861 ; 5 P. Prat, Compt. Rend., 65. 345, 511, 1867 ; . 12. Y wenne, & ., 91. 989, 1880 ; P. CJiIlis, Zeit. Chem., 11. 6 60, 1868 ; G . Gore, Chem. News, 52. 1 5, 1885; E . W edekind, Ber., 35. 2267, 1902. l D M . F araday, Phil. Tram., 134. 77, 1834 ; E xperimental Researches i n Electricity, London I . 2 27, 1849. 2 0 H . Moissan, Compt. Rend.$ 99. 6 55, 874, 1884 ; 100. 272, 1348, - 1885 ; 101. 1490, 1885 ; 10% 763, 1245, 1543, 1886; 103. 202, 256, 850, 1257, 1886; 109. 637, 862, 1889; 128. 1543, 1899 ; A nn. Chim. Phys., ( 6), 12. 472, 1887 ; ( G ) , 24. 224, 1891 ; Bull. Soc. Chim., (3), 5. 880, 1891 ; Les classiques d e kc acience, 7 , 1914. G. Gore, Phil. Tram., 159, 189, 1869 ; M. F areday, i b., 124. 77, 1834. 5 3. The Preparation of Fluorine When an electric current is passed through a conc. aq. soh. of hydrogen chloride, chlorine is liberated at the anode, and hydrogen a t the cathode. When aq, hydrofluoric acid is t reated in the same way, water done is decomposed, for oxygen is libe~atedat the anc;de, and hydrogen at the cathode. The anhydrous acid does not conduct electricity, and it cannot therefore be electrolyzed. H. Moissan found that jf potassium fluoride be dissolved in the liquid hydrogen fluoride, the soh. readily conducts electricity, and when electrolyzed, hydrogen is evolved a t the cathode, and fluorine at the anode. I n the first approximation, it is supposed that the primary products of t he electrolysis are potassium at the anode, fluorine at the cathode : 2KHPz=2HP+2K+B,. The potassium reacts with the hydrogen fluoride reforming fluorlde and liberating hydrogen : 2 K+2HF=2KF+H2. The reaction is pr~bably more complex than this, and the platinum of t he electrodes plays a part in the secondary reactions. Possibly the fluorine first forms platinum fluoride, PtF4, which produces a double compound with the potassium fluoride. INORGANIC ,4ND THEORETICAI, CHEMISTRY 8 This compound is considered t o be the electrolyte which on decomposition forms the two gases and a double potassium platinum fluoride which is deposited as a black mud. This hypothesis has been devised t o explain why the initial stage of the electrolysis is irregular and jerky, and only after the lapse of an hour, when the substances in soln. are in sufficient quantities to make the passage of the current regular, is the evolution of fluorine regular. 0. Ruff 1 h as shown that ammonium fluoride can be used in place of the pota.ssium salt. H . Moissan flrst conducted the eIectrolysis in a U-tube made from an alloy of platinum a n d iridium which is less attacked by fluorine than plabinum alone. Later experiments PIG.1.-Tube for the Electro. lysis of Hydrofluoric Acid. Fra. 2.-Moissan's Procsss for Fluorine. showed t h a t a tube of copper could be employed. The copper is attacked by the fluorine, forming a surface crust of copper fluoride which protects the tube from further action. Electfodes of pIatinum iridium alloy were used a t first, but later electrodes of pure platinum were used, even though they were rather more attacked than the alloy with 10 per cent. of iridium. The electrodes were club-shaped a t one end so that they need not be renewed so often. The positive electrode was often completely corroded during an experiment, but the U-tube scarcely suffered a t all. A copper tube is illustrated in Fig. I . T he open ends of the tube are closed with fluorspar stoppers ground to fit t he tubes and bored with holes which grip the electrodes. The joints are made air-tight with lead washers and shellac. The U-tube, during the electrolysis, i s F Kd~" s urrounded with a glass cylinder, R , i nto which liquid methy1 chloride is passed from a steels i d t he tube A , Pig. 2. L iquid methyl chlori e boils a t -23", a nd i t escapes through an exit tube. The fluorine is passed through a spiral platinum tube also placed in a bath of evaporating liquid methyl chloride, G . T his cools the spiral tuhe down to about -50°, ~ ~ ~ a n~ condenses any gaseous hydrogen fluoride, which d o ~ o / might escape with the fluorine from the u-tube. The eIectrolysis was carried out a t a low temp. in order to prevent the gaseous product being ail.w ith the vapour of hydrogen fluoride, a d rtlso t o diminish the destructive action of the fluorine on the apparatus. I n his later work, H. Moissan cooled the U-tube used for the electrolysis by using a b ath of acetone with solid carbon dioxide in suspension. This cooled the apparaFIG.3 . - - ~ h o r i n eby the Electrolysis tus down to about - 80". The temp. of tho electrolysis of Fused Alkali Hydrofluoride. vessel should not be so low that the potassium hydrogen fluoride crystallizes out. Hence, 0. Ruff and P . I psen preferred to cool the eIectrolysis vessel with a freezing mixture of calcium chloride, and condensed the hydrogen fluoride vapours in a copper condenser C, Fig. 2, cooled with liquid air. The fluorine which leaves the condenser C , t ravels through two small platinum tubes, D a nd E , containing lumps of sodium fluoride, which remove the least traces of hydrogen fluoride by forming NaF.HF. A gIass cylinder is placed outside each of the two cylinders containing methyl chloride. The outer cylinders contain a few lumps of calcium chloride, so as to dry the air in the vicinity of t'he cold jacket, and prevent the Tinder T HE HALOGEN8 9 deposition of frost on the cylinders. With a c urrent from 26 t o 2 8 B m e n cells in series, and a n apparatus containing from 9 0 t o 1 00 grms. of anhydrous hydrofluoric acid containing i n s oh. 2 0 t o 25 grms. of potassium hydrogen fluoride, H. Moissan obtained between two and three litres of fluorine per hour. C. Poulenc and M. Meslans3 have devised a copper apparatus for the preparation of fluorine on a large scale ; a nd likewise a portable laboratory apparatus, also of copper. They substitute a perforated copper diaphragm in place of the U-tube for keeping anode is hollow, and is cooled the two electrode products separate. The internally. G. Gallo did not get good results with this apparatus. W.L. Argo and co-workers pre ared fluorine by the electrolysis of molten potassium hydrofluoride in an electrica ly heated copper vessel which served as cathode, the anode being made of graphite. A copper diaphragm with slots was used as illustrated in Fig. 3. The bubbles of hydrogen evolved during the electrolysis were deflected from the interior of the diaphragm by means of a false bottom. The graphite anode was connected with a copper terminal and insulated by a packing of powdered fluorspar -current, 10 amps., 15 volts ; temp., 240"-250' ; efficiency, 70 per cent. T h b e co-workers also recommend sodium hydrofluoride because it is non-deliquescent ; decomposes below the fusion temp. ; contains more available fluorine for a given weight ; and is less expensive. P 0. Ruff, Z eit. angew. Chent., 20. 1 217, 1907 ; 0. Ruff and E. Geisel, Ber., 36. 2 677, 1903. 0.Ruff and P. I'psen, Bey., 36. 1 177, 1904. 8 C. Poulenc and M. Meslam, Rev. Ckn. A cetylene, 230, 1900 ; G. G d o , A tti A c d . I / i ~ i , (5), 19. i, 2 06, 1910; W . L . Argo, F. C. Mathew, R. Hamiston, and C. O..Anderson, Joum. Phys. Chem., 23. 3 48, 1919; Chem. Eng., 27. 107, 1919; T ram. A mw. Electrochem. Soc., 35, 3 35, 1919. 1 g 4. The Properties oi fluorine I s fluorine an element 1 Since fluorine had never been previously isolated, it remained for H. Moissan to prove that the gas he found to be liberated a t the positive pole is really fluorine. Many of its physical and chemical properties, as will be shown later, agree with those suggested by the analogy of the fluorides with the chlorides, bromide, and iodides. It was found impossible to account for its properties by assuming it to be some other gas mixed with nitric acid, chlorine, or ozone ; or that it is a hydrogen fluoride richer in fluorine than the normal hydrogen fluoride. To show the absence of hydrogen, H. M oiasm dlowed the gas to pass directly from the positive pole through a tube containing red-hot iron ; a ny hydrogen so formed was collected in an a tm. of carbon dioxide. The Iatter was removed by absorption in potassium hydroxide. In several experiments a small bubble of gas was obtained which was air, not hydrogen. The increase in weight of the tube containing the iron corresponded exactly with the fluorine eq. of the hydrogen collected a t the negative pole. The vapouw of hydrogen fluoride were retained by a tube fllled with dry potassium fluoride. For example : I n one experiment a t ube containing iron increased in weight 0 -138 g rm. while 8 0-01 c.c. of hydrogen were collected at the negative electrode. This represents 0 -00712 g rm. of hydrogen, and 0 .00712 x 1 9=Om134 g m . of fluorine. This number is virtually the same as the weight of fluorine actually weighed. Fluorine a t ordinary temp. is a greenish-yellow gas when viewed in layers a metre thick ; t he colour is paler and more yellow than that of chlorine. The liquid gas is canary-yellow ; t he solid is pale yellow or white. Moissan's gas has an intensely irritating smell said to recall the odour of hypochlorous acid or of nitrogen peroxide. Even a small trace of gas in the atm. acts quicklv on the eyes and the mucous membranes; and, in contact with the skin, it caises severe burns, and rapidly desttoys the tissues. If b ut a slight amount is present, its smell is not 10 INORGANIC A M ) T HEORETICAL CHEMISTRY unpleasant. The relative density of the gas (air unity) determined by H. Moissan in 1889, by means of a platinum flask, was 1-26; t hat calculated for a diatomic gas of at. wt. 19.8 is 1.314, and B. Brauner attributed the difference to the presence of some atomic fluorine. H. Moissan's later results (1904) rendered B. Brauner'a hypothesis unnecessary since a density of 1-31 was obtained. The gas employed previously is supposed to have been c~ntarnina~ted a l ittle hydrogen fluoride. with Most of the physical properties of fluorine at a low temp. have been determined by H. Moissan himself and in conjunction with J. Dewar.2 The sp. gr. of liquid fluorine id 1.-14a t -2W0, and 1.108 a t its b.p. -187". The sp. vol. of the liquid is 0.9025 ; a nd the mol. vol. 34.30. The capillary constant of the liquid is about one-sixth of that of liquid oxygen, and seven-tenths of that of water. The coefficient of expansion 3 of the gas ia 0'000304. The volume of the liquid changes one-fourteenth in cooling from -187" to -210". When the gas is cooled by rapidly boiling liquid air, it condenses to a clear yellow liquid which has the boiling point -18'7" a t 760 mm. press. ; a nd the liquid forms a pale yellow solid when cooled by liquid hydrogen, The solid has the melting point -233". The solid loses its yellow tint and becomes white when cooled down to -252". Chlorine, bromine, sulphur, etc., likewise lose their colour a t low temp. J. H. Gladstone's 4 e stimate for the atomic refraction of fluorine for the D-line is 0'53 ; for the A-line 0'63 ; a nd for the H-line 0.35 with the p-formula, and 0'92 and 0.84 respectively with the pLformula. F. Swarts estimated 0.94 Ha, 1.015 D, and 0.963 H y with the p2-formula for fluorine in sat. organic compounds ; a nd for unsaturated compounds with the ethylene linkage, Ha, 0.588 ; D , 0'665 ; H y, 0'638. The atomic dispersion is 0.022 with aat. and 0.05 with the unsaturated compounds. J. H. Gladstone also made several estimates of the index of refraction of fluorine, and his 1870 estimate gave 1.4 (chlorine 9.9) ; in 1885 he placed it a t 1.6 ; a nd in 1891, he considered it to be " extremely small, in fact, less than 1.0." The difficulty is due to the fact that when the magnitude of a small constant is estimated by subtraction from two large numbers the probability of error is large. A direct determination by C. Cuthbertson and E. B. R.-Prideaux gave for the index of refraction of fluorine for sodium light, p=1*000195, which makes the refractivity (p-1) x 106 to be 195. The emission spectrum of fluorine has been investigated by H. Moissan and G. Salet-"he last named, in 1873, compared the spectra of silicon chloride and fluoride, and inferred that five lines in the spectrum of silicon fluoride must be attributed to the fluorine. H. Moissan's measurements, in 1889, measured 1 3 lines in the red part of the spectrum. The lines of wave-length 677, 640.5, 634, and 623 are strong ; t he lines 714, 704, 691, 687.5, 685.5, 683.5 are faint ; a nd 749, 740, and 734 are very faint. Liquid fluorine has no absorption spectrum when in layers 1 cm. thick. According to P. Pasca1,G fluorine is diamagnetic ; t he specific magneticsusceptibility is -3.447 x 1 0-7 ; a nd the atomic susceptibilitv calculated from the additive law of mixtures for organic compounds is -63 ~ 1 0 - ' 1 . Ionic fluorine is univalent and negative. The decomposition voltage required to separate this element from its compounds is 1.75 voIts.7 The ionic velocity (transport number) of fluorine ions at 18" is 46.6, and 52.5 a t 25' with a temp. coeff. of 0-0238. Fluorine possesses special characters which place it a t the head of the halogen family. It forms certain combinations and enters into some reactions in a way which would not be expected i f t he properties of the element were predicted solely by analogy with the other members of the halogen family. From this point of view, said H. Moissan, Z'itucle des composds Jluorbs re'serve encore bien des surprises. Fluorine is the most chemically active element known. It combines additively with most of the elements, and it usually behaves like a univalent element although it is very prone to form double or complex compounds in which it probably exerts a higher valency. It also acts as an oxidizing agent. I n the electrolysis of manganese and chromium salts a higher yield of chromic acid or manganic acid is obtained in the presence of hydrofluoric acid than in the preaence of sulphuric acid9 Fluorine THE HALOGENS unites explosively with hydrogen in the dark with the production of a flame with a red border, and H. Moissan showed thia by inverting a jar of hydrogen over' the fluorine delivery tube of his apparatus. The product of the action is hydrogen fluoride which rapidly attacks the glass vessel when moisture is present, but not if the two gases are dry. Fluorine retains its great avidity for hydrogen even a t temp. as low as '-252.5" when the fluorine is solid, and the hydrogen is liquid. H. Moissan and J. Dewar.10 broke a tube of solid fluorine in liquid hydrogen. A violent explosion occurred which shattered to powder the glass apparatus in whic.h the experiment was performed. It is rather unusual for the chemical activity of a n element to persist at such a low temp. The affinity of fluorine for hydrogen is 50 great that it vigorously attacks organic substances, particularly those rich in hydrogen. The reaction is usually accompanied by the evolution of heat and light, and the total destruction of the compound. The product.^ of t he reaction are hydrogen fluoride, carbon, and carbon fluorides. The avidity of fluorine for hydrogen persists a t very low temp., for turpentine and anthracene may explode in contact f with fluorine at -210°. Even water is vigorously attacked by fluorine. I a small quantity of water is introduced into a tube containing fluorine, it is decomposed, forming hvdrogen fluoride and ozone ; t he latter imparts an indigo-blue tinge t o t he gases in the jar. By measuring the volume of oxygen liberated when fluorine reacts with water,. and measuring the exact quantity of hydrofluoric acid formed, H. Moissan showed that equal volumes of hydrogen and fluorine form hydrogen fluoride. I t he reaction between fluorine and water be symbolized, H20+Bz f =2HP+O, it follows that for every volume of hydrogen collected at the negative pole, half a v ~ l u m e oxygen should be obtained. I n one experiment H. Moissan of collected 26.10 C.C. of oxygen, 52-80 C.C. of hydrogen. I n another experiment he obtained 6.4 C.C. of oxygen per 12.5 C.C. of hydrogen and eq. of 24.9 C.C. of hydrogen fluoride. Liquid fluorine does not react with water. At -2W0, liquid fluorine can be volatilized from the surface of ice without reaction. Neither oxygen nor ozone appears to react with fluorin6,and no oxygen compound of fluorine has yet been prepared. According t o H. Moissan,ll an unstable intermediate compound df ozone and fluorine is possibly formed when water acts on fluorine to form ozonized oxygen because the ozone smell does not appear u&il some time after the fluorine has been passed into the water. 0. Ruff and J . Zedner have tried the effect of heating oxygen and fluorine in the electric arc, but obtained no signs of the formation of a compound of fluorine with oxygen or ozone, for when the gaseous product is passed over calcium chloride (which fixes the fluorine) a mixture is obtained quite free from fluorine. G. Gallo obtained signs of a very unstable compound of ozone and fluorine which is explosive at -23'. Liquid oxygen dissolves fluorine, and if the temp. rises gradually, the first fraction which volatilizes is almost pure oxygen ; t he last fraction contains most of the fluoriue. If liquid air, which has stood by itself for some time, be treated with fluorine, a precipitate is formed which is veky liable to explode. H. Moissan thinks it is probably $uorine hydrate.12 Solid sulphm9 sel&um, a nd tellurium inflame in fluorine gas at ordinary t e m p ; sulphur burns to the hexafluoride, SF6. The reactivity of sulphur or selenium with fluorine persists at -187", but tellurium is without action at this temp. Hydrogen ~ulphide nd sulphur dioxide also burn in the gas-the former produces a hydrogen fluoride and sulphur fluoride. Each bubble of sulphur dioxide led into a jar of fluorine produces an explosion and thionyl fluoride, SOY2, is formed ; b ut if the fluorine be led into the sulphur dioxide, there is no action until the sulphur f dioxide has reached a certain partial pressure when all explodes. I t he fluorine be led into an atm. of sulphur dioxide at the temp, of the reaction, sdphuryl fluoride, S021!2, is formed quietly without violence. Sulphuric acid i s scarcely affected by fluonne. Pluorine does not unite with chlorine a t ordinary temp. 0. Ruff and J. Zedner alao obtained no result by heating fluorine and chlorine at the temp. of the electric ' INORGANIC AND THEORETICAL CHENISTRY arc. Liquid chlorine dissolves fluorine, but the dissolved gas escapes as the chlorine freezes. It is inferred that the gases do not react a t the low temp. -SO0 when fluorine is dissolved in liquid chlorine because (i) the gases evolved when the s o h . is fractionally distilled showed no signs of an abrupt change in composition between 97-32 per cent. of fluorine a t the beginning and 0.63 per cent. a t the end of the operation ; (ii) on cooling a soln. of fluorine in liquid chlorine, there is a tumultuous evolution of gas when the mixture freezes-the solid is chlorine, the gas fluorine. Bromine unites with fluorine a t ordinary temp. with a luminous flame forming bromine trifluoride, BrP3. Similar rcmarks applv to iodine, where the pentafluoride, IFS, is formed. The heat of t he former rebction is small, the latter great. Liquid fluorine, however, does not react with or dissolve bromine or iodine a t -187", nor does i t liberate iodine from potassium iodide. I n the presence of water, chlorine reacts with fluorine forming hypochlorous acid ; a nd bromine, hypobromous acid ; some ckloric or bromic acid may also be formed, ahd part of t he water is also decomposed by the excess of fluorine. If fluorine be passed into a 50 per cent, s o h . of hydrofluoric acid, t here is an energetic reaction accompanied by a flame in the mid& of the liquid. The reaction of fluorine with gaseous or aq. soln. of hydrogen chloride, bromide, or iodide, is accompanied by flame. Most of the haloids of the metalloids are attacked with great energy by fluorine a t ordinary temp. Fluorine does not unite with argon even if a m ixture of the two gases be heated or sparked. Neither nitrogen or nitrous oxide, N,O, nor nitrogen peroxide, NOz, is a ttacked by fluorine at ordinary temp. 0. R uff and J. Zedner also found no reaction occurred a t the temp. of the electric arc between fluorine and nitrogen. Even a t a dull red heat nitrous oxide remains unatbcked by fluorine, but by sparking a mixture of fluorine and nitrous oxide, a mixture of nitrous oxide, nitrogen, and oxygen is formed, but no nitrogen oxyfluoride.13 A little nitric oxide, NO, unites with fluorine a t ordinary temp. ; t he reaction is attended by a pale yellow flame, and a volatile oxyfluoride is formed ; b ut if the nitric oxide be in large excess, it is simply broken down into nitrogen and oxygen, and the excess of nitric oxide forms nitrogen peroxide. Ac-cording to H , Moissan and P. Lebeau, if the fluorine be in excess, a t the temp. of liquid oxygen, a white solid is formed which, as the temp. rises, changes into a colourless liquid, boiling above SO0, and which furnishes on fractionation nitroxyl or nitryl fluoriay, N02F. Pluorine decomposes ammonia with inflammation ; and a mixture of the two gases explodes. Phosphorus inflames in fluorine gas forming the pentafluoride, PPS, if the fluorine be in excess ; a nd the trifluoride, PF3, if the phosphorus be in excess. Arsenic forms the trifluoride, ASP,, with inflammation. Similarly with antimony ; b ut bismuth is only superficially attacked. Both phosphorus and arsenic react with incandescence with liquid fluorine, but antimony remains unaltered. P~OSP~OX'+US pentoxide, P 20S,is decomposed a t a red heat forming the fluoride and oxyfluoride ; phosphorus tri- and penta-chloride a re attacked with the production of flame ; neither phosphorus pentafluoride nor phosphorus oxyfluoride i s attacked. Arsenic trioxide and arsenic trichloride a re attacked. Arsenic trifluoride, ASP,, absorbs fluorine, and the heat generated during the absorption led H. Moissan to suggest that some unstable a rsenic pentaJluoride is formed. Both boron a nd silicon u nite with fluorine gas energetically and with incandescence, forming in the one case boron trifluoride, BP3, and in t he other, silicon tetrafluoride, SiF4. Boric oxide a nd silica r eact energetically in the cold. Boron trichloride, BC13, a t ordinary temp., and silicon tetrachloride, SiCL, above 40°, both react with fluorine. Dry fluorine does not attack glass, for H. Moissan kept dry fluorine in glass vessels for two hours a t 100°, without appreciable attack. Hydrogen fluoride behaves similarly. The-slightest trace of moisture is sufficient to activate either gas. Dry lampblack becomes incandescent in fluorine ; mood charcoal fires spontaneously ; t he vigour of the reaction is reduced a t low temp., for boron, silicon, and carbon are not attacked by liquid fluorine. If ~ owdered charcoal or soot be allowed to fall into a vessel containing liquid fluorine, the particles T HE HALOGENS 13 become incandescent as they drop through the vapour, but the glow is quenched when the particles reach the liquid. The demer forms of carbon require a temp. of 50" to 100" before they become incandescent ; r etort carbon requires a red heat ; and the diamond is not affected a t that temp. Soft charcoal is quickly ignited in contact with the gas. The product of the reaction is usually a mixture of different carbon fluorides, but if the temp. of the reaction be kept low, carbon tetrafluoride alone is formed. H. Moissan 14 also found that fluorine acts on calcium carbide a t ordinary temp. giving calcium fluoride and carbon tetrafluoride. Carbon monoxide and dioxide a re not attacked in the cold ; carbon disulphide, C8,, inflames forming carbon and sulphur fluorides ; carbon tetrachloride, CCl,, reacts a t temp. exceeding 30"forming chlorine and carbon tetrafluoride ; cyanogen is decomposed a t ordinary temp. with .the production of a white flame. According to W. L . Argo and coworkers, the unlighted gas issuing from a Bunsen's burner is immediately ignited by fluorine. According to B. Humiston, acetone i n an open vessel takes fire ; chloroform forms chlorine, phosgene, and carbon fluorides. With phosgene, a compound which appears to be carbonyl fluoride, COP2, was formed. The action of fluorine on ethylene tetrachloride, C2C14,is symbolized : C2CL4+2F2=C2Y4+2Cl2, followed by Cl,+C2C&=C2Cls, and C2P4=CP4+C. The metals a re in general attacked by fluorine a t ordinary temp. ; many of them become coated with a layer of fluoride which protects them from further action. These remarks apply to the metals : aluminium, bismuth, chromium, copper, gold, iridium, iron, manganese, palladium, platinum, ruthenium, silver, tin, zinc. The formation of a protective skin of fluoride renders it possible to prepare fluorine in copper and platinum vessels a t ordinary temp. Lead is slowly converted into the f white fluoride at ordinary temp. I t he temp. be raised, nearly all the metals are vigorously at tacked with incandescence-for example; with tin and zinc, the ignition temp. is about looo, and iron and silver, a t +bout 500". Gold and platinum are slowly converted into their fluorides a t about 500" or 600". The metals of the alkalies and alkaline earths, thallium, and magnesium are converted with incandescence into their fluorides. Many of hhe metals which.in bulk are only attacked slowly, are rapidly converted into fluorides if t hey are in a finely divided condition. Thus fluorine forms a volatile fluoride with powdered molybdenum i n t he cold, but a lump of the metal is not attacked ; tungsten is attacked a t ordinary temp., and also forms a volatile fluoride.; electrolytic uranium, i n fine powder, is vigorously attacked and burns, forming a green volatile hexafluoride. If niobium (columbium) or tantalum be warmed, the pentafluorides are formed. Liquid fluorine has no action on many of the metals-e.g. iron. I mercury be quite still, a protecting f f layer of fluoride is formed, but i t he metal be agitated with the gas, it is rapidly converted into the fluoride. The chlorides, bromides, iodides, a nd cyanides a re generally vigorously attacked by fluorine in the cold ; sulphides, nitrides, a nd phosphides a re attacked in €he cold or may be when warmed a little ; the oxides of the alkalies and alkaline earths are vigorously attacked with incandescence ; t he other oxides usually require to be warmed. The sulphates usually require warming ;, t he nitrates generally resist attack even when warmed. The phosphates a re more easily attacked than the sulphates. The carbonates of sodium, lithium, calcium, and lead are decomposed at ordinary temp. with incandescence, but potassium carbonate is not decomposed even at a dull red heat. Fluorine does not act on sodium borate. Most of these reactions have been qualitatively'studied by H. Moissan,l5 and described in his monograph, Le j uor et ses compose's ( Paris, 1900). Atomic and molecular weight of fluorine.-The combining weight of fluorine has been established by converting calcium fluoride, potassium fluoride, sodium fluoride, etc., into the corresponding sulphates. I n iilustration, J. B . A. Dumas (1860) found that one gram of pure potassium fluoride furnishes 1-4991 gram of potassium sulphate. Given the conlbining weights of potassium 39'1, sulphur 32.07, oxygen 16, it follows that if x denotes the combining weight of fluorine with 14 INORGANIC AND THEOBETICAL CHEMISTRY 39.1 grams of potassium, 1 : 1*4991=21(P : K 2S04=2(39-l+x) : 174.27 ; or, 2=19. H. D avy l 6m ade the first attempt in this direction in 1814 by converting fluorspar into the corresponding sulphate. His result corresponds. with an at. wt. 18.81. J. J . Berzelius (1826) also employed a similar process and obtained first the value 19.16 and later 18.85. P. Louyet, in 1849, employed the same process, taking care that the particles of fluorspar did not escape the action of the sulphuric acid by the formation of a protective coating of sulphate. P. L ouyet obtained 18.99 with native fluorspar, and 19.03 with an artdcial calcium fluoride. I n 1860, J . B.A. Dumas obtained the value 18.95 with calcium fluoride ; S. de Luca (1860), 18.97 ; H. Moissan (1890), 19.011. P. Louyet, J. B . A. Dumas, and H. Moissan also conyerted sodium fluoride into sodium sulphate and obtained ,respectively the values 19.06, 15-08, and 19.07. P. L ouyet and H. Moissan in addition converted barium fluoride into the sulphate and obtained respectively 19.01 and 19.02 ; a nd P. Louyet's value, 19.14, was obtained with lead fluoride. 0. T . Christensen (1886) treated ammonium manganese fluoride, (N&)2MnF5, with a mixture of potassium iodide and hydrochloric acid-one mol. of the salt gives a gram-atom of iodine. The liberated iodine was titrated with sodium thiosulph.attc The value 19.038 was obtained. J. Meyer (1903) converted calcium oxide into fluoride and obtained 19,035. D. J . McAdarn and E. F. S mith (1912) obtained 19.015 by transforming sodiunz fluoride into the chloride. E. F. S mith and W. K . van Haagen obtained 19,005 by converting anhydrous borax into sodium fluoride. E. Moles and T. Batuecas estimated the at. k t . of fluorine from trhe density of methyl fluoride, and found 18.998k 0.005 when ihe at. wt. of car5on is 12.000, a i d of hydrogen, 1.0077. The best determinations range between 18'97 and 19.14, and the best representahive value of t,he combining weight of fluorine is taken to be 19. No known volatile compound of fluorine contains less than 19 parts of fluorine per molecule, and accordingly this same number is taken to represent the at. wt. The vapour density of fluorine, determined by H. Moissan, is 1-31 ( air=l), that is, 28.755 ~ 1 ~ 3 1 = 3 7 ~ 7 ( H ~ = The molecule of fluorine is therefore represented by F2. 2). Fluorine is assumed to be univalent since it forms fluorides like KF, NaP, ~ t c . with univalent elements and radicles ; CaF2, BaF2, etc., with bivalent radicles, etc. As indicated in connection with hydrogen fluoride, etc., there is, however, the great probability that fluorine also exhibits a higher valency in the more complex com~ o u n d like KF.HF, A1F3.3NaF, etc.17 This also agrees with J . Thomsen's observas tions on the heat of the reaction between the acid and silica. REFERENCES, 1 H . Moissan, C m p t . Rend., 109. 861, 1889 ; 138. 7 28, 1904 ; B. B rauner, Zeit. anorg. Chem., 1. 1 , 1894 ; J. Sperber, i h, 14. 164, 374, 1897. 2 H . Moissan and J. Dewar, Compt. Rend., 124. 1202, 1897 ; 125. 505, 1897 ; 136. 7 85, 1903. a J. Sperber, Zeit. anory. Chem., 14. 164, 1897. 4 J . H . Gladutone, Phil. Trans., 160. 2 6, 1870 ; A mer. Journ. Science, ( 3 ) , 29. 5 7, 1885 ; G. Gladstone, Phil. May., (5), 20. 483, 1885 ; J . H . and G . Gladutone, ib., ( 5 ) , 31. 9 , 1891 ; F . S warts, RuW. A d . BeEgique, ( 3), 34. 293, 1897 ; Mkm. COW. Acid. BeLiqzle, 61. 1901 ; C . C uthbertsonand E . B. R. R ideaux, Phil. Trans., 205. A , 319,1905. 6 H . Moissan, Compt. Rend., 109. 937, 1880 ; C . de Wattcville, ib., 142. 1078, 1906; G . S alet, An'n. Chim. P hys., ( 4),28. 34, 3.873. 6 P. Pascal, Compt. Rend., 152. 1010, 1911 3 Bull. Soc. Chim., (4), 9. 6 , 1911. 7 W.Abegg and C . E . I mmerwahr, Zeit. phys. Chem., 32. 142, 1900. F . K ohlrausch, Wied. Ann., 66. 7 86, 1898. B F . W . Hkirrow, Zeit. anorp. Chem., 33. 2 5, 1903 ; M. G. L evi, Chem. Ztg., 30. 4508, 1906; 1 . G. L evi and F . Ageno, Atti Accnd. Lincei, ( 5 ) ,15. i i, 549, 615, 1907. 1 10 H . Moiusan and J . Dewag, Compt. Rend., 1%. 1202, 1894 ; 136. 641, 785, 1903. 11 0 R u f f and J , Z edner, Ber., 42. 1037, 1909 ; G. Gallo, A tti Accad. Lincei, (5), 19. i , 295, . 753, 1910. l 2 H . Moissan and P. L ebeau, Ann. Chim. Phys., ( 7 ) ,2 . 5 , 1902. 6 l a H . Moissan and P . L ebeau, Compt. Rend., 140. 1573, 1905. l 4 H . Moiusan, Le four &ctriqz&, Paris, 1897 ; L ondon, 1904 ; Compi. Rend,, 110. 2 76, 1800 ; THE HALOGENS B. IIumiston, Journ. Phys. Cltem., 23. 572, 1919; W. L. A go, P. C. M ather~, . Humiston, and B C. 0.Anderson, ib., 23. 348, 1919. II. Moissan, Ann. Chim. Phys., (6), M . 224, 1891. Davy, Phil. Trans., 104. 64, 1814; J. 4. Berzelius, P oyy. Ann., 8. 1, 1 826; Ann. Chim. Phy~., 27. 53, 167, 287, 1824 ; P. Louyet, ih., ( 3), 25. 291, 1849; E. F r h y , i b., ( 3), 47, 15, (2), 1856; J. B. A. Dumas, ib., {3), 55. 129, 1859; S. de Luca, Compt, R e d . , 51. 299, 1860; H. Moissan, ih., Ill. 70, 1890 ; 0. T. Christensen, Journ. prakt. Chem., (2), 34. 41, 1886 ; ( 2 ) , 35. 541, 1887 ; 5 J. Mcycr, Zeit. anory. Chem., 36. 313, 1903 ; D. J. McAdam and E. 3 S mith, Joum. dmer. C7hem. ' . Soc., 34. 592, 1912 ; E. Moles and T. Batuecas, sourn. Chim. Phys., 17. 537, 1919; E. F. S mith and W. K ,v an Haagen, The Atomic Weights of Bormz and Pluorinc, Washington, 1918. C. W. Blomstrand, Die Chemie der Jctztzeit, Heidclberg, 210, 340, 1869 ; J. Thornsen, Vied. Ann., 138. 201, 1869; 139. 217, 1870; Ber., 3. 583, 1870. l6 " 5 5. T he Occurrence of Chlorine, Bromine, and I odine Chlorine,-Chlorine does not occur free in nature, but hydrogen chloride has been reported in the fumes from the fumeroles of volcanic districts,l Vesuvius, Hecla, eto. D . Pranco reported that the gases given off by the flowing lava of Vesuvius, during solidification, contained much hydrogen chloride, and the same gas has been found as an inclusion in minerals. Hydrogen chloride is also found in the springs and rivers of volcanic districts-c.g. the Devil's Inkpot (Yellowstone National Park), Paramo de Ruiz (Colombia), Brook Sungi Pait (Java), the Rio Vinagre (Mexico), eto. The latter is said to contain 0.091 per cent. of free hydrocliloric acid which i o eotiinated to be eq. to 42,150 k g r m . of HCI per diem.2 J. B. J. D. Boussingault suppooeo this acid to be derived from the decomposition of sodium chloride by steam. Combined chlorine is a n essential constituent of many minerals-there are sal ammoniac (ammonium chloride) ; sylvine (potassium chloride) ; halite (sodium chloride) ; chlorocalcite, CaC1, ; cerargyrite or horn silver, AgCl ; calomel, HgCl ; terlinguaite, Hg20C1 ; eglestonite, IIg,C1,0 ; m,olysite, FeCl, ; erythrosiderite, 2KC1.FeCl3.H,O ; rinneite, 3KCl.NaCl.FeC1, ; kremeruite, 2KC1.2NH4C1.2FeC13.3H,0 ; lawrenci8e, FeC1, ; douglnsite, 2KC1.FeCl,.2HaO ; accechite, MnCl, ; c dunnite, PbC1, ; rrtatlockite, PbCl,.PbO ; penfieldite, 2PbC12.Pb0 ; mendipite, PbC1,.2PbO ; Eaurionite, PbCl,.Pb(OH), ; fiedlerite, 2PbCl,.Pb(OH), ; r afaelik or paralaurionde, PbCI(0H) ; nantokite, CuCl ; melanotfiallite, CuCl,.CuO.H,O ; hydromelanothallite, CuCl,.Cu0.2H,O ; atncamite, Cu,Cl(OH), ; percylite, PbCuCl(OH), ; boleite, ,,4H,O ; 3PbCuCldOH) ,.AgCl ; footeite, CuC1,.8Cu(OH) ,.4H,O ; taltingite, CuCI2.4Cu(OH) a;felite, C U C ~ ~ . ~ C U ( O H ) ; .cH ~ O ~ umengegte, 4PbC1,.4Cu0.5H20 ; pseudobolite, 6PbCl2.4CuO. 6H,O ; phosgenite, Pb2C1,C03 ; daubreite, BiC13.2Bi,0,.3H.& ; a n d i n some Stassfurt minerals, carnallite, KC1.MgCl2.6H,O ; bischofite, MgC1,.6H,O ; tachhydrite, CaC1,.2MgCl,. 12H20; boracite, MgCl,.2Mg3B,0,, ; ebc. Chlorine also occurs in mineral phosphates e.y. i t partially replaces fluorine in t h e c hloroapatites-pyromorphite, ( PbC1)Pb4(P0,)3; mimetite, (PbCl)Pb4(As04)3 a n d uanadinite, (PbC1)Pb4(Y04)3. It o ccurs in pyrosmalite, ; H6(Fe,Mn),8i40,,C1 ; sodcclile, Na,A13Si,01,C1, a n d other silicate minerals. - Chlorides occur in sea, river, and spring water, and small quantities in rain water. Theaohes of $ants and animals contain some chlorides. The gastric juices of animals contain chlorides as well as free hydrochloric acid. The 0.2 to 0.4 per cent. of free hydrochloric acid in the gastric juices of man is thought to play an important r6Ee i n the digestion of food.3 Sodium chloride occurs in blood and in urine ; flesh contains potassium chloride ; while milk contains both of the alkaline chlorides, with potassium chloride in large excess. According to R . W a n a ~ hblood contains ,~ 0-259 per cent: of chlorine, and serum, 0.353 per cent. ; a nd according t o A. J. Carlson, J. R. Greer, and A . B. Luckhardt, there is still more chIorine in lymph. T. Gassmann found human teeth t o contain 0.25-to 0.41 per cent. of combined chlorine, and the teeth of animals rather less. Bromine.--J. H. L. Vogt5 estimates that bromine occupies about the 25th place in t he list of elements arranged in the relative order of their abundance ; a nd that the total crust of the earth has about 0-001 per ,cent. of bromine-the solid portion 0.00001 per cent. The ratio of bromine to chlorine is about the same in sea water and in the solid crust, and amounts to 1 : 150. The ratio of chlorides t o 16 INORGANIC AND THEORETICAL CHEMISTRY bromides in marine waters of the globe is almost consta,nt, excepting land-locked seas like the Black and Ba'ltic Seas. It has been estimated that there are about 120000,000000 tons of bromides present in all the marine waters of our globe. The salt lake south of Gabes in Tunis has been worked since 1915 for bromine and potash. There is no record of the occurrence of free bromine in nature, but R. V. MatteucciB has reported the presence of hydrogen bromide in the fumeroles about Venuvius. Bromine usually occurs as an alkali bromide or as silver bromide with more or less silver chloride aiid silver iodide. Thus, the Chilean mineral bromargyrite, bromyrite, or bromite approximates to AgBr ; chlorobromosilver or embolite, Ag(C1,Br) ; a nd iodobromite, or iodoembolite, Ag(Cl,Br,I). Small quantities of these minerals occur in other places. Bromine has been reported in rock salt, meerschaum, and in French phosphorites by F . Kuhlmann ;7 in Silesian zinc ores by C. F. Mentzel and M . Cochler ; i n Chili saltpetre by H. Griineberg ; in coal by A. Duflos ; and in ammonia water and artificial sal ammoniac, by C. Mkne.and others. The Stassfurt salts contain bromides, indeed, these salts are the chief source of commercial bromine.8 Perhaps twothirds of the world's annual consumption of bromine (1,500,000 kilos) was obtained in Germany from these deposits. According to H. E . Boeke, the bromine in the Stassfurt deposits is there in the form of a bromo-carnallite, MgBrz.KBr.6Hz0, in isomorphous mixture with carnallite MgClz.KC1.6HzO, L. W. Winkler reported that the potash liquors of sp. gr. 1.3, from Stassfurt, Mkcklenberg, and Hainleite respectively, have 7.492,5.398, and 3.691 grms. per litre. Bischofite and tachhydrite from Vienenburg are the richest in bromine and contain respectively 0'467 and 0.438 per cent. ; carnallite has 0.143 to 0.456 per cent. ; sylvine, 0.117 to 0'300 per cent. ; sylvinite, 0.085 to 0.331 per cent. ; H artsalz, 0.027 per cent. ; a nd langbeinite, 0'016 per cent. The presence of bromides has been detected in numerous mineral and spring waters. There is a long list of reported occurrences of bromine in mineral waters in different parts of the world arranged alphabetically in L. Gmelin and K. Kraut's Halndbuch der alnorganischeln Chemie (Heidelberg, 1. ii, 218, 1909). The waters of Anderton (Cheshire), Cheltenham (Gloucester),Harrogate (Yorkshire), Marston, Wheelock, and Winsford (Cheshire) are in the list for England. Some of the brine springs--e.g. the Congress and Excelsior Springs of Saratoga, N.Y. ; N atrona (Wyoming) ; T arentum (Pennsylvania) ; Mason City, Parkersville, etc. (West Virginia) ; Michigan, Pittsburg, Syracuse, Pomeroy, etc. (Ohio)-contain so large an amount of bromine that in importance they are second only to the Stassfurt deposits as sources of commercial supply ; and they have played an important part in keeping down the price, and preventing the Stassfurt syndicate monopolizing the world's markets. The mineral waters of Ohio are said to contain the eq. of from 3.4 to 3.9 per cent. of magnesium bromide. Bromine is present in sea water. The mixture of salts left on evaporation of the of the Atlantic Ocean contains from 0.13 to 0.19 per cent. of bromine-presumably as magnesium bromide ; t he Red Sea, 0.13 t o 0.18 ; t he Caspian Sea, 0.05 ; and the Dead Sea, 1.55 to 2.72 per cent. of bromine.9 E. Marchand l o h as reported the presence of. traces of bromine in rain and snow. The ashes of many sea weeds and sea animals contain bromine-thus, dried Fucus vesiculosus contains 0.682 per cent. of bromine.11 Bromine has been reported in human urine, salt herrings, sponges, and cod liver oil ; b ut not in bone ash. Indeed, all products directly or indirectly derived from sea-salt or from Stassfurt deposits-in the present or in the past--contain bromine. It is also said to be an essential constituent of the dye Tyrian purple which was once largely obtained from a species of marine gastropod or mollusc. Iodine.-Iodine is perhaps the least abundant of the halogens. Although widely distributed, i t always occurs in small quantities. J. H. L. Vogt 12 estimates there is about 0.0001 per cent. of iodine in the earth's crust-the solid matter containing about 0.00001 per cent. ; a nd the sea, 0.001 per cent. A. Gautier's estimate of the iodine in the sea is about one-fifth of this. Iodine occupies the THE HALOGENS 17 28th place in the list of elements arranged in their relative order of abundance, SO t hat iodine has exercised no essential cbemical or geological influence on the earth's surface. The. sea appears to be the great reservoir of iodine. The ratio ~f bromine t o iodine in sea water and in the solid crust is approximately the same, uiz. from 1 : 10 to 1 : 12 ; a nd, in sea water, the ratio of chlorine to iodine is as 1 : 0.00012. L. W. Winkler reported 1 7 mgrms. of iodine (as iodide) per litre of a natural saline water from Mecklenburg, that is, about 340 times as much as in sea water ; a sylvinite mother-liquor from Alsace contained 0.5 mgrm. of iodine per litre. Iodine does not occur free in nature, although, according to J. A. Wanklyn,ls the waters of Woodhall Spa (Lincoln) are coloured brown by this element. R. V. Matteucci reported the occurrence of hydrogen iodide in the emanations of Vesuvius, and A. Gantier found iodine in the gases disengaged from cooling lava. Iodine occurs along with bromine in iodobromite, Ag(C1, Br, I ) ; in iodyrite, AgI ; marshite, CuI ; coccinite, HgI, ; a nd in schwartzembergite, Pb(1,' C1)2.2Pb0. According to A. Guyard,l4 the iodine-up t o about 0.175 per cent. in Chile saltpetreis present as sodium iodate, NaI03, and ,periodate, NaI04 ; H. Griineberg considers a double iodide o f sodium and magnesium is also present. Potash saltpetre also has been reported to contain potassium iodate, KI03, the caliche from which Chile saltpetre is extracted forms one of the most important sources of iodine ; i t contains about 0.2 per cent. of iodine, probably as sodium iodate. Iodine has also been reported in the lead ores of Catorce (Mexico) ;16 in malachite-0.08 to 0'40 per cent. (W. Autenrieth) ; Silesian zinc ores (C. F. Mentzel and ill. Cochler) ; t he clay shales of Latorp in Sweden fJ. G. Gentele) ; t he limestones of Lyon and Montpellier (G. Lembert) ; t he bituminous shales of Wurtemberg (G. C. L . Sigwart) ; t he dolomites of Saxony (L. R . Rivier von Fellenberg) : rock salt ( 0.H enry) ; t he phosphorites of France (F. Kuhlmlmn) ; t he phosphates of Quercy (H. ~ a s n e ; granites ) (A. Gautier) ; Norwegian apatite (A. Gautier) ; coal, and ammonium salts derived from coal ( A. Duflos) ; guano of C u r a ~ a o( H. Steffens) ; a nd the Stassfurt salt deposits (A. Prank)-although P. R inne and E. Erdmann failed to confirm A. Frank's results. The presence of iodides has aIeo been recorded in a number of spring waters, brines, etc.16 I n Great Britain it occurs in the waters of Leamington (Warwickshire), Bath (Somerset), Cheltenham (Glouceater), Harrogate ( Yo'rkshire), Woodhall Spa (Lincoln), Bonnington (near Leith), Shotley Bridge (Durham), etc. Iodine occurs in small quantities in sea water ; E. S onstadt 17 estimated that there is about one part of calcium iodate per 250,000 parts of sea water ; b ut acsording to A. Gautier, the iodine in the surface water of the Mediterranean Sea is found only in the organic matter which can be separated from the wafer itself by filtration ; but at depths below 800 metres, he found iodine t o be in water itself as soluble iodides. According to A. Gautier, also, the waters of the Atlantic contain 2.240 mgrrns. per litre ; a nd according to L. W. Winkler, the waters of the Adriatic Sea, 0.038 rngrm. per litre. A. Goebel reported 0'11 per cent. o f iodine in the salts from Red Lake (Perekop, Crimea) ; a nd H. Fresenius 0.0000247 per cent. in t h e waters of the Dead Sea. The amount is so small that analysts have usually ignored the iodine, or reported mere traces. Similar remarks apply to the brines from the waters of closed baains. Iodine has been reported in rain and anow,lB and A. Chatin found iodine universally present in small quantities in the atm., rain water, and running streams. A. Gautier reported in 1889 that the air of Paris contained less than 0.002 mgrm. of free iodine or an iodine compound in about 4000 litres ; b ut 100 litres of air in Paris contained 0.0013 mgrm. in a form insoluble in water, generally the spores of a l p , mosses, lichens, etc., suspended in the air. A. Gautier alss found sea air t o contain 0.0167 mgrm. of iodine per 100 litres. The amount of iodine in mountain air and the air of forests is less than in other parts. The iodine in the atm. is supposed t o be of marine origin. The presence of iodine as a normal constituent of the atrn. has been denied,lQ but A. Chatin'a conclusions were confirmed by J. A. Barral, A. A. B. Bussy, and A. Gautier. Marine animals and plants assimilate VOTA,IT. 0 INORGANIC AND THEORETICAL CHEMISTRY iodine from sea water ; most of the iodine can be extracted by water from the ash of these organisms. I t appears strange that the marine algae should select iodine from sea water and practically leave the bromine which is present in much larger proportions. The pelagic seaweeds ( a l p ) in favourable localities cover the ocean about the 10-fathoms line with dense fields of floating foliage ; t he littoral seaweeds grow nearer shore a t about the limit of extreme low tide. The deep-sea algw usually have a greater proportion of iodine than those which grow in shallow water. According to E. C. C. Stanford,zo the percentage amounts of iodine in a few dried plants are as follows : - LITTOR& (SHORE) SEAWEEDS Fucua fllium 0.089 Fucua digitatus 0.135 Fucua nodosus 0.057 Fucua serratus 0.085 Fucus vesiculosua 0 .001 Ulva umbilicalis 0.059 . . . . . . . . . . PELAGIC (DEEP-SEA)EAWEEDS S Nereocystis leutkeana Macrocystis pyrifera Pelagophycus porra Laminaria digitata Laminaria stenophylla Laminaria saccharins . . . . . . . . . . . . 0.521 0.205 0.24 0.374 0 .478 0 .255 A. M. Ossendowsky studied the algae employed in northern Japan for making iodine. According to J. Pellieux, seaweed grown in minter usually carries more iodine than that grown in summer ; t hat grown in the north more than that grown in the south ; and the younger parts of the algae more than in the older parts. Iodine has not been found in the gelatinous varieties of marine algae-4.g. the chondrus crispus or Irish moss, and the eucheum spinosum or agar-agar ; nor 'has it been found in the enderomrpha compressa, or common sea-grass. Some plants which grow near the sea--e.g. the salsola kali, or salt-wort of salt-marshes, from which baralp is made 2'-are almost free from iodine. Smaller amounts of iodine have been found in fresh-water plants than in land plants, and smaller proportions of potash are found in them also. According to A . Gautier,22 iodine must be a constituent of the chlorophyll or reserve protoplasm of plants because plants containing chlorophyll contain more than the algae and fungi which are free from chlorophyll. Iodine is found in tobacco (A. Gautier), and in beetroot ( M. J. Personne), and the potash derived from these products, as well as other plants, also contains iodine. It is found in fossil plants ; and hence also its occurrence in coal, and in the ammoniacal products derived from coal. Turkey sponge has 0.2 per cent., the honeycomb sponge 0.054 per cent., and according to F. Hundeshagen,zs the sponges from tropical seas contain up to 14 per cent. of iodine ; A. Fyfe, and K. Stratingh found none in corals ; b ut E. Drechsel isolated from certain corals what he considered to be iodogorgic acid, C4H81N02. Minute quantities of iodine have also been reported in nearlv all marine worms, molluscs, fish, and other marine animals which have been examined. For example, oysters have been reported with 0.00004 per cent. of iodine; prawns, 0'00044; cockles, 0.00214 ; mussels, 0'0357 ; salt herrings, 0-00065 ; cod-fish, 0.00016 ; a nd cod's liver, 0'00016 per cent. It occurs i n most fish oils-cod-liver oil, for instance, contains from 0'0003 to 0,0008 per cent. of iodine ; whale oil has 0*0001per cent. ; and seal oil, 0'00005 per cent. Iodine is a normal constituent of animals where it probably occurs as a complex organic compound. The iodine of the thyroid gland is present a s a kind of albumen containing phosphorus and about 9 per cent. o f iodine. This has been isolated by digesting the gland with sulphuric acid, and precipitating with alcohol. The iodine seems to play a most important part ip the animal economy. The proportion of iodine is smaller in young people than in adults, and the amount becomes less and less with the aged.24 According to J . J ustus, the amount of iodine in milligrams per 100 grms. pf t he various organs of human beings is : thyroid gland, 9.76 mgrms. ; liver, 1'214 ; kidney, 1.053 ; stomach, 0.989 ; ~tkin,0,879 ; hair, 0.844 ; nails, 0'800; prostate, 0'689 ; lymphatic gland, 0.600 ; spleen, 0560 ; testicle, 0'500 ; pancrow, 0'431 ; virginal uterus, 0'413 ; lungs, 0.320 ; nerves, 0.200 ; small THE HALOGENS intestine, 0.119 ; f atty tissue, traces. The proportion in the corresponding parts of animals is smaller. Only a very small proportion is found in blood and muscle. 0. Loeb 2hould find none in the brains, spinal marrow, fat, and bones. Iodine has been reported in wine and in eggs. E.Winterstein found no iodine in milk, cheese, or corn's urine ; b ut he found iodine in thirty-five phitnerogams-in beetroot, c,elery, lettuce, and carrots, but not in mushrooms or yellow boletus. REFERENCES. 1 a R. Bunsen, Liebig's Ann., 62. 1 , 1847 ; U. Franco, Ann. Chim, Phya., (41, 30. 87, 1873 J. R . J. D. Boussingault, Compt. Rend., 78. 453, 526, 593, 1874 ; A nn. C h h . Phys., (51, 2. 80, 1 874; F. A. Gooch and J . E . W hitfield, Bull. U S . Geol. Sur., 47. 80, 1888. 8 P. S ommwfcld, Biochem. Zeit,, 9. 352, 1908 : J. C hriutian~en, ib., 4 2.4, 60, 71, 82, 1912. 6. R . W anach, Chem. Centr,, ( 4 ) , I , 355, 1889; T , G assmann, Zeit. physiol. Chem., 55. 455, 1908; A. J. Carlson, J. R . Greer, and A. 11. L uckhardt, Amer. Journ. Phyaiol. 23. 91, 1008. 6 J. H. L. V ogt, Zeit. pakt. Geol., 225, 314, 377, 413, 1898 ; 1 0, 274, 1899 ; F. W . Clarke, Bull. Washingt~n Phil. Soc., 11. 131, 1889 ; Proc. Amer. Phil. Boc., 51. 214, 1912 ; W . Ackroyd, Chem. News, 86. 187, 1902. 6 R . V . Matteucci, C m p t . Rend., 129. 65, 1899. 7 F . K uhlmann, Conrpt. Rend., 75, 1678, 1872 ; C. p&ne, ib., 30. 612, 1850 ; C. F . M entzel, Kastner's Archiv., 12. 252, 1827 ; M . Cochler, ib., 13. 33G, 1828; A . Duflou, Arch. P h r m . , 49. 29, 1848; H . Griineberg, Journ. pakt. Chem., ( I ) , 80. 172, 1853. K. K u bierechey, Die dcubche Kaliindustrie, Hallo a. S., 1907 ; L. W . Winkler, Zeit. nngew. Ckm., 10. 95, 1897 ; H . E. B oeke, ib., 21. 705, 1908 ; Sitzber. Akad. Berlin, 439, 1908. * F.W , Clarke, The Data of Cfeochemistry, Wauhington, 1916. l6 E. Marchand, Journ. P h r m . Cliim., (3), 17. 356, 1850 ; I . G uareschi, Atti Accad. Tiwino, 47. 988, 1912. 1 1 P Marsson, Arch. Pharm., 6 . 281, 1851. . 6 l a J. H L Y ~ g tZeit. pakt. Geol., 225, 314, 377, 413,1898; 100, 274, 1899 ; F . W. C larke, .. , BuU. Washington Phil. Boc., 11. 131, 1889 ; Proc. Amer. Phil. Soc., 51. 214, 1912 ; L. W . W inkler, Zeit. angew. Chem., 29. 451, 1916 ;A. Gautier, Bull. Soc. Chim., ( 3 ) ,21. 456, 1899 ; W . Ackroyd, Ckm. New8,86. 187, 1902. l 8 J . A. W anklyn,Chem. News, 54. 300, 1886 ; R. V . Matteucci, Compt. Rend., 129. 65, 1899 ; A. Gautier, ib., 129. 66, 189, 1899 ; H . E r d m a m , Zeit. Naturwise., 69. 47, 1896 ; R. B rando~, Schweigger's Journ., 15. 32, 225, 1827 ; A. F. Bergeat, Zeit. prakt. Geol., 43, 1899 ; P.W . D afert, M omtsh., 69. 235, 1908. ' L Faure, Chem. Uaz., 13. 199, 1855 ; Brit. Pai!. No. 344, 1854 ; A . G uyard, Bull. S w . 1 . Chim., (2),22. 60,1874 ; V . A. Jacqnelain, Bull. Soc. Em., 54. 652, 185.5 ; H. Griineberg, Journ. ' prakt. Chem., ( l ) ,6 172, 1853. 0. W . A u t a i e t h , Zeit. phpio!. Chem., 22. 608,1897 ;Chem. Ztg., 23. 626,1899 ; C . F . Mentzel, Kastner's Archiv., l a . 252,1827 ; M . Cochler, ib., 13. 336,1828; J. G . Centele, Proc. AccuE. Stockholm, 131, 1848 ; G . Lembert, Journ. Phurm., ( 3), 19. 240, 15151 ; G . C. L. S igwart, Jahresb. Natureu. Wurlembq, 43, 1853 ; L. R. R ivier von Fellenberg, Bull. Soc. Vnud. Scien. Nat., 924, 1853 ; 0.Henry, Journ. Chim.'MU., ( 3 ) , 5. 81, 1849 ; F. Kuhlmann, Compt. Rend,, 75. 1678, 1872; T. Petersen, Jahrb. Min., 96,(1872 ; H . L ame, Bull. Soc. Chim., ( 3 ) , 2. 313, 1889 ; A. Gautior, Compt. Rend., 12$.66,1899; 132.932, 1901 ;H . S teffens,Zeitl anal. Chem., 19.60,1880; A. Dnflos, Arch. Pham., 4 29, 1848 ; A. F rank, Zeit. angew. Chem., 2 . 1279, 1907 ; P. R inne, ib., 2 9. 0 0. 1031,1907 ; E. E rdmann, ib., 21. 1693, 1908 ; H . E . B oeke, i6., 21. 705, 1908. I s L. G melin and K . K raut, Hadbzcch &r anorganischen Chemie, Heidelberg, 1. ii, 287, 1909. 17 E S onstadt, C h m . News, 25. 196,231,241,1872 ; 7 4,316,1896 ;A. G autier, C m p t . Rend., . 128. 1069, 1899 ; 129. 9, 1899 ; A. Goebel, Mtd. Chim. Phys., 5. 326, 1864 ; H . Freuenius, Verh. Gea. deut. Natu-rfoac'h. Aerete, 118, 1913 ; L.W . Winkler, Zeit. angew. Chem., 29. 205, 191 6. l 8 E . Marchand, Compt. Rend., 31. 495, 1850 ;A. G autier, ib., 128. 643, 1899 ; A. C hatin, ib., 31. 868,1850 ; 32. 669, 1851 ; 33. 529, 584, 1851 ; 34. 14, 51, 409, 519, 529, 584, 1952 ; 35. 4 6, 107,605,1852 ; 87. 487, 723, 958, 1853 ; 88. 83, 1854 ; 39. 1083, 1854 ; 4 .390, 1858 ; 50. 420, 6 1860; 51. 496, 1860 ; J . A. Barral. ib., 35. 427, 1852 ; A. A. B . B u q , ib., 30 537, 1860; 35. 508, 1852. . l 9 F . Gaxxi ov, Compt. Rend., 128. 884, 1899. 8 0 IT C. C. t anford, Chem. N e w , 85. 172, 1877; Dinqler's Journ., 226. 85, 1877 ; J. P ellieux, ib., 284. 216, 1879 ; A. G autier, Compt. Rend., 129. 189, 1899 ; F. J . Cameron, Journ. Franklin Iwt., 176, 346, 1913; A. M. Ossendowsky, Journ. R t m . Phys. Chem. ~SOC., 1081, 1906. 38. H. Davy, Phil. Tmns., 104. 487, 1814; A. Fyfe, Edin. Phil. Journ., I . 254, 1819. z2 A. Cautier, Compt. Rend., 128. 643, 1069, 1899; 129. 60, 189, 1899; M. J. Personne, ib., SO. 478, 1850. F Hund~hagen, . Zeit. angew. Chem., 8. 473, 189.5 ; K . S tratingh, Repert, Pharm., 15. 282, 1823 ; E. Dxechsel, Cetrtr. Pbpaiol., 9. 704, 1907; A. F yfe, &din, P hil. Jowm., 1. 254, 1819. % INORGANIC AND THEORETICAL CHEMISTRY 24 E. Baumann, Zeit. p Jyswl. Chem., 21. 3 19, 1895 ; 22. 1 ,1896 ; J . Justue, Y irchozu'~Archiv, 170. 501, 1903; 176. 1, 1904. a s 0.Loeb, Arch. E xp. Path., 56. 320, 1897 ; H . Zenger, Arch. Pharm., ( 3), 6. 137, 1876 ; E. Wintemtein, Zeit. physid. Chem., 1W. 51, 1919. f j 6 The . History of Chlorine, Bromine, and Iodine La vrai chirnia ne date que de l'emploi bien Ctabli des acides mindraux, qni sont veri( 1842). tables dissolvants des metaux.-F. HOEPER Sodium chloride mas known as salt from the earliest times. About 77 A.D., Pliny, in his Naturalis Historice (33. 251, described the purification of gold by heating it with salt, misy (iron or copper sulphate), and schistos (clsy). This mixture would give off fumes of hydrogen chloride. The attention, however, was focussed on the effect of the treatment on the metal ; no notice was taken of the effluvia. I n the Akhimia Gebwi (Bern, 1545)-supposed to have been written in the thirteenth century or afterwards-there is an account of the preparation of nitric acid by distilling a mixture of saltpetre, copper sulphate, and clay ; and Geber adds that the product is a more active solvent if some sal ammonincus is mixed with the ingredients. Thus Gebelr prepared aqua regia. Raymond Lully 1 called the former aqua salis nitri, and the latter aqua salis armoniaci; Albertus Magnus2 called the first aqua prima (nitric acid), the second aqua seeunda (aqua regia). J. R . Glauber (1648) 3 prepared aqua regia by distilling nitric acid with common salt. While the Arabian alchemists were probably acquainted with the mixture of nitric and hydrochloric acid known as aqua regia, there is nothing to show that they were acquainted with hydrochloric, acid. The method of making hydrochloric acid, first called spiritus salis, dates from the end of t he sixteenth century. Although there is no record, this acid was probably made earlier than this because it was the custom of the then chemists to collect the products of the distillation of mixtures o f various salts and earths ; and the necessary ingredients were in their hands. For example, in preparing aqua regia it merely required the substitution of sal naturus for nitmcm-which, as Pliny said, " do not greatly differ in their properties "-to furnish spiritus salis. It is almost inconceivable that this was not done. The preparation of the acid by distilling salt with clay id mentioned in the Alchmia (Francof~rti~ 1595) of A. Libavius; and in the Triumphwagen des Antimonii (Leipzig, 1624) of the anonymous writer Basil Valentine, J. R. Glauber (1648) described the preparation of spiritus salis by distilling common salt with oil of vitriol or with alum. J. R . Glauber also described the salient properties of t his acid, and especially remarked on its solvent action on the metals. He mentioned that silver resists it,s action. H. Boerhaave knew that lead resists the action' of, the acid ; R . Boyle referred to the action of the acid on soh. containing silver, and he noted.that the salt which this acid forms with the alkalies effervesces and fumes when treated with sulphuric acid. Stephen Hales (1727) noticed that a gas very soluble in water is given off when sal ammoniac is heated with sulphuric ac,id, and, in 1772, J. Priestley collected the gas over mercury and called it marine aeid air in reference to its production from sea-salt. For a similar reason, the French term for the acid was acide marin. In A. L . Lavoisier's nomenclature (1787), the acid was designated aeide nturiatique, or muriatic acid ; and after H. Davy's investigations on chlorine, the name was changed to acide chlorhydrique or hydroehbric aeid. There can be little doubt that the corrosive, suffocdting, greenish-yellow fumes of chlorine must have been known onwards from the thirteenth century by all those who made and used aqua regit~--e.g. . R . Glauber's rectified spirit of s alt J mentioned above. Early in the seventeenth century, J. B. van Helmont mentioned that when sal marin (sodium chloride) or sal armeniacus (ammonium chloride) and T H E HALOGENS 21 aqua chrysulca (nitric acid) are mixed together, a $atus incoercz3ik is evolved. J. Glauber (1648) also appears t o have obtained a similar gas by heating zinc chloride and sand ; h e also said that by distilling spirit of salt with metal oxides, he obtained in the receiver a spirit the colour of fire, which dissolved all the metals and nearly all minerals. This liquid-was no doubt chlorine water ; J. R . Glauber called it rectified spirit o salt, and he said that it can be used for making many products f useful in medicine, in alchemy, and in the arts. He gives an example by pointing out that when treated with alcohol, spirit of salt furnishes oleum vini, which is very agreeable, and an excellent cordial. The meaning of these observations was not understood until C. W. Scheele published his De magnesia nigra 4 in 1774. C. W. Scheele found that when hydrochloric acid is heated with manganese dioxide, a yellowish-green gas, with a smell resembling warm aqua regia, is given off. C. W. Scheele's directions for preparing the gas are : Common muriat.ic acid is t'o be mixed with levigated manganese ( i.e. pyrolusite) in an quantity in a glass retort, which is to be put into warm sand, and a glass receiver applied: capable of containing about, 12 oz. of water. Into the receiver put about 2 drms. of water ; the joints are to be luted with a p i ~ c e blotting paper tied round them. I n a quarter of of an honr, or a little longer, a qnantity of the acid, going over into the receiver, gives the air contained in it a yellow colour, and then it is to he separated from the retort. He remarked that the gas is soluble in water ; t hat it corroded corks yellow as i they had been treated with nitric acid; that it bleached paper coloured with f litmus ; t hat it bleached green vegetables, and red, blue, and yellow flowers nearly white, and the colour was not restored by treatment with acids or alkalies; that it converted mercuric sulphide into the chloride and sodium hydroxide, common salt ; etc. C. W. Scheele considered the yellowish-green gas to be muriatic acid freed from hydrogen (then believed to be phlogiston) ; accordingly, in the language of his time, it was called dephlogisticated muriatic acid. A. L. Lavoisier (1789) 5 named the gas o x~muriatic aid, or oxygenated muriatic acid, because he considered i t to s be an oxide of muriatic (i.e. hydrochloric) acid ; a nd, consistent with his oxygen theory of acids, A . L. Lavoisier considered muriatic acid to be a compound of oxygen with an hypothetical muriatic base-murium ;t his imaginary element was later symbolized Mu ; hydrochloric acid was symbolized MuOz ; a nd C. W. Scheele's gas, Mu03. Hence, added A . L . Lavoisier, muriatic and oxymuriatic acids are related t o each other like su!phurous and sulphnric acids ; a nd he was a t first inclined t o call the one acide muriateux, and the other acide mur<atipue. This certainly seemed to be the most plausible explanation of the reactions. Lavoisier's hypothesis was supported by an observation of C. L. Berthollet (1785),6 that if the manganese dioxide be deprived of some of its oxygen bv calcinatiou, i.t furnishes a much smaller quantity of Scheele's gas. Hence, concluded C. L. Berthollet, " i t is to the vital air (oxygen) of the manganese dioxide, which combines with the muriatic acid, that the formation of dephlogisticated marine acid is due." H e did not succeed in oxidizing muriatic acid to oxymuriatic acid. because he considered that the elastic state of muriatic acid gas prevents it uniting directly with oxygen. However, C. L. Berthollet supposed that he had succeeded in decomposing dephlogisticated marine acid into muriatic .acid and oxygen, for he noticed that an aq. soh. of C. W. Scheele's gas-the so-called oxymuriatic acid-when exposed to sunlight, gives off bubbles of oxygen gas, and forms muriatic acid. In 1800, W. H enry7 passed electric sparks through muriatic acid gas and obtained a little hydrogen which he supposed to come from the moisture in the gas ; on sparking a mixture of oxygen and muriatic acid gas he obtained a little oxymuriatic acid gas which he supposed was formed by the electric sparks decomposing Rome of t he moisture in the muriatic acid gas into oxygen, and the union of the oxygen with the muriatic acid gas to form oxymuriatic acid gas. The experiments INORGANIC AND THEORETICAL CHEMISTRY of J. L . Gay Lussac and L. J. ThBnard,8 and the earlier experiments (1809) of H. D avy,Q were complicated by the assumption that muriatic acid gas contains water, or the principles which constitute water, intimately combined. According to J. L. Gay Lussac and L. J. T h h a r d , muriatic acid gas and oxygen are formed when a mixture of steam and oxymuriatic acid is passed through a heated porcelain tube. It was here assumed that the muriatic acid gas in virtue of its great affinity for water, leaves the oxygen with which it is combined in oxymuriatic acid gas, and combines with the water to form muriatic acid gas. L. J. Gay Lussac and L . J. Thenard also found that by heating a mixture of hydrogen and oxymuriatic acid gas, there is a " violent inflammation with the production of muriatic acid " ; t hey also showed that a mixture of equal volumes of hydrogen and oxymuriatic acid gas does not change in darkness, but in light, the mixture is transformed into muriatic acid ; a nd in bright sunlight, the combination is attended by a violent explosion. It was assumed in these experiments that water is formed by the hydrogen removing the necessary oxygen from oxymuriatic acid gas, and leaves behind fiuriatic acid gas which is eager to combine with water. J. L . Gay Lussac andL. 3. T h h a r d also tried t o deoxidize oxymuriatic acid, so as to isolate the hypothetical muriatic base of A . L . Lavoisier, by passing the dry gas over red-hot carbon, but when the carbon was freed from hydrogen no change was observed even though " urged to the most violent heat of the forge." I n any case, the attempt to separate from oxymuriatic acid anything but itself was a failure. While favouring Lavoisier's hypothesis, J. L . Gay Lussac and L. J. T henard added : " t he facts can also be explained on the hypothesis that oxymuriatic acid is an elementary body." Here, then, are two rival hypotheses as to the nature of oxymuriatic acid-the yellowishgreen gas discovered by ScheeIe ! According to C. W. Scheele's hypothesis, oxymuriatic acid=muriatic acid less hydrogen ; according ,to C. L. Berthollet and A. L. Lavoisier's hypothesis, oxymuriatic acid=muriatic acid plus oxygen. I n his Researches on the Oxymuriatic Acid (1810), H. D avy described the attempts which he- made, without success, to decompose oxymuriatic acid gas. He a.lso found that when dried muriatic acid gas is heated with metallic sodium or potassium, the metallic muriate and hydrogen are formed, but neither water nor oxygen is obtained. Hence, no oxygen can be found in either muriatic acid gas or oxymuriatic acid gas, H. Ziiblin also tried in 1881, and likewise failed to decompose chlorine. Accordingly, H. D avy claimed that C. W. Scheele's view is an expression of the facts, while Lavoisier's theory, though " beautiful and satisfactory," is based upon a dubious hypothesis-uiq. the presence of oxygen in gases where none can be found. The definition of an element will not permit us to assume that oxymuriatic acid is a compound, because, in spite of repeated efforts, nothing simpler than itself has ever been obtained from the gas. I n order to avoid the hypothesis implied in the term oxymuriatic acid, H. Ditvy proposed the alternative term chlorine and symbol C1-from the Greek XXwpds,green. The term chlorine is thus " founded upon one of the obvious and characteristic properties of the gas-its colour.? According to H. Davy's theory, C. L . Berthollet's observation on the action of oxymuriatic acid gas on water, and J. L . Gay Lussac and L. J. Thhnard's observation of the action of steam on oxymuriatic acid gas, are explained b y the equation: 2H20+2~l2=4HC1+O2 ; t hat is, the oxygen comes from the water, not from the chlorine. Similarly, the formation of. chlorine by the action of oxidizing agents upon hydrochloric acid is due to the removal of hydrogen. I n symbols : 4HC1+02=2H20+2C12. H. D avy also showed that muriatic acid gas when adequately dried contains nothing but hydrogen and chlorine, and he showed that when the gas is decomposed by heating it with mercury, potassium, zinc, or tin, a chloride of the metal is formed, and one volume of hydrogen chloride furnishes half a volume of hydrogen-the other half volume must be chlorine. H. D avy summarized his conclusions from these and other experiments : " There may be oxygen in oxymuriatic gas, but I can find none " ; t he hypothesis that oxymuriatic acid is a simple substance, and that muriatic acid is a compound of this substance with hydrogen, T H E HALOGENS 23 explains all the facts in a simple and direct manner ; t he alternative hypothesis of the French school rests, in the present state of our knowledge, on hypothetical grounds. The French hypothesis died a lingering death. J. J. Berzelius in 1813 lo tried to argue.that the French schools were right ; he even expressed hia surprise that Davy's hypothesis " could ever gain credit." J. J. Berzeliua seems to have misunderstood H, Davy's experiments, but he too accepted Davy's conclusions a few years later. H , Davy's theory is orthodox to-day. New facts as they arrived have fallen harmoniously into their places and arranged themselves about Davy's theory as naturally as do the particles of a salt in soIution about the enlarging nucleus of a crystal. Iodine.-During the Napoleonic wars, nitre beds were cultivated in various parts of France, and from these saltpetre was obtained artificially. About 1811, Bernard Coudois, a manufacturer bf saltpetre, near Paris, used an aq. extract of varec or kelp for decomposing the calcium nitrate from the nitre beds ; he noticed that the copper vats in which the nitrate was decomposed were rapidly corroded by the liquid, and he traced the effects to a reaction between the copper and an unknown substance in the lye obtained by extracting the varec or kelp with water, B. Courtois isolated this new substance and ascertained its more obvious properties. In his paper entitled D6cduverte d'une substance nouvelb dam le vareck, and published about two years after his discovery,l"e said : The mother-liquors of the lye obtained from varec contain a of a singular and curious substance. I t can easily be obtained. sufficient to pour sulphuric w i d upon the mother-hquid and to connected with a receiver. The new substance which, on the addition of the sulphuric wid, is at once thrown down as a black powder is converted on heating into a vapour of a superb violet colour ; t his vapour coridenaes in the tube of the retort and in the receiver i n the form of b r i l h n t crystalline plates, having a lustre equal to that of crystallized lead sulphide. On washing these plates with a little distilled water the substance is obtained in a state of purity. The wonderful colour of its vapour suffices to distinguish i t from all other substances known up to the present time, a n d i t has further remarkable properties which render its discovery of the greatest interest. B. Courtois commu~iicated tidings of this discovery to . Clement and J. B. Nsormes, and they published some results of their study of this new aubstance ; early in December, 1813 ; 12 a few days later J. 1 . Gay Lussac, who also had received some of B. Courtois' preparation, gave a preliminary account of some researches on B. Courtois' new substance, a t a meeting of 1'1mtitut 1mperiaZ de France on Dec. 6th, 1813. I n this communication, J. 1;. Gay Lussac demonstrated some striking analogies between Courtois' preparation and chlorine ; h e made clear its elementary nature ; a nd designated it iode, t he French eq. of its present namefrom the Greek locr8ijs, violet. Its hydrogen compound was prepared and likewise given the very name.which it has to-day. J. L. Gay Lussac sttid : After this account one can only compare iode with clilore, and the new gmeous acid with T he phenomena of which we have just spoken can all be explained muriatic wid. . either by supposing that iode i s a n element and that i t formsan acid when i t combines with hydrogen, or by supposing that the latter acid is a compound of water with a n unknown snbstance, and that iode is this snbstance combined with oxygen. Considering all the facts recounted, the first view appears more probable than the other, and serves a t the same time to give probability to that hypothesis according to which oxygenated muriatic acid is regarded as a simple body. Adopting this hypothesis hydriodic acid appears 8 suitable name for the new acid. .. O n Dec. 20th, 1813, J. 1;. Gay Lussac 13 r ead a further memoir on the combinations of the new element with oxygen ; o n March 21st, 1814, J. J. Colin and H. G. de Claubry communicated observations made in J. L. Gay Lussac's laboratory on the action of iodine on organic compounds ; by June 4th, 1814, L. N. Vauquelin had studied the action of iodine on ammonia, iron, mercury, and alcohol ; a nd finally,' on Bug. lst, 1814, J. L. Gay Lussac communicated his famous 1Me'moire sur Z'iode.14 I n thig 24 INORGANIC AND THEORETICAL CHEMISTRY paper, said P . D. Chattaway,l5 the whole chemical behaviour of iodine is described in such a masterly fashion that it remains to this day a model of what such an investigation ought to be. H. Davy played a not too glorious part in the history of iodine, and his action roused the ire of J. L . Gay Lussac. Humphry Davy was in Paris a t the very time of the excitement consequent on these reports of the properties of B. Courtois' new element; R. D avy received a complimentary specimen from A. M. Ampere on Nov. 23rd, 1813, and on Dec. l l t h , 1813, details of his observations on this substance were also communicated by M. le Chevalier Cuvier to l'lnstitut Imperial de France. H. Davy confirmed the conclusions of J. L. Gay Lussac read a t l'lnstitut seven days previously. H. D avy sent a fuller account to the Royal Xociety,l6 Jan. 20thJ 1814, and another on June 16thJ 1814, and yet a third on April 20th, 1815. ' I n these memoirs, says F. D . Chattaway, " H . Davy did little more than make the discoveries n of B. Courtois and J. L. Gay Lussac known i England." I n 1881, H. Ziiblin tried to decompose iodine, but failed. It must be added that in 1767, in a paper on La soude de uarech, L. C. Cadet 17 spoke of a blue and green substance which is obtained by treating the aq. extract of varec by sulphuric or nitric acid; and he attributed the cause of the coloration it vrne surabondance d'une terre jaune martial. F. Hoefer asks : aurait-il entrevu l'existence de 1'iod.e ? Bromine.-Of the three halogens, chlorine, bromine, and iodine, bromine has the least eventful history. Its elemental nature and its relation to chlorine and iodine were recognized from the very first. While studying the mother-liquid which remains after the crystallization of salt from the water of the salt-marshes of Montpellier, A. J. B alard was attracted by the intense yellow coloration developed when chlorine water is added to the liquid. A. J. Balard digested the yellor liquid with ether ; d ecanted off the supernatant ethereal s o h . ;a nd treated this with potassium hydroxide. The colour was destroyed. The residue resembled potassium chloride ; b ut unlike the chloride, when heated with manganese dioxide and sulphuric acid i t furnished red fumes which condensed to a d ark brown liquid with an unpleasant smell. A. J. Balard submitted a pli cachet4 to the Acad'emie des Sciences in 1824, and published an account of his work in his Mdmoire sur une substance partieditre contenue dam I'eau de la mer,l8 in 1826. He related that he was a t first inclined to regard the substance as a chloride of iodine, but he tried in vain t o establish the presence of iodine. He said : I ts refusal to colour starch blue, and the white precipitate which i t formed with the protonitrate o mercury and with nitrate of lead msured me that no iodine was contained f in it. On the other hand, I could not; detect any indication of decomposition when i t was submitted successively to the action of the voltaic pile and to high temp. Such a resistance to decomposition could not fail to suggest to me the idea that I h ad to deal with a simple body, or with one comporting itself as a simple body, and indeed I was confumed in this view when I regarded the entire treatment to which I h ad subjected the substance. I came, therefore, to the conclusion that I h ad found out a new simple substance closely resembling chlorine and iodine in its chemical aptitudes, and forming absolutely analogous compounds, but showing marked point,sof difference from them both in its physical properties and chemical behaviour, and clearly to be distinguished from them. A t first, A. J. B alard called this substance muride, but afterwards brominefrom the Greek /?pGpos, a stench. A. J. Balard prepared hydrobromic, hypobromous, and bromous acids ; a nd he concluded his memoir by summing up the arguments in favour of the elementary nature of bromine : A substance which in the free state resists as effectively as does bromine all attempts to decompose it, which is expelled by chlorine from all i ts compounds possessed of exactly its original properties, which when allowed to act on compounds containing iodine substitutes itself in every case for the latter element to play a similar part in the new products, find which, in spite of some differeaces o action, is connected with both chlorine and i d n e f THE HALOGENS ,as by the most sustained analogies, seems t o possess t h e s a m e right t o be considered as a sim le body. I t hese results are confirmed by other chemists, bromine mus* as a s imple f bo y rank along with chlorine a n d iodine, a n d i t is manifestly between these elements t h a t it must be placed. B J. R . Joss 1 9 seems t o h ave obtained this element in 1 824 ; h e noticed that a red colour was developed in preparing hydrogen chloride by heating a sample of rock salt with sulphuric acid, but he attributed the coloration to the presence of selenium in the acid employed. J. Volhard also narrates that J. von Liebig had bromine in hand a month before A . J . Balard, but mistook it for iodine chloride : J. v on Liebig related t h a t some years before Balard's discovery he received, from a salt manufactory i n Germany-the Kreuznacher saIt springs-a vessel containing bromine, or a t least a product very rich i n bromine, with a request t o examine it. Believing the liquid t o be iodine chloride, h e did n o t subject t h e specimen t o a very exhaustive study. When he heard of the discovery of B d a r d , Liebig saw his blunder, a n d placed t h e vessel i n a special cabinet for storing mistakes-l'armoire dm fautm. Liebig pointed this out t o his friends t o show how easily one could get vary close t o a discovery of the first rank a n d yet fail to grasp the facts when guided by preconceived ideas. I n 1881, H. Zublin tried to decompose bromine into simpler constituents, but failed. R. Lully, Arbor scientice venernbilis et ccelitu,~,Lugduni Batavorum, 1515. Albertus Magnus, De nlchemia (Theatrum chemicum), Argentomti, 2. 423, 1659. 8 J. R. Glauber, Philosophische Oefen, Amsterdam, 1648 ; H . Bocrhaave, Elementa chemice, Lugduni Batavorum, 1732 ; R. Boyle, The Usefulness o Experimentab Philosophy, Oxford, 1663 ; f S. Hales, Ve.gatnbZe Staticks, London, 1727 ; J. Prieetley, Observntions on Different Kinds of Air, London, 3. 208, 1779 ; J. B. van Helmont, Ortus medicince, Amsterdam, 68, 1648. C. W. Schcele, K6nig. Vetens. AEad. Stockholm, 25. 89, 1774 ; Opuscuh chimica e t physica, f Scheek, London, 52, 1901 ; Alembic Club Leipzig, I . 232, 1788 ; The Chemical Essays o C. Reprinls, 13, 1897. A; L. Lavoisier, Trnitk klkmentnire de chimie, Paris, 1789 : G. d e Marveau, A. L. Lavoisier, C. L. Bartholiet, A . E". d e Fourcroy, Mbthode de nbmemlature ciimipe, Paris, 1787. C. L. Bertholle't, Me'm. Acod., 270, 1785. 7 W. Henry, Phil. Trans., 90. 188, 1800 ; 99. 430, 1809. 8 J. L. Gay Lussac and L. J. T hknard, Mbm. Soc. Arcueil, 2. 295, 1809 ; RecJberchm physicochimiques, Paris, 1811 ; H. D avy, Phil. Trans., 99. 39, 450, 1809 ; Alembic Club Reprints, 13, 1897. 9 H. Davy, Phil. Trans., 100. 231, 1810 ; Alembic C l d Reprints, 9, 1894 ; R . Ziiblin, Liebig's Ann,, 209. 277, 1881. l o J. J. Berzeliue, TIw~nsen's Ann. Phil., 2. 254, 1813 ; J. Martin, Nicholson's Journ., 13. 237, 1806. l1 B. Collrtois (F. Clement and J. B. Dksormes), Ann. Chim. PAys., (I), 88. 394, 1815. l a F. Clement and J. B. DBsormes, Le Moniteur Universel, Dec. 2nd, 1813 ;J. L. G ay Lussac, a?., Dw. 12th, 1813. l a J. L. Gay Lussac., 'Ann. Chim. Phys., (I), 88. 310, 1813 ; J. J. Colin and H. G. d e Claubry, i t., ( I), 90. 87, 1814 ; L. N. Vauquelin, ib., ( I), 90. 239, 1814. J. 1,. G ay Lussac, Ann. Chim. Phys., ( I ) , 91. 5, 1814 ; Ostwald'.? Khssiker, 4, 1890. 15 F. D. C bttrtway, Chem. News, 99. 193, 1909 ; H. Ziiblin, Liebig's Ann., 209. 277, 1881. l6 H. Davy, Phil. Trans., 104. 74, 487, 1814 ; 105. 215, 1815 ; H. Ziiblin, Liebig's Ann., 209. 277,1881. 17 L. C. Cadet, Mbm. Acad., 487, 1767 ; F. Hoefer, Histoire de la chimie, Paris, 2. 399, 1843. 18 A. J. Balard, Ann. Chim. Phys., (2), 32. 337, 1826; F. D. Chattaway, Chem. News, 99. 206, 1909. 1 B J. R. Joss, Journ. prak. Chem., (I), I 129, 1834 ; H . Ziiblin, Liebig's Ann., 209. 277, 1881; . J. von Liebig, ib., 25. 29, 1838 ; J. V'olhard, Justus v o Liebig, Leipzig, 1 192, 1908 ; W. H iittner, on . K ali, 1 . 19% 1917. 1 1 8 n7. 5 7. The Preparation of Chlorine Chlorine is nearly always prepared in the laboratory by the action of a n oxidizing agent-manganese dioxide, lead dioxide, barium dioxide, potassium dichromate, 26 INORGANIC AND THEORETICAL CHEMISTRY potassium permanganate, air, etc.-either directly on hydrochloric acid, or indirectly through the medium of a chloride. Manganese dioxide is the oxidizing agent most commonly used. Much chlorine is prepared for industrial purposes by the electrolysis of soln. of sodium chloride. I. The preparation of chlorine by the action of heat on the chlorides of the h e a metals.--Gold a nd platinum chlorides give off chlorine when heated, but these ~ compounds are far too expensive for the preparation of chlorine, except for very special purposes, such as V. a nd C. Meyer's work 1 o n the vapour density of chlorine, where platinous chloride, PtCl,, was used as the source of chlorine : PtCl2+Pt+Cl2, This salt was selected because it is easily decomposed-about 360"-and is not f deliquescent. I moisture be present, some hydrogen chloride and oxygen will be formed. W. Wahl (1913) used gold chloride. M. Wildermann passed purified chlorine through a tube of hard glass containing reduced copper ; cupric chloride, CuC12, is formed. The chlorine was washed out of the tube by dry air ; t he tube was sealed a t one end ; a nd heated by a combustion furnace. Chlorine gas was evolved : 2CuC1,+2CuCl+Cl,. I n special cases, too, highly purified fused silver chloride can be electrolyzed to furnish chlorine of a high degree of purity.2 II. The preparation of chlorine by the oxidbation of hydrochloric acid.-I . Manganese d ioxide.--C. W. Scheele, the discoverer of chlorine, obtained this gas by warming manganese dioxide with hydrochloric acid: a mixture of sulphuric acid with manganese dioxide and sodium chloride may also be used. I n the latter case, a mixture of one part of pyrolusite with from 1'5 to 2'5 parts of sodium chloride and 2.5 to 3 parts of conc. sulphuric acid dil. with its own volume of water is rnade.3 The equation representing the reaction is : MnO, 2NaC1+ 2H,S04 = M&04 2 H20 N a2S04 +C1,, but some manganese chloride, MnC12, and sodium bisulphate may be simultaneously formed. I n Scheele's process, the mixture may contain one part pyrolusite with four parts of commercial acid, or an excess of coarsely crushed fragments of the pyrolusite may be used ; a nd after the process is over, the excess can be washed and used again. The end products of the reaction are indicated in the equation : Mn02+4HC1 Fra. 4.-The Preparation of Chlorine. =M&l2 +2Hz0 +C1,. The manganese dioxide, or the mixture of manganese dioxide, and salt is placed in a flask A , Fig. 4, fitted with a wash-bottle, C, or other washing and scrubbing train, and the acid poured into the tube funnel, B. T he gas cannot be collected over mercury because that metal is immediately attacked by chlorine forming the chloride. Por lecture experiments, the gas c$n be collected over warm water, or water sat. with salt, or by the upward displacement of the air. The preparation under these conditions should be conductd in a well-ventilated fume chamber since the gas is most objectionable in the atm. The purification of chlorine.-.The gas can be washed with water in order to remove most of the fumes of hydrogen chloride carried over with the chlorine, but to remove the last traces of hydrogen chloride, P. Stolba 4 recommended the introduction of a wash bottle with a soln. of copper s u l ~ h a t eor a tube of solid copper , sulphate or bleaching powder, and then washing the gas with water. According to A. Michaelis, the bleaching powder contaminates the gas with hypochlorous acid. B. Mohr recommended removing the gas by scrubbing it in a tube packed with manganese dioxide, and H. Moissan and A. B. d u Jassonneix kept the tube warm a t about 50'. With the same object, W. H ampe and H. Ditz washed the gas in-8 conc. soh. of potassium permanganate. To avoid contamination with carbon dioxide, H. Ditz recommended washing the manganese dioxide first with nitric or dilute sulphuric acid and then with water to remove carbonates. F. P. TreadwpU + , ' + + THE H ALOGENS 37 and W. A. K . Christie removed chlorine oxides from the gas by passing it through a tube packed with asbestos, heated to redness. The gas ean be dried by conc. sulphuric acid, calcium chloride, or phosphorus pentoxide. J . A. Harker 5 still further purified chlorine by passing the purified gas into cold water so as to form the hydrate, C128H20, which was found to keep very well in darkness below 9". When the hydrate is warmed slightly, it gives off chlorine with less than 0.2 per cent. of impurity. H. Moissan and A. B. d u Jassonneix purified the dried gas by liquefaction, and, after prolonged contact with calcium chJoride, solidifying the liquid so as to enable traces of dissolved gaseous impurities to be pumped off. L . Moser recommended removing air and carbon oxide by liquefying the gas with a freezing mixture of carbon dioxide and ether, and redistilling. The mechanism of the reaction between manganese dioxide and hydrochloric acid.-The reaction between hydrochloric acid and manganese dioxide has given rise to much discussion. When manganese dioxide is treated with cold conc. hydrochloric acid, a dark brown liquid is formed, and chlorine is slowly evolved at ordinary temp., more quickly if the mixture be warmed. The liquid finally becomes colourless and it contains manganous chloride, MnC12. G. Forchhammer, in 1821, showed that if the freshly prepared brown liquid be largely diluted with water, it remains clear for a few seconds, and then becomes turbid owing to the formation of a brown precipitate of hydrated manganese dioxide. Quite a similar precipitate is formed if either the red oxide of manganese, Mn304, or the sesquioxide of manganese, Mn203, be treated in place of the dioxide, Mn02. According to S. U . Pickering,G however, the composition of the precipitate varies with the- conditions of the experiment from about 3Mn02+Mn0 t o 7Mn02+Mn0. Attempts have been made to find what the brown soh. contains. Most are agreed that either manganese trichloride, MnC13, or manganese tetrachloride, MnC14, is first formed as an intermediate product. M. Berthelot 7 considers it improbable that the simple manganese tetrachloride is produced when a soh. of manganous chloride, MnC12, in hydrochloric acid is treated with chlorine ; r ather is the first product an easily decomposed perchlorinated compound, HC13.nMnC13or MnC14.nHC1. The formation and decomposition of such a product would explain the observed phenomena. The isolation of the product of the reaction has proved very difficult because it decomposes so readily. Indirect evidence has been obtained by determining the ratio of the manganese to the available chlorine ; b ut the results are not decisive ; a nd hence some have considered the trichloride is formed ; others, the tetrachloride. C. Naumann isolated the double salts (NH4)2MnC15 and K2MnC15, and hence argued in favour of the trichloride. J. NicklBs a t reated manganese dioxide suspended in ether, C4HI00, with hydrogen chloride, and obtained a green liquid which changed to a deep violet colour on adding more ether. The green oil has a composition corresponding with MnC14.12C4H100.2H20; hence, argued J. Nicklbs, the green oil contains manganese tetrachloride. J. Nicklbs did not succeed in isolating a definite product, and, since he found his analyses variaient singulidrement par leur composition, 8. U. Pickering considers that it is just as likely tLat MMnC12CaH~00.2HC1.4H20 might have been present. W. B. Holmes 9 used carbon tetrachloride in place of ether, and obtained both the tri- and +tetrachlorides of manganese. He therefore inferred that the reaction between hydrochloric acid and manganese dioxide furnishes a soln. containing both chlorides : MnO2+4HCl=MnC&+2H2O and 2MnO2+8HC1=2MnC1,+Cl2+4Hz0 ; b ut in conjunctionwith E. F. Manuel, he gave up the tetrachloride hypothesis, and expressed the belief that the trichloride alone is formed. As an alternative hypothesis, B. F ranke assumed that hydrochloromanganic acid, H2MnC16,is formed as an intermediate product : Mno2+6HCl=H2MnCl6+2H20 ; whichdecomposed : H2MnC16=2HC1+C12-$-MnC12. In the presence of manganous chloride and a n excess of water-still more complex reactions occurred, finally furnishing Mn02.H20. Apart from electrolytic chlorine, by far the largest proportion of chlorine wed in the industries is made by the oxidation of hydrocbJoric; acid, generally INORGANIC AND THEORETICAL CHEMISTRY by manganese dioxide either as native manganese ore, or as " recovered manganese "-the so-called Weldon mud. T he precess w ith hydrochloric acid is conducted as an annexe to the salt-cake works wbere the a,cid is a by-product of the process-otherwise the cost of transport of hydrochloric acid would not enable the process to live against competitive prices. The stills for generating the chlorine are not made of lead because that metal is attacked too readily by the mid ; vessels of stoneware were formerly used ; t o-day the stills are built witb flags of siliceous sandstone which are sometimes first boiled in tar. The volvic lava from Puy-de-DBme (France) is preferred even t o the best sandstone. Acid-resisting bricks are also used. The flags are clamped together by iron tie-rods, and tbe joints either made tight with indiarubber cord, or tongueand-groove joints are employed with a cement of fireclay and tar. The mixture is heated by blowing in steam. A section of one type of cblorine still is illustrated i n F ig. 5. T he manganese ore rests on the perforated plate A ; B is a stoneware steam pipe ; t he steam enters the still v id B below the perforated false bottom ; C i s the exit pipe for chlorine ; D i s a man-hole ; E a n acid safety funnel which has various designs ; a nd P i s a n opening for running off the spent acid, i t is closed by a wooden plug. The lid is of lead. The ex'FIG-5.-Chlorine Still. bausted still liquor contains manganous chloride, and some ferric and other metal chlorides derived from impurities in the manganese ore. There is also some free chlorine, and free hydrochloric acid-e.g. the composition of a still liquor approximated : R Cl Cia (free) AIC1, FPC& MnCl, Water 6 .7 0 .1 2.8 1 6.5 73.3 per cent. 0 .6 It will be observed that even under the ideally perfect conditions represented by only half the chlorine of the the equation, Mn0,+4HC1=MnCl2+2H20+C1,, acid can be obtained in the free state. The hydrochloric acid consumed by the impurities in the manganese ore ; b y forms of manganese oxides with a lower power of oxidation than the dioxide ; a nd the excess hydrochloric acid which escapes oxidation all tend to reduce the efficiency of the process. Various patents have been granted for utilizing the still 1iquor-e.g. it has served for the manufacture of manganese carbonate for purifying coal gas (R. Laming) ; l o for deodorizing facal matters (J.Dales) ; for converting barium sulphate to the chloride (I?. Kuhlmann) ; for the manufacture of brown pigments of various kinds by the precipitation of the manganese oxide with lime (C. Chockford) ; a nd for the manufacture of pure mazlganous chloride (K. Muspratt and B. W. Gerland). Various other proposals for utilizing the free acid of the still liquor have been made, and patents have been taken for recovery of manganese dioxide by precipitating manganese oxide with lime and subsequently oxidizing the precipitate with air. None of tb.ese processes can be regarded as successful ; 11 t he recovered manganese in some cases cost more than the original ore, or else it did not work satisfactorily. W. Weldon's improved process is founded on the fact that the freshly precipitated manganese hydroxide suspended in a s o h . of calcium chloride, is easily converted into the dioxide when an excess of lime is present. Many of the older processes recovered manganese dioxide by the action of a current of air upon the manganese hydroxide precipitated by lime, but the oxidation is so slow as to be useless in practice, and even then only about half is converted into the peroxide. for the oxidation seems to stop when the manganese is oxidized t o the sesquioxide, Mn20s. In Weldon's recovery process, i t is t he excess of lime which led t o commercial success, for there is a complete conversion to the dioxide in less than one-tenth the time required for maximum conversion when there is no excess of lime. W. 'Weldon showed that the complete oxidation of the precipitated manganese hydroxide can take place only in the presence of a strong base. Manganese dioxide has weakly acidic properties, and, in the absence of strong bases, it c~mbines with unchanged manganese sxide, MnO, T HE HALOGENS 29 to form manganese manganite, Mn0.Mn02, i .e. Mn,03. I lime is present, the f formation of calcium manganite, Ca0.Mn02 and Ca0.2Mn02, is much faster than the formation of manganese manganite, and all the manganese is then converted t o the higher oxide. The application of Weldon's recovery process involves treating the acid still liquor with ground chalk or limestone to neutralize the free acid and precipitate the oxides of iron. The clarified liquid is run into a tall cylindrical vessel, where i t i s mixed with milk of lime in sufficient qua.ntity to precipitate all the manganese as manganese hydroxide. Mn(OH),. The reaction is symbolized : MnCl,+Ca(OH), =Mn(OH), fCaC1,. An additional amount of l i m e f r o m one-fifth to one-third the quantity previously employedis introduced ; t he liquid is warmed to about 55' or 60" by blowing in steam ; a nd air from a compressor is driven through the liquid for about 29 hours. The manganese is converted into the dioxide, and the contents of the vessel run into a vat, where the m a anese dioxide is precipitated as calcium manganite, CaO,MnO,. The reaction is symbo ized : 2Mn(OH),+2Ca(OH), + 02 2(MnO,.CaO) +4H20. More still liquor is added and the = blowing continued whereby CaO.MnO, is converted into Ca0.2Mn02. The reaction is represented : 2MnO,.CaO +2CaO +2MnCI, +O, =2(Ca0,2Mn02)+CaCI,. The slurry is run into settlers, where the manganite deposits as a thin mud-Weldon mud. The clear liquid-largely calcium chloride-is run to waste ; n o use has been found for but a very limited quantity of this by-product. The mud is pumped iuto special chlorino stills where it ismixed with hydrochloric acid. The mixture is warmed by blowing i n steam. After the chlorine has ceased t o be evolved, the residual liquid is treated as before. The circulation or transport of the iiquids and slimes is effected wholly by pumping machinery ; t he time required for completing one cycle is comparatively short ; a nd the plant is simple and inexpensive. These advantages secured a n unprecedented success for this process of manufacturing chlorine. l4 "81 The theory of the process is incomplete, but that outlined above, based on the acidic qualities of manganese dioxide, is due t o W. Weldon.13 It will be observed that Weldon's mud has a composition approximating to C a0.2Mn02,and, in the chlorine still, this reacts : CaO.2MnO~+1OHCl=CaC1,+2MnCl, +2C12+5H20 ; if manganese sesquioxide were T he used: Mn,03+6HC1=2MlzC1,+3H,0+Cl,. consumption of acid per unit of chlorine, as well as the reduction in the time required for the oxida; tion is therefore in favour of Weldon's process. It is, however, a blemish that it produces less chlorine per unit of acid than native manganese ore of good quality ; and, as W. Weldon himself has said,,it is a barbarous process that yields only one-third of the total chlorine of the acid treated, and thatthe other two-thirds are wasted ascalcium chloride. 2 . The oxidation o hydroclzloric acid b y potasf sium permanganate.--H. B. Condy 1 4 obtained a Dkc. 6.-E. Wedekind and 8. J. provisional patent for the preparation of chlorine ~ ~Chlorine Generator. ~ i ~ l by heating a mixture of sodium chloride. sodium ss pkmangaiate, and sulphuric acid. ~ h e ' ~ r o c e was to be used when a fairly puke product was required for manufacturing purposes. C. M. T. du Motay developed aproeess in which hydrochloric acid is passed over a heated mixture of lime and potassium permanganate. This salt is a very convenient oxidizing agent for preparing chlorine on a small scale, a t ordinary temp. A flask containing some c q d d s of 10 grms.-is fitted as indicated in Pig. 4 a nd potassium permanganate-say connected with a wash-bottle containing conc. sulphuric acid. Dilute hydrochloricacid-60 to 65 c.c. of acid of sp. gr. 1.17-is run, drop by drop, fram a t ap funnel, when chlorine is evolved by the reaction : 2KMn04+16HC1+8H20+2KCI 4-2MnCI2+5C1,. S. J. Lewis and E. Wedekind7s15 a pparatus is illustrated in Pig. 6. The apparatus, Fig. 8, i s intended to furnish a r e e a r supply of gas under control. The permmganate is placed in ,the f l a k A furnished w i t h a stoppered funnel from which 3 0 INORGANIC AND THEORETICAL CHEMISTRY the supply of acid is regulated. The gas passes through the exit tube B , where it is to be used, and any excess escapes wid C, D to the outside air. The bottle E contains water or suiphuric acid, and it acts as a safety tube. If the pipette P be lifted above the surface of the liquid in E , the gas will escape through P and D into the outside air. When the supply of acid is cut off, the evolution of gtw ceases when all the acid is used up. According to A. Scott,l6 potassium permanganate is veryliable to be contaminated with chlorates which introduce chlorine oxides into the gas. The only safe test for these impurities is to absorb the gas in a neutral soln. of potassium iodide, and just decolorize the soh. with sodium thiosulphate. If t he addition of pure dilute hydrochloric acid does not restore the blue colour, the absence of chlorine oxides may be inferred: 3C120+7KI=312+KI03+6KC1, a nd 6C102+10KI=3~+4KIOs +6KC1. The hydrochloric acid destroys the iodate : KI03+6HC1+5KI=6KC1 +3H20 +312. 3. The oxidation o hydrochloric acid by chloratix-L. von Pebal (1875) 17 a nd f G . Schacherl (1876) have investigated the preparation of chlorine by the action of potassium or sodium chlorate on hot conc. hydrochloric acid. I t he temp. is low, f the gas will be contaminated with chlorine oxides and oxygen. The reaction is somewhat complex.ls The two main reactions are : HC103+5HC1+3C12+3H20 ; a nd 2HC103+2HC1+2C102+C12+2HZ0. F. A.Gooch andD. A . Kreider recommend washing the gas in a warm soln. of manganous chloride in conc. hydrochloric acid, or better, by passing the gas through a heated tube packed with asbestos. This process offers no advantages over other processes ; i t is far less convenient.; and there is a risk of explosions. 4. The oxidaidation o hydrochloric acid by nitric acid.-4. Watt and T. R. Tebutt lo f proposed t o make chlorine by heating lead chloride with nitric acid, but the patent was of no technical importance. C. Dunlop heated a mixture of sodium nitrate and chloride with sulphuric acid, and passed the mixture of chlorine and nitrous gases through conc. sulphuric acid-the chlorine passes on, the nitrous gases are :etained by the sulphuric acid. This process was in use a t St. Rollox for some years where the chlorine was used for making bl9aching powder, the by-product of nitrous vitriol was utilized in the sulphuric acid process. H. Goldschmidt 2 0 supposed that the reactions furnished nitric and.hydrochloric acids-aqua regia, in fact--which . decomposed into nitrosyl chloride and chlorine : 3HC1+HN03=2H20 +NOCI+C12 ; a nd in contact with water or sulphuric acid, NOCI+H20=HN02+HCI. These reactions are but approximations to the far more complex changes which actually occur. G. Lunge represents the reaction : 3NaCI+NaN03+4H&304=NOCl+C12 + 2H20 +4NaHS04, which corresponds with those of H. Goldschmidt. Many other modifications have been proposed.21 I n T. Schlosing's process, a mixture of nitric and hydrochloric acid reacts with manganese dioxide, chlorine is evolved and manganese nitrate is formed : Mn02+2HN03+2HC1+2H20+C12+Mn(N03)2. T he latter when dried and calcined,furnishes manganese dioxide, Mn02, and nitrous fumes, Mn(N03)2+Mn02+N204, which can be converted into nitric acid; 2N2O4 +02+2H20=4HN03. There is a loss of about 10 per cent. of the nitric acid in the working of the process. This process is a variant of a patent by F. A. Gattyin 1857. 5. The oxidation o hydrochloric acid by chromates or dichromaies.-Ia 1848 f A. MacDougal and H. Rawson 2 2 p atented the manufacture of chlorine by heating chromtes or dichromates-preferably those of calcium-with hydrochIoric acid, directly or indirectly. The process with potassium dichromate was recommended by E.M. PBligot, J. G. Gentele, and H. E. Roscoe for preparing a fairly pure gas : K2Cr,07+14HC1=2CrC1,+2KC1+7H20 +SC12. 6. The oxidation o hydroclzloric acid by bleaching powder.--Chlorine can be made f by the action of an excess of hydrochloric acid on an alkaline hypochIorite or bleaching powder. The process was suggested by M. Boissenot 23 i n 1849, and later recommended by A. Mermet and H. Kiimmerer. ~ c c b r d i n ~ C. Winkler, the to bleaching powder may be compressed into cubes with suitable binding agent---say plaster of Paris. The gas comes off a t ordinary temp,, and the cubes used i n Kipp's T HE HALOGENS 31 apparatus with hydrochloric acid of sp. gr. 1'124 dil. with its own volume of water -but no sulphuric acid. According to J. Thiele, no binding agent is necessav. 7 . T h e catalytic oxidation of hydrochloric acid by atmospheric air.-R. Oxland,24 in 1845, patested a process for making chlorine by passing a mixture of hydrogen chloride and air through red-hot pumice and washing out the undecomposed gas by water. Ten years later, H. Vogel proposed to prepare chlorine by heating cupric chloride to dull redness, about 500") whereby cuprous chloride is formed : 2CuC12 =2CuC1+Cl2. The cuprous chloride was then mixed with hydrochloric acid and oxidized by air when he supposed cupric oxychloride, CuCl2.2Cu0.3Hz0, was first formed, and then cupric chloride itself, so that the end-product of the reaction is regenerated cupric chloride : 4CuCI+4HC1+O2-4CuC?l2+2H2O. $tripped from the accessory reaction, it will be observed that fundamentally the reaction may be symbolized : 4HC1+02=2C12+2H20. I n practice on a large scale only one-third of the chlorine of cuprio chloride was obtained ; t he copper *chlorides quickly corrode stoneware, firebricks, etc. ; t he manipulation is dangerous to health ; a nd the cost is high owing to the loss of copper. Various modifications were proposed by C. P. P. L aurent, F. M. A . de Tregornain, and J. T. A. Mallet, but none were successful until 1868, when H. W. Deacon 25 arranged the reactions so that the process is continuous. I n the early process the yield was rather small, but R.Ha~enclever2~ obtained better results, introducing the hy+ochloric acid in a continuous stream into hot sulphuric acid contained in a series of stoneware vessels, and driving out the hydrochloric acid by a stream of air. Although this added to the cost of the operation, since the sulphuric acid had t o be conc. again, the process worked more regularly, and the purification of the hydmhloric acid effected by this treatment has added much to the successful working of Deacon's process. It was also found that with impure gases containing sulphur oxides, arsenic oxide, carbon dioxide, etc., the activity of the catalytic copper is rapidly destroyed.27 H. W. Deacon showed that the oxidation of hydrogen in hydrogen chloride can be effected by atm. oxygen, by passing the mixed gases through a tube a t a high temp. The action takes place below 400" i n the presence of pumice-stone sat. with cuprous chloride-CuC1. The result of th6reaction is represented bv the equation : 4HCl+O2+CuC1=2H2O+2Cl2+CuCl. T hecuprous chloride remaking a t the end of thereaction has the same composition as a t the beginning. It is supposed that the h t action results in the formation of a copper oxychloride : 4CuC1+02=2Cu20C12 ; followed by : C U ~ O C ~ ~ + ~ H C ~ = ~ C U C J .a~ + H ~ O by : 2 C~C1~=2CuCl+C1~. ; nd finally Several other guesses z8 have been made on the nature of the cyclic reactions between the catalytic agent and the reacting gases. Iron, nickel, cerium, and other chlorides can be used in place of copper chloride. H . Ditz and B. M. Margosches2D have patented the use of the chlorides of the rare earths which occur as a by-product in the manufacture of thoria for gas mantles. The chlorine is necessarily contaminated with undecomposed hydrogen chloride, atmospheric nitrogen, atmospheric oxygen, and steam. The steam and hydrogen chloride can be removed by washing, etc. The chlorine so prepared is used in.the manufacture of bleaching powder, where the presence-of the impurities does no particular harm. The reaction can be illustrated by the zpparatus 'shown in Fig. 7 . Air is forced from a gas holder through a wash-bottle containing hydrpchloric acid, and then through a hot porcelain tube containing pumice-stone impregnated with a soh. of cupric chloride and dried. The chlorine gas obtained a t the exit can be collected in the usual manner. It is, of course, mixed with the excess o air, nitrogen, etc. f I n the reaction: 4HC1+Oe+2C1?+2H20, for equilibrium, [H,0]2.[C12]2 =K'[HCl]?[02], where K' is the equilibrium constant, and the symbols in brackets represent the concentrations of the reacting substances; if partial press. be used PI for the hydrogen chloride ; P 2 for the oxygen ; pl for the chlorine ; a nd p2 for steam, the 32 INORGANIC AND THEORETICAL CHEMISTRY equilibrium conditions are PPP22=KVp14p2, where K" is constant, and if K"= K4, Plt.P2t=Kppp2*. The latter represents the equilibrium condition corresponding with the decomposition of one mol. of hydrogen chloride. Observations showed that a mixture containing 92.7 mol. of oxygen and 100 mol. of hydrogen chloride a t 352", reacts until 86.95 per cent. of the hydrogen chloride is decomposed when the system is in equilibrium. Consequently, the equilibrium mixture a t 352" contains 100-86-95=13.05 mol. of hydrogen chloride ; 1 of 86.95=43'45 mol. of steam ; a nd 92.7-4 of 86.95 or 71.0 mol. of oxygen. The total press. of the gases was virtually one atm. and therefore the partial press. of the gases must be proportional t o these figures, and their sum must be unity, or the partial press. are : HC1, 0.0763 ; Cl,, 0.2542 ; H 20, 0.2542 ; a nd 02, 0.4152. From the mass law,. therefore a + K = P1 P2 . Or, K = 4 . 5 2x 40.254'2 04 2 PIP? ' 0.0763 x 2/0;4fEf or, 1<=4.15 at 352". The heat evolved during the reaction HC1+i02=H,0 -+CI+6'9 Cals. G N. Lewis,30 assuming that the sp. ht. of the four gases do n ot FIG. 7.-Illustration of Deacon's Process for Chlorine. v ary appreciably for temp, ranging between 300"-4W0, regarded the heat of the reaction to be independent of the temp. and van't Hoff's equation showing the influence of temperature on the reaction, is then represented by log K=Q/RT+Ct, where C' is a constant, R=lm985,these and Q=+6.9. Substituting these values of Q and R, and passing from natural to common logarithms by diiriding by 2.3, there follows : loglo K=1509/T+C, where C is a constant. Substituting the observed values of K and T, and logl, 4~15=1509/629+C, or C=ln78. Hence, as a first approximation, the expression loglo K=lGOgT -1-1 -78 enables corresponding values of K a nd T t o be computed a t 386", K=2'94 (observed) and K=3.02 (cnlculated) ; a t 419", K=2'40 (obs.),2'35 (calc.). At 25" the computed value of K is 1800. I allowance be made for the variation of the heat of the reaction Q with temp., f K. V. vonPa.lckenstein obtains, as a second approximation, log K = 14Nn5T-1-0.534 Iog T-O*00021425T+1*7075 x 10-8T2+0'074. W . D. Treadwell used the expression log Kp=6034T-1-6.972 over the range 300" to 1800°, and t h formula is approxi~ mately correct down to room temp. The value of Kp at 352" is 2.68 ; a t 6W0, -0-06 ; and a t 1984", -4'30. The reaction never runs completely to an end, but rather approaches a state of equilibrium : 4HCI+O2+2Cl2+2H20, which fixes a definite limit to the yield of chlorine which can be obtained a t any particu1,ar temperature and concentration of T HE HALOG-ENS 33 the reacting gases. The most favourable practical conditions were worked out by G. Lunge a nd E . iMarmier.31 I n the reaction : 4HC1+O2-+2C12+2Hz0, 27'6 Cals., both chlorine and oxygen are competing for the hydrogen; a t 577" both appear equally strong, for the hydrogen is distributed equally between the chlorine and oxygen. At higher temp. the chlorine is stronger than oxygen, because less free chlorine is obtained than a t lower temp., when the affinity of oxygen for the hydrogen is the stronger. I n consequence, a greater yield of free chlorine is obtained a t temp. lower than 577". This agrees with the effect of temp. on chemical reactions deduced thermodynamicalIy. Since the reaction is exothermal, the lowering of the temp. favours the formation of chlorine. The temp., however, cannot be reduced indefinitely because the reaction would then become inconveniently slow, even in the presence .of the catalytic agent-cuprous chloride. The catalytic agent begins t o volatilize a t temp. even below 430". According to K. V. v on Palckenstein, the best yield is obtained with a mixture of 40 per cent. NCI and 60 per cent. air, when about 70 per cent. of the hydrogen chloride can be oxidized to chlorine. I n accord with the rule that an increase of pressure favours the system with the smaller volume, and remembering that over 100°, the system 4HCl+02 occupies five volumes when the system 2H20+Clz occupies three volumes, it follows that an increase of press. should favour the oxidation of hydrogen chloride and augment the yield of chlorine. 3. Quincke 32 recommended using oxygen in place of air. This raises the partial press. of the oxygen, and induces a more complete oxidation of the hydrogen chloride to chlorine. In practice, the mixture of air and hydrogen chloride from the salt-cake gases is driven through cooling pipes and scrubbers t o remove moisture, and dried in a sulphuric acid tower. There are two sets of cylinders heated t o about 450"by waste heat. The cylinders contain broken bricks dipped in a s o h . of cupric chloride. The cylinders are recharged about once a fortnight. The exit gases containing 5 to 10 per cent. of chlorine are dried in a sulphuric acid tower and used for making bleaching powder. A bout two-thirds of the hydrogen chloride is converted into chlorine. III. The preparation ot chlorine by the oxidation of the metal chlorides.Chlorine can be obtained by the action of oxygen or sulphur upon certain chlorides. The electrical energy required for the electrolysis of the fused chlorides is nearly proportional t o their heats of formationHeats of formation . . zSaC1 195.4 CeCI, 169'8 MgCI, 151- 0 Cals. In the idealized reactions 2MCl +O = M20 +C12, where M represents a unlvalent or an eq. bivalent element, energy respectively eq. t o 151.0, 39.0, and 7.0 Cab. per mol. of chlorine is needed. Hence, the manufacture of chlorine by the calcination of magnesium chloride in a current of air appears far more feasible than the treatment of sodium chloride, because the latter requires the expenditure of over twenty times more thermal energy per mol.\ of-chlorine. None the less, patents for the treatment of all these chlorides, and others, have been taken. Thus, W. Longmaid 33 proposes t o obtain chlorine by calcining the chlorides of manganese, copper, iron, zinc, or lead with an excess of air. J . Nargreaves and T. Robinson used a mixture of ferric chloride or chlorine oxide and salt ; W. Weldon, a mixture of ferrous sulphate and salt; and A. R . A rrott, a mixture of ferrous phosphate and salt. I n general, the expulsion of the chlorine may also be facilitated by mixing the sodium chloride with a sulphide, sulphate, silica, bonc oxide, stannic oxide, phosphoric oxide, alumina, clay, etc., or a mixture of air with sulphur oxide can be used.34 Similar remarks apply t o calciuni chloride. The enormous quantities of calciuni chloride produced in the ammonia-soda process has attracted inventors who have made verypersistent eff orts t o separate the chlorine by a cheap process. Thus, E. Solvay 35 had a series of about twenty patents between 1877 and 1888 directed to the decomposition of the chloride by heating a mixture with sand or clay in a stream of VOL. 11. D 34 INORGANIC AND T HEORETICAL CHEMISTRY air. F. H urtur has shown that the great amount of thermal energy required for the decomposition of sodium and calcium chlorides by chemical process makes i t . probable that it would be really cheaper to decompose these chlorides by sulphuric acid than by Solvay's process. I n an attempt to recover the chlorine from the stillliquor in t he manufacture of chlorine from manganese dioxide and hydrochloric acid, W . Weldon 36 mixed the acid liquid with magnesite and heated the dried residue ; although the process was not successful industrially, he got the idea 37 t hat chlorine could be obtained from hydrochloric acid by converting the latter into magnesium chloride : MgO+2HCl=MgC12+H20 ; a nd then heating the magnesium chloride in a stream of air : 2MgCl2+O2=2MgO+2Cl2. The reaction between oxygen and magnesium chloride is reversible,38 and measurements of the equilibrium conditions when the different reacting members are heated in a closed tube shows that theequilibrium constant K agrees best with the assumption that the reaction proceeds : 2MgC12+02=2Mg0+2C12, and accordingly K=C12/C2, when C1 denotes the concentration (partial press.) of the chlorine, and C2 that of the oxygen. The observed values of K were 0.0324 a t 586" ; a nd 0,0625 a t 675". If water vapour be present, equilibrium is established between 350" and 505" through the relation : MgC1,+H20 =MgCl.OH+HCl ; a nd above 510°, equilibrium is established through the relation : H 20+MgC12=Mg0 +2HC1. Between 505" and 510°, the oxychloride MgC1.OH is decomposed : MgCI.OH=HCl+MgO. Technical details of W. Weldon's process were developed in conjunction with A. R . PBchiney, and the process-called the Weldon-Pechiney process-was worked in a continuous cycle of operations: (i) dissolving the magnesia in hydrochloric acid ; (ii) mixing the magnesium chloride with a fresh supply of magnesia so as t o form magnesium oxychloride, and evaporating to dryness ; (iii) breaking, crushing, and sifting the magnesium oxychloride ; ( iv) heating the magnesium oxychloride to a high temp. when any water present is converted into hydrochloric acid, and the remaining cblorine is given off in a free state ; ( v) converting the resulting magnesia back to the oxychloride and so o n in a continuous cycle of operations. Probably the magnesia acta a s a catalytic agent-like copper oxide in Deacon's process-and, in the furnace, converts parts of the hydrogen chloride into chlorine and water. The process was used for a time a t Salindres (Prance), where the mother liquors from the evaporation of sea-water for salt were treated. Many modifications have been patented. 5 . Mond 39 t ried to recover the chlorine from the waste liquors of the.amrnonia-soda process. I n Mond9s chlorine process the ammonium chloride vapour is led over nickel or other metal oxide a t about 4 0O0, t he chlorine is retained : NiO+2N&+2HC1=NiCl2+H20+2NH3, t he ammonia passes on. The ammonia gas is then washed out of the apparatus by aspirating an inert gas-producer or flue gas-through the system. If d ry air be then led over the nickel chloride a t 500", chlorine is given o f f : NiC12+02=Ni0+C12 ; a nd if steam be used, hydrogen chloride is formed : NiCl2+HzO =NiO +2HCI. Ij. Mond, however, returned to the use of magnesia in which magnesium oxychloride was produced in place of nickel chloride; &nd the oxychloride was decomposed by heating in dry air a t 800" : 2Mg20C12+02=4Mg0+2C12. The Solvay Process Co. heated mixture of alkali chloride and ferric sulphate in the presence of oxygen. N. Electrolytic processes for c h l o ~ m d alkaline hydroxides.--In the elece trolysis of conc. hydrochloric acid, with carbon or platinum electrodes, chlorine is evolved a t the anode, hydrogen a t the cathode. When the conc. acid, dil. with eight volumes of water, is electrolyzed, some oxygen is evolved along with tho' chlorine ; with nine volumes of water, still more oxygen is evolved. The more dil. the acid, the greater the amount of oxygen, until, with water acidified with a few drops of acid, no chlorine, but oxygen alone is obtained a t the anode.40 The relation between the yields of chlorine and of oxygen with acids of different concentrations is shown in Pig. 8 ; with 3N-HC1, the amount of oxygen obtained was scarcely appreciable. The less the current density the greater the yield of oxygen with soh. more dil. than &N-HCl ; a nd with more conc. acids, the converse is true. . , T HE HALOGENS 35 Other products besides chlorine, hydrogen, and oxygen are obtained. Por.example, some hypochlorous acid could always be detected in the anode gas obtained in the electrolysis of the more dil. acid ; chloric acid accompanied by hydrogen peroxide in small quantities' was formed with acids between normal and &N ; a nd perchloric acid is formed when the more dil. acid is electrolyzed. I an aq. soln. of potassium or sodium chloride be electrolyeed, chlorine (anion) f appears a t the anode, and the metal (cation) a t the cathode. I n the case of sodium chloride, the primary separation is symbolized : NaCI=Na+Cl-96.4 Cak. The metal then reacts with the water, liberating hydrogen and forming sodium hydroxide : 2Na 60 + 2H20=2NaOH+H2+86% Cals. The ther- 3 ma1 energy required to decompose one mol. of sodium chloride and produce a mol. of sodium hydroxide, and a gram-atom each of hydrogen F - , ~ and chlorine is -96'4+$ of 8 6 . 8 -53-0 Cals. g ~ The net result of the electrolysis is accord0 ingly symbolized : NaCl+H20=$C12+$Hz Concent&jon af dcid +NaOH--53.0. Cals. Here, 53,000 cals. of F lu. 8.-Relation between the Yields energy a re eq' to 53,000t0'24 of Chlorine Oxygen with Hydro=220,800 joules of eIectrica1 energy per gram ,hlOric Acid of Different Concentra eq. ; and since 96,540 coulombs will liberate a tiofis. gram-eq. of sodium and of chlorine ; a nd electrical energy in joules is equal to the ~ r o d u cof the capacity factor in coulombs and t density factor in volts, 220,800+96;540=2.3 volts are needed to liberate a gram-eq. of sodium and chlorine, on the assumption that the electrical energy is eq. to the thermal value of the reaction. I n practice a greater voltage is needed; this is in part due to the expenditure of energy in overcoming the resistance of the electrolyte. The probability of manufacturing products by the electrolysis of fused salts was foreseen by M. P araday 4 1 i n 1834 before the development of the dynamoJ. which replaced the voltaic battery as a source of electricity on a manufacturing scale. 3 The capability of decomposing fused chlorides, iodides and other compounds and the opportunity o f collecting certain of the products without any loss by the use of simple sppsratw, render i t probable that the voltaic battery may become a useful and even manufacturing instrument. The hydrogen gas produced during the electrolysis appears a t the cathode and chlorine at the anode. I n practice it is found that in the electrolysis of the alkali chlorides, the two electrodes must be separated t o prevent the sodium hydroxide formed a t the cathode mixing with the chlorine discharged at the anode, There are three main types with many differences in detail.42 ( I) Solid diaphragm process.-A porous diaphragm-Podand cement, earthenware, asbestos, limestone, etc. This permits electrolytic conduction, and greatly retards the mnixini of t he soln. The GriesAeim cell 4 3 is a rectangular steam-jacketed iron box which contains about six cement cases, the walls of which act as diaphragms. Each cement box is provided with an anode made of magnetite, add a lid and exit pipe for the chlorine. The iron box forms the cathode. WIG. 9.-E. A . le Sueur's Bell Cell (2) Bell process.--The anode is enclo~ed an in (Diagrammatic). inverted non-conducting bell with a cathode outside.'". A . le Sueur's cell ja illustrated diagrammatically in Fig. 9. T he electric current flows from anode t o cathode. Chlorine gas passes out of the bell, and the 36 INORGANIC AND THEORETICAL CHEMISTRY alkali hydroxide forms about the cathode in the compartment outside the bell. The diaphragm is made of asbestos covered with iron gauze which aho forms the cathode. Fresh brine flows into the anode compartment through a hole in the lid, and the alkali hydroxide runs from the exit pipe of the outer cell. The electrolyt,e is always moving from within outwards. ( 3) Mercury cathode process.-The sodium is dissolved by the mercury to form an amalgam. The amalgam is removed from the cell and treated with water, when sodium hydroxide and mercury are obtained. The mercury is returned to the cell to be used over, and over again. I n E. S olray's ccll(1898),45 the mercury is circulated through the cell by gravity whence it flows into a se arated trough, where i t is decomposed by a counter stream of water. The alka ine liquors are drawn off, and the regenerated mercury returns to the cell. The mercury is circulated by an Archimedean screw. I n H. Y. CastnerTsprocess (1893),46illustrated in Fig. 10, the reaction between the water and the mercury amalgam takes place in the cell itself. The cell is nearly obsolete, but it illustrates the principle very well. P E ach cell has three compartments. The two out,er compartments are fitt,ed with graphite anodes f + ) ; a nd the middIe compar-tment is fitted with an i ron grid ( - ) t o serve as cathode. The non-porous partitions do not reach quite t o the boGom ofAthecell but dip into a layer of mercury covering the bottom. A soln. of alkali chloride flows through the two outer cells, and water through the inner compartment. The brine in the outer compartment is decomposed by the electric current into chlorine a t the anode and sodium a t the cathode. The latter dissolves in the mercury, a t the cathode, and the chlorine a t the anodes escapes vid t he exit pipes. The sodium amalgam diffusesinto the inner chamber, and there. comine into contact with the water. is immadiately dvecomposed .i nto sodium hy: FIG.10.-El. Y. Castner's'Rocking Cell. droxide and mercury. The hydrogen escapes through the loosely fitting cover. The sodium hydroxide is run into a special tank as required. A slow rocking motion is imparted t o the cell during the electrolysis, b y a n eccentric wheel, so as to make the mercury flow irom one compartment t o the other along the bottom of the cell. I n C. E . AckerTsprocess (1898), now abandoned,47 sodium chIoride was electrolyzed in a cell in which molten lead was used as cathode, and a carbon rod as anode. During the electrolysis, the molten lead dissolved the sodium forming an alloy ; t he chlorine was drawn off from t he anodes. The alloy of lead and sodium was decomposed by steam t o form hydrogen and sodium hydroxide. 1 V. a nd C. Meyer, Rer., 12. 1426, 1879; W. A. Shcnstone and C. R. Beck, Proc. Chem. Soc., 4. 38, 1893 ; M. Wildermann, Phil. Trans., 199. 337, 1902 ; W. WahI, Proc. Roy. Soc., 88. A, 348, 1913. 2 W. H. Shenstone, Jozwn. Chem. Soc., ? I. 471, 1897 ; J. W. MeIlor and E. J. Russell, i b., 81. 1272, 1902. a P. Klason, Rer,, 23. 330, 1890. 4 F. Stolba, Rer. bdhm. Ces. Wiss., 11. 7, 1873 ; .Dingier's J ourn., 211. 232, 1874 ; A. Michaelis, Lehrbz~chder anorganischen Chewtie, Braunschwsig, 1. 272, 1878 ; P. Mohr, Commentar zur 3weuss$chen Pharrnacopoe, Braunschweig, 98, 1858 ; H. Moissan and A. B. dc Jassonneix, Compt. Rend., 137. 1108, 1903 ; W. Hampe, Chem. Z tq., 14. 1777, 1890 ; H. Ditz, Zed. angeoo. Chem., 14. 6, 1901 ; F. P. Treadwell and W. A. K. C h~jstie,ib., 18. 1931, 1905; L. Moser, Zeit. awrg. Cfhem., 110. 125, 1920. J. A . Harker, Z i ,phys. Chem., 9. 673, 1892. et. G. Forchhammer, Ann. Phil., (2), I , 50, 1821.; S. U. Pickoring, Journ.Chem.Soc., 35. (54, 1879 ; Phil. Mag., (5), 33. 284, 1808 ; C. Naumann, Monatsh., 15. 450, 1894. 7 M. Bcrthelot, Compt. Rend., 91. 251, 1880. 8 W. W. Fischer, Journ. Chem. Soc., 33. 409, 1878 ; H. M . Vernon, Proc. Chenz. SOC., 58, 6. 1890 ; Phil. Mag., ( 5), 31. 460,1891 ; R . J. Mcyer and H. Best, Zeit. anorg. Chem., 22. 169,1899 ; THE HALOGENS 37 J . NickEs, C m p t . R d . , 80. 479, 1865 ; A nn. Chim. Phys., f4\, 5. 161, 1865 ; B. Franke, J wra prakt. Chem., (2), 36. 31, 451, 1887 ; 1.W acker, Chem. Ztg., 24. 285, 1900. ; C. N aumann, Monatsh., 15. 483, 1894. l0 R . Laming, Brit. Pat. No. 11944, 1847 ; J . Dales, ib., 2157, 1859 ; C. Crockford, ib., 1860, 18G3 ; K . Muepratt and B. W . Gerland, z k , 2922, 1856 ; 1589, 1857 ; I. K uhlmann, ib., 2650, ? 1856 ; C m p t . R e d . , 47. 164, 1858 ; T . L eykauf. Dingler's Journ., 190. 70, 1868. l 1 W . Gossaga, Brit. Pat. No. 7416,1837; C. Binks, ib., 7 963,1839 ;C. Rinks and J . Macqueen, z . 1240, 1860 ; W . W eldon, ib., 1948, 1866 ; C. Dunlop, ib., 1243, 2637, 1855 ; W . Gosea@, i 3, b., 1963, 1855 ; C. Jezler, Di.ngZer's Joxrn., 215. 446, 1875. l a L Mond, R. A. Rep., 734, 1890 ; Journ.. Soc. Chem. i d.,15. 713, 1896 ; G . Lunge, T h Manufucture of Acid and AZkaZi, London, 3. 325, 1911. l8 . Weldon, Chem. News, 22. 145, 1.870; 41. 129, 179, 181, 1880; 42. 1 0, 19, 1880; Rer., W 16. 398, 1882 ; Journ. Soc. Chem. I d . , 10. 387, 1884 ; G . Lunge, Chem. News, 41. 129, 1880 ; D iykr's Joumz., 235.300,1880 ; 2 36.231,236,1880 ; 242.371,1581 ; C T. L unge and B. Zahorsky, Zeit. angew. Chem., 6. 631, 1892 ; J . Volhard, Liebig's Ann.., 198. 318, 1870 ; J . P ost, Rer., 12. 1454, 1539, 1879 ; Dingler's Journ., 236. 225, 1880 ; 239. 74, 1881. l4 H. B. Condy, Brit. Pat. No. 3100,1866 ;W . H. Balmain, ib., 1059,1869 ; C. M . T . d u Motay, Ih'ykr's Journ., 205. 356, 1872. l6 C . Graebe, Rer., 35. 4 3, 1902 ; S . J . L ewis and E . W edekind, Zeit. aryew. Chem., 22. 580, 1909 ; Proc. C'hem. Soc., 25. 59, 1909 ; L . Cohn, Arbeitsmethoden fur organitsch-chemischeLaboratorien, Hamburg, 1901. l 6 A , S cott, Proc. Chem. Soc., 25. 59, 1909. l7 L. v on Pebal, Lzebig's Ann., 177. 1, 1875 ; G . Schacherl, ib., 182. 193, 1876 ;F. A. Gooch and D. A. Freider, Zeit. anorg. Chem., 7 . 17, 1894 ; C. Graebe, Rer., 84. 645, 1901 ; B . Merck, Phurm. Ztg., 48. 894, 1903. l B. S and, Zeit. phys. Chem., 50. 4 56, 1901 ; H . Sirk, Zeif. Ekktrochem., 11. 261, 1905. J l B C . W a t t and T . R. T ebutt, Brit. Pat. No. 7531, 7538, 1838 ; C. R i n h , i b., 7473, 1839 ; 1319, 1853 ; C. Dunlop, ib., 11624, 1847 ; T . R oberts and J. Dale, ib., 2242, 1858. a B H G oldschmidt, Liebig's Ann., 205. 372, 1880 ; G . Lunge, Zeit. angew. Chem., 8. 3, 1895. . 21 C . M. T d u Motay, BUZZ. Soc. Chim., (2),22.48,1874 ; J . T aylor, Chem. I nd., 1 4. 361,1891 ; . Brit. Pal. No. 13025,1884 ; G. and E. D a v i ~ ib., 6416, 6698, 6831,1890 ; F. A. G atty, ib., 2230, , 1857; 0. Krieg, Dingkr's Journ., 151. 48, 1859 ; E . J . Maumen6, Compt. Renal., 33. 301,1851 ; T, Schlosing, ib., 55. 284, 1862. Pa A. MacDoilgal and H. Rawson, Brit. Pat. No. 12333, 1848 ; L. Mond and J . H argreaves, 13., 1312,1870 ; J . S hanks, z%., 2018,1855 ; D . G . Fitzgerald, ib., 5542,1886 ; C. F. Claus, ib., 1054, 1867 ; E. M. Pkligo t , A nn. Chim. Phys., (2), 52. 267, 1833 ; J . G . Gentde, fin&v's Joztm., 125. 492,1861 ; H. E. Roscoe, Liebig's Ann.,95. 357, 1865. 23 I B o i ~ e n o t Journ. Pharm., ( 3 ) , 15. 185, 1849 ; A. M ermet, BUZZ. SOC. d. , Chim., ( 2 ) ,21. 541, 1874 ; H. Kiimmerer, Be9.., 9. 1548, 1876 ; C. Winkler, ib., 20. 184, 1887 ; 22. 1076, 1889 ; P. &son, ib., 23. 330, 1890 ; C. G ~ a e b eib., 34. 645, 1001 ; J. T hiele, Liebig's Ann., 253. 239, , 1889 ; B. Merck, Pharm. Ztg., 48. 894, 1903 ; H. Kreis, Chem. Ztg., 27. 281, 1903 ; C. Eckart, Zeil. aml. Chem., 44. 398, 1905 ; , J . Hargreaves, Brit. Pat. No. 20835, 1904 ; F . W . Bartelt, ib., 3395, 1909. R. Oxland, Brit. Pat. No. 10528, 1845 ; C. P. P. Laurent', z%., 163, 1860 ; J . T . A. Mallet, ih., 33171,1866 ; P. M. A. de Tregomain, z%., 5 , 1864 ; H . V o p l , Diryler's Jou.r?b., 136. 237, 1861 ; A. W . H ofmann, Report by the Juries, London, 35, 1862. 2 5 H. W . Deacon, Brit. Pat. No. 1403, 1868 ; C?iem. News, 22. 157, 1870 ; J ourn. Chem. Soc., 10. 7 25, 1872. R. Hasenclever, Brit. Pat. No. 3393, 1883. e7 K . J urimh, Dingier's Journ., 221. 356, 1876. C. Hensgen, Dingler's Journ., 227. 3Ti9, 1878 ; Ber., 9. 1676, 1876; I. de Lalande and ? M.Prud'homme, RuZl. Soc. Chim., (2),1'7. 290, 1872 ; (2), 20. 74, 1873 ; H. Lamy, ib., ( 2 ) ,20. 2, 1873; F . H urtur, Bingler's Journ., 223. 200, 1877 ; J . Wialicanns, Arch. Sciences Qenkve, (2), 48. 64, 1 873; M. G. Leir and V . Bettoni, CTazz. Chim. Itnl., 35. i, 320, 1905 ; M . G. Levi and V. Voghera, ib., 36. i, 513, 1906. ED H . Ditz and B. M. Margosches, German Pat., D.R.P. 150226, 1902 ; M. G. L evi and V . Voghem, Gazz. Chim. ItaZ., 36. i, 513, 1906. S B 0. . Lewis, Jown. Amer. Chem. Soc., 28.1380,1906 ;F. Haber, Thmrnodynamik technischer N Uameakiimn, Miinchon, 162, 1905 ; L ondon, 181, 333,1908 ; K . V . v o n Falckenetein, Zea. p h y . Chem., 59, 313, 1907 ; 65. 371, 1909 ; Zeit. Ekktrochem., 12. 763, 1906 ; G. BGdlander, ib., 8. 833, 1902 ; W . D. Treadwell, Zeit. Elektrochem., 23. 177, 1919. G. Lunge and E . Mmmier, 2 4 . angew. C k m . , 10. 105, 1897 ;E . Marmier, Ueber die DarsteZlunq vm Chlor nach dem T'erfahren von Deacon und Mond, Zurich, 1897 ; P. Nautefeuille and J . Margottet, Compt. Rend., 109. 641, 1889 ; H . le Chatelier, ib., 109. 664, 1889. ZE F Q uincke, German Pat., D. R. P. 88002, 1902. . 8 8 W . Longmaid, Brit. Pat. Xo. 10797, 1845 ; J . Hargreaves and T . R obinson, z . 668, 1 870 ; 3, 508,1872 ; W . Weldon, ib., 134, 1867 ; A. R. A rrott, ib., 2236, 1873. . 8VP. de Lalande and M . P rud'homme, Bull. Soc. Chim., ( 2 ) )17. 7 , 1872 ; ( 2), 18. 74, 1873. 8 5 E . S olvay, B i t . Pat. No. 77, 171, 1877 ; F. Hurtur, Journ. SOC. Chem. I d , 2. 103, 1883. W . W eldon, Brit. Pat. No. 2389, 1871 ; 317, 2044, 1872. 38 I NORGANIC A ND T HEORETICAL CHEMXSTRY s7 W. Weldon, Brit. Pat. No. 967, 9 68, 1881 ; A. R. Pbchiney and W. Weldon, i 3., 9305, 11035, 1884; 14653, 14654, 1887; F. M. Lyte a nd J. G. T atters, ib., 17217, 1889 ; J. Dewar, Journ. Soc. Chcm. Ind., 6. 775, 1887 ; C. T. Kingzett, ih., 7. 286, 1888 ; B. Kosmann, Die Darstellu~y Chlor und CIdorwasserstoff~~ure Chlormagnesizcm, Berlin, 1891 ; G. Escheliman, won aua Chem. I d., 12. 2, 25, 51,1889 ; F. Fischer, Zeit. angew. Ch.em., 1. 549, 1888 ; R. Nahnsen, ib., 20. 673, 1889 ; H. Kunheim, Ueber Einwirkung decr Waaserdampfees a z ~ChEorrnetaUe bei @her Tern j peratur, Gzttingen, 1861. S s F. Haber and S. Tollocz-ko, Zeit. anorg. Chem., 41. 407, 1904 ; F. Haber and F Fleisclunann, i b., 51. 336, 1906 ; 52. 168, 1907 ; W. Moldenhauer, i h., 51. 369, 1906 ; K. V. von Palcken~tein, Zeit. Elektrochem., 12. 763, 1906. 8 9 L . Mond, Brit. Pat. XO. 65, 66, 1049, 3238, 188G; Solvay Process Co., U .S. Pat. A70. 1236570, 1917. '* F. Haber aud S. Grinberg, Zed. anorg. Chem., 16. 108, 329, 438, 1898 ; 18. 37, 1898. 1 M. F araday, PhtE. T r a n ~ .124. 425, 1834. ' , A. J. Allmand, The Principles o j Applied Elecl.iochemisslry, London, 1012 ; M. d e K a y Thompson, Applied Electrochemisty, New Y ork, 1911 ; F. Foer sber, Elektrackemie wiiseeriger L6aungen, Leipzig, 1905 ;R. Lorenz, Die Elektrolyse geschmolzelze~ Salze, Halle, 1 9 0 5 4 ; R. Lucion, Die elt?lctrolytische Alkdichloridzerlegung mit Jliissigen Metalkathden, Halle, 1906 ; J. Billiter, Die elekbrolytische AEkalichloridzerlegung mit stamen Blektroden, Halle, 1912 ; V. Engelhardt, Hypochlorite und elektroly:isc7~eBleiche, Halle, 1903. 4 3 E. M atthcs and Weber, German Pat., D.R.P. 34885, 1886; J. H argreaves and T. Bird, t x., 18871, 1892 ; 5197, 1893 ; C. Hiiussermann, Dingkr's Journ., 315. 1, 1900 ; B. Lepsius, Chem. Ztg., 33. 299, 1909 ; C. P. Townsend, Electrochem.. I d . , 5. 209, 1907; 7. 313, 1900 ; L. TT. Backeland, ib., 5. 43, 259, 1907 ; R. H. F. and A. H. Finlay, Brit. Pat. No. 1711, 1906 ; 26853, 17492,1907. 4 4 W. Bein, & m a n Pat., U.R.P. 85547, 1893 ; 107917, 142245, 1898 ; E. A. le Sueur, Brit. Pat. No. 15050,1891 ; Zed, Elektrochem., 1. 140, 1894 ; 4.215, 1897 ; 5. 29, 1808 ;J. Billiter, Brit. f i t . No. 11693, 1900. 4T. Solvay, Geman Pat., D. . P. 104900,1898 ; Zeit. Ebctmchem., 7. 241, 1900. R '6 W. H. Walker, Electrochem. Ind., 1, 14, 1902 ; W. C. Carey, Journ. Soc. Chem. Ind., 32. 995, 1913 ; H. Y. Castner, Brat. Pat. No. 16046, 1892 ; C. Rellner, i b., 20060, 1891 : 9346, 1892 ; 20259, 1894; B. E. F. Rhodin, Journ. Soc. Chcm. I d . , 21. 449, 1902 ; J. G. -4. Rhodin, ib., 19. 417, 1900 ; Brit. Pat. No. 21509, 1896. C. E. Acker, Brit. Pat. No. 14269, 1898. $ 8. The Preparation of Bromine Bromine is not usually prepared i n t he laboratory, but if desired it can be liberated from potassium bromide by processes analogous with those employed with sodium chloride, namely, the action of heat on a mixture of manganese dioxide, potassium bromide, and sulphuric acid. Unlike chlorine, the bromine evolved readily condenses to a liquid a t ordinary temp. so that in place of a fla~k, tubulated retort is employed a as illustrated in Pig. 11. I n all technical processes the bromine is liberated from its compounds by the action of chlorine, for if chlorine be passed into a liquid containing u bromide until the yellow or reddish colour no longer increases in intensity, the chlorine will have displaced all the bromine : MgBrz+C1, FIG.11.-The Preparation of Bromine. = l - - r I a n excess of chlorine . -f be used, some bromine chloride will be formed. I n A. J . Balard's historic process, the liquid was shaken up with etheror some other suitable solvent, say, chloroform-in a separating funnel ; under these condi$ions, most of the bromine leaves the a q. liquid and collects in the light,er ethereal layer. The lower aq. layer is run off, and an aq. soln, of potassium hydroxide T HE HALOGENS is added to the reddish-brown ethereal layer, the bromine forms a mixtwe of potassium bromide and bromate. The soln. is e vaporated to dryness and treated with manganese dioxide and sulphuric acid. This process is too costly on a large scale. ~ r o m i n ewas formerly obtained from the mother liquid remaining after the separation of sodium chloride from sea-water, a certain proportion of bromine was also extracted from the lixivium of the ash of seaweed, but the proportion in seaweed is small-about one-tenth that of the iodine. The manufacture of bromine fromthe brine springs of America was commenced a t Freeport in 1846 by D. D. Alter ; and the manufacture from the saline waters about Stassfurt 1 was commenced in Germany in 1865. The discovery that bromine could be profitably extracted from the Stassfurt salts reduced the price of that element from about 38s. Od. t o 1s. ad, per lb. About 1837, F.Mohr separated the bromine in the mother liquids of salt springs by treating them with pyrolusite and sulphuric or hydrochloric acid. He showed that at least 5 per cent. of acid must be present or an appreciable amount of bromine will not be formed. The raw material now employed is carnallite, which contains from 0.25 t o 0 .42 per cent. of bromine. The mother liquid remaining after the separation of the potassium chloride from carnallite contains from about 0.15 to about 0 95 per cent. of bromine in the form of bromocarnallite. I n k Prank's first process 3 t he bromine wars obtained by mixing the mother liquid with suldhuric acid and manganese dioxide, and heating the mixture by blowing in steam. The mixture of steam, bromine, and chlorine was cooled in a spiral tube. The condensate separated into two layers, a heavy layer of bromine contaminated with chlorine, and a lighter aq. soln. of chlo- . rine and bromine. The latter was i added to the next charge in the still, the former was reserved for purification. The fumes from the condenser : were passed through a tower packed FIG. 12.-Recovery of Bromine from Ctlrnallite Mother-liquors with iron turnings which arrested the chlorine and bromine. It was found that a considerable loss of time, and of chlorine and bromine were involved in periodically emptying and charging the still. Continuous processes were therefore devised by R. Wiinsche and R. Sauerbrey,4 and K. KubierschI~~. , The principles of the continuous system are a s follows : T he hot mother-liquor containing the bromine is allowed to percolate steadily down a tower A , F ig. 12, packed with earthenware bdls ; t he descending liquid meets a n ascending stream of chlorine gas. The The excess of chlorine, magnesium bromide is decomposed : MgBr, +C1, =MgC1, +Br,. and the bromine, rising from a n exit a t the top of the tower, descend, and pass t o the condenser where most of the bromine is condensed in a cooled worm-tube, B, a nd collects in the bottle C : t he uncondensed bromine vapours and chlorine pass into a tower D packed with iron turnings kept moist by water. The liquid leaves the base of the tower along the same pipe which brings in the chlorine gas, and runs into a tank beneath the ground. The liquid is forced to flow through this tank t o the exit pipe near the bottom i n a zig-zag direction. The inRow and outflow of the liquid in this tank is so arranged t h a t the level remains the same. Steam is blown into the liquor vici t he pipe E . Most of the chlorine and bromine rise to the space above the level of the liqnid where they are carried along to the tower with the stream of chlorine from the generator. Waste liquid rum f rom the underground tank. The electrolytic processes 5 for separating chlorine from liquors containing INORGANIC AND THEORETICAL CHEMISTRY bromides are analogous with those employed for chlorine. The liberated bromine, however, remains in soln., and has to be separated by distillation. The purification of bromine.--The first product of the extraction contains as impurities,6 iodine cyanide, carbon bromides, bromoform, and 1 t o 4 per cent. of chlorine. This is purified before it is put on the market, and much of the bromine on the market is either free from chlorine or contains a t most up to 0'3 per cent. The chlorine is removed by distillation from alkali, calcium, or ferrous bromide.7 This is t he most convenient and effective method of eliminating chlorine. If t he bromine is initially very much contaminated with chlorine, a repetition of the treatment after converting the bromine into a bromide may be advisable where a very high degree of purity is desired. S. Piria8 added baryta water until the bromine is decolorized and evaporated the mixture to dryness. On calcination the organic matter is destroyed, and oxy-halides are transformed into the simple halides. The mass is then treated with sufficient bromine to displace iodine from iodides and againevaporated to dryness and calcined. The remaining mixture of barium chloride and bromide was leached with alcohol, which is said to dissolve the bromide and leave the chloride undissolved. The bromine is recovered from the bromide by heating i t with manganese dioxide and sulphuric acid. A . A. B. Bussy removed iodine by precipitation as cuprous iodide ; A. Adriani, by treatment with starch paste. The most convenient and effective method of removing iodine is that suggested by S. Piria, namely, converting t.he bromine into a soluble bromide, and boiling the soln. with a little free bromine. A repetition of the treatment is advisable when the bromine is very much contaminated with iodine. P. C . E . M , Terwogt agitated the bromine with an excess of water for two or three hours, and removed the bromine by means of a separating funnel. The bromine was then mixed with potassium bromide and a little zinc oxide, and distilled. 'She distillate was collected under water and redistilled. The product was dried by standing over phosphorus pentoxide, and finally distilled once again. J. S. S tas 1 employed a somewhat analogous process in 1881, According to A . S cott, this treatment removes chlorine and iodine but not organic chlorides and iodides ; t here is also a risk of the zinc oxide contaminating the product with nitric acid owing to the difficulty of purifying zinc oxide free from nitrogen. A. Scott prepared highly pure bromine by boiling 1500 grrns. potassium bromide in as e qual weight of water first with a few crystals of potassium metabisulphite and about 5 C.C. of conc. sulphuric acid, then adding 100 c.c. of sat. bromine water, distilling off the excess of bromine, adding another 100 C.C. of bromine water, distilling off once more, and, after neutralization with potassium carbonate, evaporating to dryness. The dried bromide was now fused with potassium dichromate (which had previously been fused) in the proportion of 500 grms. of bromide to 200 grrns. of dichromate. This left an excess of bromide sufficient to retain any quantity of chlorine likely to be present. 1050 grms. of the fused mass, broken u p into pieces the size of hazel nuts, were now treated with a cold mixture of 450 c.c. of conc. sulphuric acid w ith 700 c.c, of water. Any organic matter i n this mixture was destroyed by adding a small quantity of potassium permanganate. The above quantities gave about 470 grms. of bromine, and on the addition of an excess of dichromate a further 30 grrns. of bromine were obtained, which ought to contain all the chlorine. The purified bromine was dissolved in a soln. of potassium bromide-freed from iodine-and distilled. 1 E. Pfeiffer, Handbz~d der Kali-Industrie, Bra'unschweig, 1887 ;K. Kubierechky, Die Deutsche Kaliindustra'e, Halle a. S., 1907 ; M. Mitreiter, Die Cewinnung des Rroms in der Kaliindustrie, Halle a. S., 1910. a F. Mohr, Liebig's Ann., 22. 66, 1837 ; R. Hermann, Pogg. Ann., 13. 175, 1828 ; 14.613, 1828, 8 A. P rank, German Pat., D.R.P. 2251, 1877; Chem. Ind., 1. 329, 1878; K . P ietrusk~, Chem. Ind., 30. 85, 1907; F. J. H. Merril, Zeit. angew. Chena., 19. 1783, 1908; C . F. Chandler, Chem. News, 23. 77,1871. R. Wiinsche and R. Sauerbrey, Ccrman. Pat., D. R. P . 158715, 1905 ; K. Kubierschky, G., 194567, 1907. 6 lf. Kossath, Qererman Pat., D.R.P. 103644, 1897 ; F . Mchns, 2%., 134975, 1902 ; lf. Pemsel, THE HALOGENS 41 ib., 145879, 1903 ; B. R inck, ib., 182298, 1906; C . H iipfner, ib., 30222, 1881; G. A. Nahnaen, ib., 53395, 1899 ; R. W iinche, Jafirb. EEktrochem., 7 . 347, 1900 ; A. G. B etts, Bng. M ia. Journ., 783, 1901 ; M . S chlotter, Ueber die eleklrolytische Gewinwng von Brom und Jod, Halle a. S., 1907. T . L. Phipson, Chem. News, 28. 51, 1873 ; J . C. H amilton, ib., 42. 288, 1881 ; M. H ermann, Journ. prakt. Chem., ( I ) , 60.285, 1853 ; Liehig's Ann., 95. 211, I855 ; H. Poselger, Pogg. Ann., 71. 297, 1847 ; S. Reymann, Ber., 8. 792, 1875. W. R amsay and S. Young, J o u n ~ Chem. Soc., 49.453,1886. 8 S. Piria, Journ. Chim. Med., (2),4. 6 5, 1839. A. A. B. Buesy, Journ. Pharm., 23. 19, 1867 ; A. Adriani, J w ~ n Pharm. Chim., (4), 11. . 20, 1847. lo J S . Stas, Bull. A d . Belgique, ( 2 ) ,1 208, 1860 ; Mdm. Acad. Belgiq?ce, 3 3 , 1885 ; 4 2 . 0. 5. 90, 1881 ; J . D. v an der Plaats, Rec. Trav. Chim. Pays-Bas, 5. 34, 1886 ;' J . Pierre, Ann. Chim P h y ~ . ( 3f, 20. 45, 1847 ; B. Brauner, Monutsh., 10. 411, 1889 ; 8. J ahn, Sitzber. AkcuZ. W i m , , 85. 778, 1882 ; P. C . E . M . T erwogt, Zed. anoq. Chem., 47. 203, 1905; G. P. B axter, Journ. Amer. Chem. Soc., 28. 1322, 1906 ; G . P. B axter, C. J . Moore, and A. C . R.oylston, i b., M.259, 1912; T . W . Richards and R. C. W ells, Proc. Amer. Acad., 41. 410, 2906; A. S cott, J o w n . Chem. Soc., 103. 817, 1913 ; B. G. E ggink, Zeit. phyy. Chem., 54. 449, 1908. $ 9, The Preparation of Iodine Iodine is dispIaced from the iodides by any one of the other three haIogensf fluorine, chlorhe, or bromine. I a n excess be used, a compound of iodine with the halogen may. be formed. Iodine can be obtained from iodides by a process analogous t o that employed for chlorine from chlorides and bromine from bromides, namelv, by heating the iodides with manganese dioxide and sulphuric acid: ~ K I ~ f ~ ~ ~ S ~ ~ = K ~ S O~ S f 2~ ~ + I~ . 0 he iodine vapour conM HO O +M , O ~T denses in the cooler part of the retort in almost black crystals. Some potassium bjsulphate, KHS04, as well as the normal sulphate, K,S04, is formed a t the same time. The direct action of nitric acid or conc. sulphuric ac,id also liberates iodine from iodides. I n symbols : 3H2S04+2KI=2KHSU4+2H,0 +S02+12, in the latter case, and 4HN03+2KI=2KN03+2N02+2H,0+I, i n the latter. Va,riousoxidizing agents also liberate iodine from the iodides-thus, with ferric chloride, PeCl,, the reaction is symbolized : 2 K I f 2 FeC13.=2KCl +2FeCl,+I,. Most of the iodine of commerce is derived from the ashes of certain varietiee of seaweed, or from the mother-liquor--aqua vieja-remaining after the extraction of sodium nitrate from the caliche of Chili. Methods have been also proposed for extracting iodine from blast furnace gases,l from natural waters,z and from natural phosphates ; 3 b ut they have not proved to be of any commercial importance in view of the relative abundance of the Chilian supply. Not much progress has been made by electrolytic processes. T. Parker and A. E. R obinson4 proposed to electroylze a soln. of the alkali iodide acidified with sulphuric acid, in a cell with a diaphragm separating the platinum or carbon anode from the iron cathode. The iodine which separated a t the anode was to be washed with water, and dried by hot air. Formerly all the iodine was made from the ash of seaweed, and potash was a remunerative appendix to the iodine industry ; b ut just as the Stassfurt salts killed those industries which extracted potash from other sources, so did the separation of iodine from the caliche mother-liquors threaten the industrial extraction of iodine from seaweed with extinction. Iodine in a very crude form was exported from Chili in 1874-e.g. a sample was reported with iodine 52.5 per cent. ; iodine chloride, 3 .3 ; sodium iodate, 1 .3 ; potassium and sodium nitrate and sulphate, 15.9 ; magnesium chloride, 0.4; insoluble matter, 1.5 ; water, 25.2 per cent. About that time much of the iodine was imported as cuprous iodide. This rendered necessary the purification of the Chilian product ; b ut now the iodine is pursed in Chili,before it is exported. The capacity of the Chilian nitre works for the extraation of iodine Is greater than the world's demand. It is said that the existing Chilian factories could produce about 5100 t ons of iodine per annum whereas the INORGANIC AND THEORETICAL CHEMISTRY world's annual consumption is about 500 tons.5 I n order to prevent the cutting of prices by competition, the Chilian manufacturers have combined to restrict the output, and keep the supply as nearly as possible equal to the demand. As a result, a definite maximum is allocated for each works per annum; so that the plant for the extraction of iodine in any particular works is rarely a t work more than a few months each year. A l ittle iodine is yet made from seaweed in Scotland, in Norway, and in Japan. The extraction of iodine from seaweed.-The deep-sea drift-weed which is washed on t o t he western coasts o f Ireland,B Scotland,7 France, and Norway during the stormy months of spring is collected, dried, and burnt in shallow pits. The product is called kelp (formerly k ilpe, a word of unknown origin) or varec or kelp-ash i n the United Kingdom, and in Prance, says P. Lebeau,s on le designe parfois sous le nom de kelp ou salin. On the Normandy coast, the term varech is applied generally to the drift-weed or wrack which is thrown on the coast b y the sea ; a nd the ash is called kelp or s alin de varech or cendres de varech. I t he drift-weed were t o be burnt to a loose ash, i t would furnish 25 t o 30 lbs, of iodine f per ton ; i n practice, i t rarely contains more than 12 lbs. per ton. The low yield is due to f aulty treatment in calcination-e-g. ( i) b urning a t too high a temp. which causes the volatilization of part of the iodine, and the fusion or fluxing of the ash with sand and pebbles ; a nd (ii) imperfect protection of the kelp-ash from the weather whereby some of the soluble iodides are washed out by rain. High temp. burning also reduces some of the sulphates to sulphides, which later causes a high consumption of acid per unit of iodine. I n 1862, E . C. C. Stanford9 proposed the carbonization of the drift-weed in closed retorts so as t o recover tar and ammoniacal liquor in suitable condensers. This modification did not flourish because of t he subsequent difficulties in extracting soluble iodides from the charcoal. V. Vincent (1916) claims that soh. containing aluminium sulphate extract the alkali iodides from seaweed leaving behind the organic matter which prevents the direct precipitation of iodine or iodides. The alkali iodide soh. is treated with copper sulphate for cuprous iodide, or by soln. of sulphites for iodine. M. Paraf and J , A. Wanklyn proposed to heat the drift-weed first with alkali hydroxide so as t o form oxalic and acetic acids, which could he crystallized from the lixivium. The economical treatment of seaweed for iodine has been discussed by A . Puge. E . C. C. S tanford proposed boiling the seaweed with sodium carbonate, the washed residue being termed algulose ; t he aciclified filtrate gives a precipitate of what he called a lginic acid or i,mbluble algin. T he filtrate w as e vaporated to dryness, carbonized, and called kelp substitute ; i t contains the iodine and potassium salts of the original seaweed. The insoluble a k i n was converted into tlie sodium salt a n d called a k i n or soluble a b i n . Alginates of aluminium, iron, copper, and many other metals are precipitated directly by adding a metallic salt to t he soln. of algin. It was further proposed to use a lgul~w f or making a transparent tough paper, and imitations of bone or ivory ; algin, as a substitute for gelatine ; a luminium alginate in making a waterproof varnish, and as a m ordant in dyeing ; copper alginate dissolved in ammonia, as a waterproof varnish. The proposals did not prove a commorcial success. T he kelp contains 45 to 70 per cent. of soluble salts ; 0 .5 t o 1 .3 per cent. of iodine ; a nd 30 t o 50 per cent. of insoluble matters. The kelp is extracted with hot water, and the s o h . fractionally crystallized ; t he mother liqnid is treated for iodine. T he kelp is crushed into lumps-say, one to two inches diameter-and extracted with wat.er i n rectangular iron vats with false bottoms, heated by steam. The liquid of sp. gr. 1.200 to 1.258 is decanted into open iron boiling pans where i t is evaporated down to a sp. gr. of 1.325 ; t he salts-mainly potassium sulphate (80-60 per cent.) mixed with sodium sulphate and chloride--which separate by crystallization during the evaporation are removed. The hot liquid is run into cooling vats where crystals of potassium chloride separate. The liquid is again boiled down, and crystals consisting mainly of sodium chloride with 8 t o 10 per cent. of sodium carbonate-and called kelp salt---separate from T HE HALOGENS 43 the hot liquid ; t he hot liqnid, decanted into cooling vats, furnishes crystals containing 80 t o 95 per cent. potassium chloride and called muriate. These operations are repeated several times. The mother liquid is mixed with about one-seventh its volume of sulphuric acid, free from arsenic; the polysulphides and thiosulphates are decomposed with tho mparation of sulphur ; s ome hydrogen sulphide and sulphur dioxide is evolved. The liquid is d o w e d t o settle for 24 hrs. in closed led-lined wooden vats, and decanted for the extraction of iodine. The sediment is steamed for a long time t o remove adsorbed iodides. The dried mass---sulphur waste--containing 70 p er cent, of sulphur, is used in making sulphuric acid. The sulphate salts are used for manurial purposes in agriculture ; t he muriats can be used in the manufacture of potash salts ; t he kelp salt is unsaleable ; a nd the imoluble matter in the lixiviation vats, which consists principally of calcium and magnesium carbonates and phosphates, was once used in making common bottle glass. Many processes have been proposed for separating the iodide from the motherliquor resulting from the fractional crystal1ization of the aq. extract of kelp. A. Payen l o treated the acidified mother liquid with chlorine, or with potassium chlorate which in conjunction with the hydrochloric acid furnishes chlorine. An excess is avoided or iodine chloride will be formed. The precipitated iodine is washed, dried, and sublimed. Bromine is extracted from the mother liquid by neutralizing the acid, evaporating to dryness, and distillating with pyrolusite and aulphuric acid. P. J. Persoz, L. Boirault, and others have recommended adding copper sulphate mixed with iron filings or ferrous sulphate. The precipitated cuprous iodide, CuI, when heated with manganese dioxide, gives off iodine : FIG.13.-Iodine Still with Two Trains o Udell Condensers. E 3Mn02+2CuI=2Cu0+Mn304+1,. T. Schmidt precipitated the iodine as Iead iodide, PbI,. E. Moride precipitated the iodine by treating the liquor with nitric acid; J. P ellieu and M. Launay, used nitrous vapours ; R . von Wagner, ferric chloride ; M. Luchs, a mixture of potassium djchromate and sulphuric acid ; a nd I. hiercelin and L. Raure, sulphur dioxide or bisulphites. I n th; Scottish process, T the liquor was heated in stills with sulphuric and manganese dioxide as recommended by W. H. Wollaston, J. J. Berzelius, and others. The iodine sublimes. Each still consists of an iron pot covered with a leaden lid t o which are luted, say, two earthenware arms each of which is connected with a train of about five stoneware aludels or udelb supported on a wooden framework as illustrated in Fig. 1 3. E ach udell has a stoneware stopper below so as t o permit condensed water (with the 3 halogens in solution) to be drained off as required. Manganese dioxide is adcled t o the liquid. The still is heated by an open fire, and iodine and steam are evolved. The liquor from the still is a troublesome waste product. The udells are emptied when required. The bromine is in too small a quantity to pay to collect. The iodine from the udells requires further purification. The extraction of iodine from caliche.--The mother-liquor-aqua oiejaremaining after the extraction of sodium nitrate from caliche in Chili, contains sodium nitrate, chloride, sulphate, and iodate as well as magnesium suIphate. The iodine content of this liquid amounts up to about 0.3 per cent. ; as the original caljche has abont 0.02 per cent., the iodine thus accumulates in the mother liquid during the extraction of the nitrate. The mother liquid is run into wooden vatu, INORGANIC AND THEORETICAL CHEMISTRY and treated with sufficient sodium bisulphate to reduce part of the contained iodic acid to hydriodic acid : GNaHSO3+2HIO3=3Na2SO4+3H2SO4+2HI ; a nd t o get the right proportion of iodic acid to reduce the remaining hydriodic acid to The soh. is well agitated, and then neutralized iodine : HIOs+5H1=3H2O+31,?. with the so-called sal natron. liquor. After agitation, the liquid is allowed to settle. The iodine is removed, pessed into cakes, dried, and slowly sublimed in cast-iron retorts fitted with udells, as indicated in Pig. 13. The mother-liquor-called aqua fibte-is returned to the nitrate extraction tanks and is used over again.11 T he sal natron liquor, prepared by heating crude nitrate from the a qua v ieja tanks w ith 15 per cent. of coal dust, is made into a cone 5 f t . high with a kind of moat dug round the base of the cone. The cone is sat. w ith water and ignited. The crude sodium carbonate formed fuses and runs into %hepit. The product dissolved in water forms the sal natron liquor. The sodium bisulphate soln. is made by passing the fumes of burning sulphur into the s d natron liquor. The liquid acid is then acid enough t o liberate iodiq acid from iodates. The recovery of iodine from waste liquids.--F. Beilstein 1 2 recovered iodine from laboratory residues by evaporation to dryness with an excess of sodium c arbonate and calcination until the organic matter is all oxidized. The mass is dissolved in sulphuric acid and treated with the nitrous fumes, obtained by treating st,arch with nitfic acid, until all the iodine is preuipitated. The iodine is washed in cold water, dried over sulphuric acid, and sublimed. Other oxidizing agents less unpleasant than the nitrous fumes employed by P. B edstein-e.g. hydrogen peroxide-were recommended by G;. Torossian for the residues obtained in copper titrations. P. Beilstein's process is applicable to soluble but not to insoluble, oxidized forms of iodine. F. D. C hattaway and K. J. P. Orton found it b etter to heat on a water bath the waste material, solid or soln., with. aqua regia containing a slight excess of hydrochloric acid. This transforms the iodine 14---Purifka- i nto iodine chloride. The iodine chloride may be divided into tion of Iodine by t wo parts, one half decolorized by the addition of sulphurous Sublimation. acid or a sulphite, and this mixed with the other half. All of the iodine precipitates a t once. .Iodine chloride may be dil. with an excess water and allowed to stand, when the element separates in long crystals. The purification of iodine.-Crude iodine contains from 75 to 90 per cent. of iodine, some iodine chloride, iodine cyanide, water, and different salts. The iodine of commerce is purified by washing it wit'h cold water, drying by press., and subliming from heated iron retorts into ude11s.ls The following is a convenient way of purifying small quantities of iodine for analytic,al purp0se.s : 1 4 Grind, say, 6 grms. of commercial iodine with 2 grms. of potassium iodide. P u t the The beaker dry mixture in a small d ry beaker (Pig. 14) fitted with a GGckel's is surrounded with a cylindrical asbestos jmket (not shown in the diagram). Plme the beaker on a wire gauze, or a hot plate, and heat the apparatus by means of a small flame. The condenser is full of cold water, a t the temp. of the room. When violet vapours have ceased to come from the bottom of the beaker, let the ap aratus cool. A c rust of iodine will be found on the condenser. Pass a current of col water through the condenser. The glass contracts, and the crust of iodine can be easily removed by pushing i t with a glass rod into a similar beaker. The sublimation is repeated without the potassium iodide a t as low a temp. as possible. Grind the iodine in an agate mortar, and d ry i n a desiccator over calcium chloride-not sulphuric acid or the iodine may be contaminated. If t he cover of the desiccator is greased, the iodine may attack the grease, forming hydriodic mid, which might contaminate the iodine. B Numerous other modes of purification have been recommended Gom the simple process of G . S. Shrullas in which the iodine is dissolved in alcohol, the soh. filtered, and the iodine precipitated with an excess of water, to the elaborate process of T HE HALOGENS 45 J. S. Stas, in which the utmost degree of purity was desired. L. L. de Koninck heated a mixture of potassium iodide with about twice its weight of potassium dichromate; C. Meineke precipitated iodine from a soh. of potassium iodide by the permanganate ; a nd G. P. B axter converted iodine into hydriodic acid by meam of hydrogen sulphide, and after boiling the filtered s o h . a few hours to expel hydrogen cyanide, he distilled the iodine with potassium permanganate. Three repetitions of the process gave iodine quite free from cyanides. B. Lean and W. H. Whatmough converted the iodine into cuprous or palladious iodide, and sublimed, the iodine by dry distillation of the salt at about 250'. A. Ladenburg precipitated silver iodide from a soh, of potassium iodide ; and, after shaking the mixture with aq. ammonia for 24 hrs. to dissolve out the silver chloride, he reduced the silver iodide with zinc and sulphuric acid. The resulting zinc iodide was treated with nitrous acid ; the precipitated iodine distilled in a current of steam ; a nd the iodine dried over calcium chloride. M . Baubigny and P. Rivals converted the iodide into iodate by adding potassium permanganate to a soh. of the iodide in sodium carbonate. Five-sixths of the s o h , was reduced with neutral sodium sulphite and the iodine was precipitated when the reduced soln. was mixed with the remaining sixth : $HI+ HI0,=3H20+312. The iodine was filtered, washed, dried, and snblimed. The product was freed from chlorine, bromine, and cyanides. G . P. B axter oxidized the iodine to iodic acid and recrystallized the latter a number of times from conc. nhic acid. The iodic acid was heated t o 100' to drive off moisture, the temp. was then raised to about 240°, and finally heated to 350°.in a current of air. The iodide was condensed and finally remelted to remove all traces of moisture. P. Kiithner and E. Aeuer found ethyl iodide boiled a t 7Z0, t he chloride a t abont 12", a nd the bromide a t 38". Hence, if the halogen is transformed into the ethyl salt, fractional distillation enables the iodide to be separated from the chloride and bromide. I n one of J. S. Stas' processes, the iodine was dissolved in a s o h . of potassium iodide. The soh. was diluted with water until a precipitate began to form, and then three-fourths of the amount of water required t o precipitate all the iodine were added. The separated iodine was washed free from potassium iodide by decantation, the crystals, after draining, were dried over calcium nitrate in vacuo, and then distilled twice from barium oxide. I n another process, J . S. S tas purified the iodine by hst treating the iodide with ammonia which converts about 95 per cent. of it into the explosive nitrogen iodide. The washed nitrogen iodide decomposes quietly when warmed with an excess of water. J. S. Stas thus describes the procedure : Powdered iodine is added to a cold conc. soln. of ammonia in a largs flask until the dark brown liquid is nearly colourless. The resulting nitrogen iodide is washed by decantation with cold conc. ammonia until the ammonium iodide is removed. The nitrogen iodide is placed on a funnel with its neck drawn t o a fine point, and washed with cold water until the colour of the compound changes to brown, and the wash water is yellowish-brown. Themoist iodide is placed in a large glass fiask with ten times its weight of water, and .slowly heated on awater bath to 60' or 65O. I t he temp. be raised above 65O, before decomposition f is complete, c explosion may occur. The nitrogen iodide decomposes forming crystals of m iodine, asolution of iodine i n a mmonium iodide, and a white substance-possibly ammonium iodate. When t h e decomposition appsars complete, the liquid is warmed up to 100° for a few minutes. The solid iodine which separates out on cooling is washed with water on a funnel with a drawn-out neck ; a nd afterwards distilled in steam. The iodate is n ot volatilized. The iodine is dried over cdcium nitrate ; twice distilled from admixture with %bout5 per cent. of finely powdered purified barium oxide ; a nd finally sublimed alone. P. Gredt, German Pat., D.R.P. 83070, 1895. %. Carnpni, L'Orosi, 15. 263, 1892. 3 L. Thiercelin, Bull. Xoc. Chim., (2), 22. 435, 1874 ; P. Thibaalt, Dingier's Journ., 212. 339, 1874 ; Compt. Rend., 79. 384, 1874. I Parker and A. E. Robinson, Brit. Pat. No. 11479, 1888 ; C. Luckow, Z eit. anal. Chem., ! . 19. 1,1880 ; M. Schlijtter, Ueber die elektrolytiscl~e ewinnung ?*onBrom und Jod, Halle ;t. S., 1907. G 6 G . G. Henderson, Thorpe's Dictionary o Applied Chem.istry, London, 3. 142, 1912. f G . H.Kinahan, Quart. J ourn. Science, 6. 331, 1869. 1 INORGANIC AND THEORETICAL CHEMISTRY P. N eil, D. J . R oberton, and J . Macdonald, Rep Bd. Agric. Scotland, 174, 1914. P . L ebeau, H . H oissan's Traitd dc chimie midrule, Paris, I . 157, 1904. E. C. C . S tanford, Pharm. Joumt., (3), 21. 495,1862 ; M. Paraf and J, A. W anklyn, Bull. Soc. Chim., !2),7.89,1867; G. G. Henderson, m T . E . Thorpe, A Dictionary qf Applied Chemistry, London, 3. 142, 1912 ; R. Galloway, Chem. News, 38. 146, 1878; E. C . C. Stanford, ib., 35. 172, 1877 ; E . S onstadt, ib., 26. 183, 1872 ; 28. 241, 1873 ; W . H . Chandler, ib., 23. 77, 1871 ; L. Herland, I Dirzgier's Jowrn., 222. 400, 1867 ; B . Wetzig, ib., 2 N . 21 6, 1879 ; L . Thiercelin, C ~ C N .d . . 3. 199, 1880; E. Moride, Compt. Rend., 62. 1002, 1866; W . Wallace, B. A. Rep., 88, 1859 ; J. Pellieux and E . A llary, Bull. Soc. Chim., ( 2)) 34. 197. 627, 1880 ; V V incent, French Pa:. No. 480014,1916; A. Puge, L'IntE. Chim., 0. 231, 1919. l o A . B. B. Bussy, Journ Pharm., 23. 17, 1837 ; 25. 718, 1840 ; F . J Persoz, Journ. Pharm. . Chim., ( 3 ) , 12. 105, 1847 ; E . B echi, 8.,3), 20. 5, 1851 ; E . Soubeiran, ib., (3), 13. 421, 1848 ; ( L . Boirault, French Pat. No. 393668, 1907 ; J Pellieux and M. Launay, Bzdl. Soc. Chim., ( 2 ) , . 18. 44, 1872 ; M. Luchs, Pharm. VierteEjahres., 10. 536, 1861 ; T . Schmidt, Chem. News, 37. 66, 1878 ; R. v on Wagner, Wurzburqer Natunois. Zeit.,.(2), I . 94, 1860 ; M. Lauroy, Dilbgkr'~Journ. 192. 172, 1869; 211. 74, 1874; L. K rieg, ib., 154. 374, 1859; M . W hytelaw, Pogg. Ann., 39. 199, 1836 ; W . H . W oIlaston, Schweigger's Jozlrn., 40. 465, 1814 ; J . J . B enelius, Annuaire &s sciences chimiques, Paris, 8. 84, 1829 ; L. Thiercclin, Ber., 2. 79, 1869 ; L . Faure, C h m . Gdz., 13. 199, 1855 ; B rit. Pat. hTo. 355, 1864 ; E . Moride, Compt. Bend., 62. 1002, 1866 ; A . Paye& Ann. CfLim. Phy.~., (4), 4. 221, 1866 ; (41, 7. 376, 1866. W . Newton, Journ. Soc. Chem. I d . , 22. 469, 1903 ; M. Schlijlter, Ueber die elektrolytische Gewinnuq von Brom undJod, Halle a. S ., 1907 ; L . Faure, Chem. G az., 1 3. 199,1855 ; M . Loire and M . W iessflog,Diqler's Journ., 253.48,1884 ; C*. Langbein, ib., 231,375,1879 ; L. Thiercelin, Bull. Soc. Chim., (21, 11. 186, 1869 ; J. B uchanan, Berg. H u t . Ztg., 53. 237, 1894. l. " BeiIstein, Zeit. Chem., 13. 528, 1870 ;L . Henry, Ber., 2. 599, 1869 ; F . D. C hattaway and K . J . P. O rton, Journ 8oc. Chem. I d . , 18.560,1899 ; J . H . Gladetone and A. T ribe, Journ. C h m . Soc., 43. 345, 1883 ; R. Dieterich, P h a m . Cent&, 24, 1896 ; A. Olig and J . T illmans, Zeil. Unters. Nahr. Genuss., 11. 95, 1906 ; A. h r g e u , Compt. Rend., 102. 1164, 1886 ; H. W . Gill, Analyst, 38. 409, 1913 ; G . Torossian, Journ Ind. E ng. Chem., 6. 83, 1914. l a W . Newton, Journ. Soc. Chem. I nd., 22. 469, 1903 ;J :Hertkorn, Chem. Ztg., 16. 795, 1802. l4 J . W . Mellor, A Treatitye on Quantitative Inorganic Analysis, London, 298, 1913. l6 H. Gocliel, Zeit. angew. Chem., 12, 494, 1899. 16 L . L. de Koninck, Bull. As.wc. Chim. Beb., 17. 15, 1904 ; J . S . Stas, Ndm. Acad. BeZp$ue, 35. 3 , 1865 ; CE'ztvreb ComplQtes,Bruxelles, I. 663, 1894 ; C . C . W illstein, Diqler's Journ., 200. 310, 1871 ; C. F. Mohr, Lehrbuch &r chemische-analytischen Titrirmethode, Braunschweig, 269, 1874; G. P. Baxter, Journ. Am&. C h m . Soc., 26. 1577, 1904; 32. 1591, 1910; A. Gross, ib., 25. 987, 1903; Ckem. News, 88. 274, 1903 ; L . W . .Andrews, ib., 90. 27, 1904; Amer. Chem. Journ., 30.428,1903 ; B . Lean and W . 8.W hatmough, Proc. Chem. Soc., 14. 5,1898 ; C . Meineke, Chem. News, 68. 272, 1893 ; Chem. Ztg,, 16, 1210, 1230, 1892 ; Z . Musset, Zed. a n d . Chem., 30. 48, 1891 ; G . Lunge, Zeit. angew. Chem., 7 . 234, 1894 ; A. Ladenburg, Ber., 35. 1256, 1902 ; P. Rijthner and E. Aeuer, ih., 37. 2536, 1904 ; M. B aubigny and P. Rivals, C m p t . Rend., 137. 927, 1903 ; J. L. Mayer, Amer. Journ. Pharm., 87. 154, 1915; G. Fouqu6, Bull. Soc. Chim., (4), 19. 370, 1916 ;W . H. Miller, Ann. Chim. Phys., (3),9. 400, 1843; A. M ikeherlich, Sitzher. A k d Berlin, 409, 1855 ; G . S . SC?rullas,Ann. CRim. Phys., (2), 42. 260, 1829. " § 10. The Physical Properties of ChIorine, Bromine, a nd Iodine A t ordinary temp. fluorine is a gas with a pale canary-yellow colour, chlorine is a gas with a greenish-yellow colour, bromine is a dark reddish-brown liquid which readily forms a reddish-brown vapour when warmed ; a nd iodine is a dark bluishblack crystalline solid which gives a violet-coloured vapour when heated. The colour of the halogen gases is therefore deeper and more inclined to the violet end of the spectrum, the greater the at. w t. The colour of bromine and chlorine gradually becomes paler as the temp. is reduced. At the temp. of liquid air, bromine is pale yellow, chlorine almost colourless. J. H. Kastle 1 t ried t o show that the characteristic colours of the halogens can be explained on the assumption that the molecules are slightly dissociated even in the solid state, and the observed colour is that of the dissociated halogen. He based his argument mainly on the facts : ( I) t he colour of the halogens is invers,ely as their chemical activity ; (2) t he least stable halogen compounds are the most highly coloured ; (3) on heating, the halogen compounds become deeper in tint ; (4) t he change of colour of bromine on cooling is thus said to be an effect of diminished dissociation (iodine is steel-grey in colour a t -190" a nd a t ordinary temp.). FHE HALOGENS 47 The odour of chlorine is most disagreeable and suffocating ; if but small quantities of the gas are present its odour recalls that of seaweed. Chlorine attacks the membrane of the nose, throat and lungs producing irritation, a kind of bronchal coughing, and spitting of blood ; t he lungs become inflamed, and this is followed by painful death. M . von Pettenkofer and K. B. Lehmann 2 found that 0.001 to 0.005 per cent, of chlorine in the air affected the respiratory organs ; 0 . 0 4 4 9 6 produced dangerous symptoms., whilst concentrations exceeding 0.06 per cent. fatal. Bromine like chlorine has a very unpleasant irritating smell ; rapidly it attacks the eyes very painfully, and is an irritant poison.; when in direct contact with the skin it produces troublesome sores. M. von Pettenkofer and K. B. Lehmann add that men cannot stand more than 0.002 to 0.004 per cent. if not habituated; and if h abituated, not more than 0.01 per cent. The smell of iodine is not so obtrusive since it is solid a t ordinary temp. ; it too has a faint smell recalling that of chlorine, but is less unpleasant and less irritating. The crystalline f o r m of t he halogens.-According to W. Wahl, the crystal8 of chlorine; bromine, and iodine belong to the rhombic system,3 and they are isomorphous. The crystals are strongly doubly refracting in sections both parallel and at right angles to the longest axis ; a nd the extinction between crossed nicols is parallel to the longest axis. The optical properties of crystals of bromine are similar to those of chlorine; the crystals have a tendency to develop prismatic forms, and the prismatic cleavages-angle 70"-are very distinct. H. Arctowsky obtained slender carmine-red needle-like crystals of bromine which recall those of chromic anhydride. From B. J. K arsten's and P. C. E . & I. Terwogt7s observations on the m.p. of the binary systems, Cl-Br, Cl-I, and Br-I, t he halogens can form a continuous series of mixed crystals, and they thus appear isomorphous. E. Mitscherlich showed that it crystallizes in the rhombic system with a priam angle of 67" 12' ; a nd E. 8. von Federoff found that some crystals on the asbestos stopper of a reagent bottle contained both the ordinary rhombic tablets and prisms belonging to the monoclinic system. Both forms can be obtained from soh. in carbon disulphide, chloroform, petroleum ether, and alcohol. The rnonoclinic crystals are formed by rapid evaporation, the rhombic form by slow evaporation. V. Kurbatoff found the sublimation of iodine above 46.5' gave the ordinary rhombic crystals, and a t lower temp. the monoclinic crystals. Hence, it h as been said the iodine is dimorphous with a transition point a t 46.5'. W. Wahl could not find a transition point by cooling the ordinary form down t o -180°, and considers it is not probable that there is a transition point. He believes that the ordinary form is stable a t all temp., and that the monoclinic prisms belong to a monotropic form with a marlred temp. limit of formation, and apparently with a low velocity of transformation a t ordinary temp. With chlorine, the polarized light travelling parallel to the cleavage axis is more strongly absorbed, and the transmitted light a deeper yellow with a greenish tint than that passing in directions a t right angles to the principal axis. There is a slight difference in the degree of absorption in the two directions a t right angles to the principal axis. The crystals of bromine are pleochroic, being dark brownish-red in the direction of the prism axis ; yellow-red in the direction of a line bisecting the smaller prism angle ; a nd pale yellowish-green in the direction of the line bisecting the larger prism angle. The change in colour of solid bromine from brownish-red to nearly black a t the m.p., and pale yellow a t the temp. of liquid air, and to a still paler tint a t the temp. of liquid hydrogen, is principally due to a gradual disappearance of the strong trichroism which i t preserves near the m.p. The crystals of iodine s appear black or light reddish or leather brown according as the polarized light i transmitted wit'h the direction of the principal axis parallel to the plane of polarization, or a t right angles to the principal axis. The pleochroism of these three members of the halogen group increases in strength and character as the at. wt. increasesthe colour of the strongest absorption in chlorine is the same as the weakest in bromine ; a nd the strongest in bromine, about the same as the weakest in iodine. INORGANIC AND THEORETICAL CIHEMISTRY The behaviour of t he halogens towards the gas laws.-One litre of chlorine gas a t 0" a nd 760 mm., latitude 45", a nd a t sea-level, weighs 3.1667 grms., when 61 l itre of oxygen under similar conditions weighs 1.42900 grms. The theoretical density with respect to air is 2-4494-under standard conditiow. The value observed by J . L . Gay Lussac and L . J. ThBnard G s 2.47 ; R . Bunsen, 2'4482 ; A. Leduc, 2.491 ; H. Moissan and A. B. du Jassonneix, 2.490 grms. According to E. Ludwig's the vapour density, D, of the gas a t a temp. 0" between 0" and 200" is D=2.4855-Ob000170. According to E. J ahn, above 200°, chlorine follows Boyle's law exactly ; M. Pier found that the coeff. of expansion a in the expression v=vl(l+aO) between 0" and 50s24", a=0'003873 ; between 0" and 100'4", a=OW003833 between 0' and 150*7", a=0'003814 ; a nd between 0" and 184", ; a=OS003804. Similar observations apply to the constant /3 i n the expression q=%(1+/36)' which decreases from 0.003807 a t 100" to 0.003774 a t 184-4". With rising temp. or decreasing press., the behaviour of chlorine approximates more and more nearly to that of an ideal gas. R. Knietsch's values 6 for J. D. v an der Waals' constants are : a nd for bromine, a=0.01434 ; b=od00202. The v apour densities,-The observed densities of chlorine, air unity, are Vapour density . . 0" 2 ,4910 100° ZOO0 2.4615 2.4520 O 0 t o 1200' 2,450 1400' c. 2 .02 The ideal value of Cle is 2.4494. The greater density of chlorine below 200" is attributed by M. Pier t o polymerization into C14 molecules ; a nd on this assumption he has calculated the degree of di~socia~tion C14=2C12, corresponding with the deviations of the observed densities from the ideal value. This assumption is not supported by other evidence, for the molecular condition of liquid chlorine appears to be the same as the gas, elz. The vapour densities of chlorine a t high temp., determined by V. Meyer and his co-workers,7 have a value lower than the norma.1, and indicate that chlorine is appreciably dissociated into one-atom molecules: C12=2C1, a t temp. exceeding GOO0, or else that the coeff. of expansion of chlorine is greater than normal. M. Reinganum obtained no evidence of dissociation at 1137" if precautions be taken to prevent errors arising from the tendency of chlorine t o diffuse through the walls of the apparatus. Bromine forms a reddish-brown vapour a t ordinary temp. with a vap. press. of 138.1 mm. a t 15") a t which temp. E. P. P ermans found the vapour density at 15" to be quite normal, but according to H. J ahn, it is rather higher than the normaI value 5.5149 if a ir be unity or 79'92 if oxygen 16 be the unit, for a t 102'6" he found 5.7280 ; a t 175'58", 5'6040 ; a t 227'92", 5.5243 ; a nd he represents the observed vapour density, D , a t 8" by the empirical formula D=5.8691-0'001530. H. J s h n assumes that the molecules are polymerized a t the lower temp. and that with rising temp. or decreasing press., say, by dilution with nitrogen, t h i density approaches the normal value corresponding with two-atom molecules.- At higher temp. the vapour densities are lower than the normal value, and this the more the higher the temp. It is therefore assumed that there is a n increasing dissociation : Br2+2Br with rise of temp. ; a nd calculations from the observed deviat,ions of the vapour density from the ideal value for Br2 show that the percentage dissociation a t Dissociation .. 800" 850' 900" 950" 100" 1060' 0 .18 0.20 1.48 2 .53 3 -98 6.30 1284' 18.3 per cent. M . Bodenstein and F. Cramer represent the relation between the dis~ocia~tion and temp. by the formula T HE HALOGENS Similar results have been observed with iodine, but the dissociation 9 is much more marked a t even lower temp. The theoretical density for the two-atom molecule is 8,758, air unity ; or 126.92 oxygen=16 ; a nd for the one-atom molecule the theoretical density is 4.379, air unity. When iodine vapour is heated above 700" its density diminishes steadily up to about 1700", when i t becomes constant a t half , ' its value a t the lower temp. . Temp. Vapour density Dissociation a . . 4 80' 8.74 0 855' 8.07 8'6 1043" 7.01 25.5 1275" 5.82 50'5 1390" 5.27 66'2 1468" 5.06 73.1 per cent. Without, doubt, the iodine molecule, 12, dissociates into atoms : I z=I+I. The state of t he system in equilibrium will be represented by kC1,=k'C~2. If x denotes the proportion of iodine dissociated, and v the volume of t he iodine vapour, then, since v volumes of iodine vapour become 2v volumes of dissociated iodine vapour, it follows that the concentration of the dissociated iodine will be xlv, and of the undissociated iodine ( I -z)lv. Hence for equilibrium In every mol. of iodine (Iz) a t 1043", 0'25 mol. will be dissociated ; hence, x2=0*0625 ; 1-x=OB75 ; a nd K=0-0833/v. To evaluate v, remember that one mol. of iodine vapour a t O0 a nd 760 mm. occupies 22.3 litres ; a nd a t 1043", 107.5 litres. This quantity of gas contains 0.25 more molecules of iodine because of dissociation, and hence its volume is 107-5+$ of 107.5=134.4 litres. Hence K=0'0833+134.4 =On00062; or k : k'=Om00062 : 1 ; or 1 : 1600 (nearly). Otherwise expressed, Clz=1600 ,2, that is, the atoms of iodine will u'nite 1600 times as fast as the molecules C dissociate under such conditions that unit concentration of each is present. The dissociation of'iodine molecules is a unirnolecular reaction because one molecule is concerned in the reaction; and the formation of the two-atom molecule by the union of two one-atom molecules is a bimolecular react'ion becanse two molecules are concerned in the process. G. Starck and M. Bodenstein represented the relation between the dissociation constant K , a nd the absolute temp. T, by the relation loglo K,=-7761'96T-1+lS75 log T-0.00041566T+0.422, which is closely in accord with observations : 1173' 1273" 1373" 3473" R. Temp. 1073' - 1.325 -0.782 -0.309 -0.091 Log,, K v (Observed) . - 1.945 - 0.084 - 0 .771 -0.31 1 Log,, K v (Calculated) . - 1,956 - 1.340 For the sake of comparison, the dissociation constants K o f chlorine, bromine, and iodine are respectively 0.01 (1670°), 0.06 (1050°), and 0.66 (1390"). The heat of dissociation 12+21, calculated from G. Starck and M. Bodenstein's equilibrium measurements is 35-67.cals.a t 1073O K., and 39'64 cals. a t 14-73' K . ; while G. N. Lewis and M. Randall calculate that the increase of free energy in passing from I, t o 21 is 33560-3.50T log T+0.0020T2-1-99T cals. I. Langmuir obtained evidence of the formation of atomic chlorine b y heating chlorine under a low press. by means o a tungsten filament as in the analogous production of atomic hydrogen-3.v. f The specific gravities of liquid and solid.-The sp. gr. of liquid chlorine has been determined by R. Knietsch 10 over a range of temp. from -80" to 7T0, and A . Lange has also obtained results in close agreement with those of R. Knietmh. The latter represents his sp. gr. D a t the temp. 8, by the empirical formuIa: D=ls6588346-0.002003753(8 +8O) -O.OOOOO4559674(8 +8O)2, with a mean error k0'00148; P. M. G. Johnson and D . McIntoeh give the formula D=lm725 -0'00243 (100+8). A selection of A . Lange's results for the sp. gr. of liquid chlorine are . Sp. gr. . S p.gr.. - 50° 1.5950 30° 1.3799 VOL. T. I - 40" 1.5709 40' 1-3477 - 30' 1,5468 60" 1.3141 -20° 1.5216 60" 1.2789 - 10° 1,4957 70" 1.2421 0' 1.4485 80" 1.2028 10° 20° 1.4402 1.4108 l oo0 90" 1.1602 1-1.134 E INORGANIC A ND T HEORETICAL CHEMISTRY M. Pellaton represented the sp. gr. of liquid chlotine a t 0" by the formula a nd for the sat. vapour, D=0'48219+0902451(144--8)+0.068526(144-8)t. The results are in close agreement with the law of rectilinear diameters. The sp. gr. of liquid chlorine at the b.p, is 1.568 ; a nd for liquid bromine, according to W. Ramsay and D. 0 . Masson, 2.9483. T he values for the sp. gr. of liquid bromine by the early investigators show considerable variations because the bromine they used was impure and probably contaminated with much chlorine. A. J . Balard's 11 number was 2.966. T . E . Thorpe's number is 3,18828, a t 0" water a t 4" unitv ; a nd 3.15787 a t 9.1". L. W. Andrews and H. A. Carlton give 3.11932 a t 20°, 3'10227 a t 25O, a nd 3.08479 a t 30". According to H. Billet, the sp. gr. of liquideiodine at 107" is 4.004 ; 3'944 a t 124.3" ; 3'918 a t 133.5' ; 3.866 a t 151'0" ; a nd 3.796 a t 170" ; a nd, according to J. D rugman and W. R amsay, 3.706 a t 184-8". J , Dewar obtained for the sp. gr. of solid iodine, 4.8943 a t - 38'85; A. Ladenburg, 4.933 a t 4' ; J. S. S tas, 4.948 a t 17" ; H . Billet, 4,917 a t 40" ; 4.886 a t 60" ; 4.857 a t 79.6' ; 4.841 a t 89.8" ; a nd 4.825 a t 107". G. 1 eBas estimates the at. vol. of the haIogen atoms in combination relative t o combined hydrogen, a t the critical point, to be D=Os687014 3-0-a002379(144 -0) + 0 ° 0 6 2 2 1 0 9 ( , Critical temp. . H T he compressibility coefl[icients,-The I? C1 22.6 2.3 5 8.7 6-0 Br 71.6 7 .7 I 100-6 10-4 mean compressibility of liquid chlorine a t 20" under the influence of one megabar, i .e. 0'987 a tm., is 0.000116 f or press. between 0 a nd 100 megabars, and 0.000095 between 100 a nd 500 megabars ; for liquid bromine between 0 a nd 100 megabars press., 0'0000613, a nd between 100 a nd 500 megabars, 0.0000518 ; for solid iodine, 0'000013 b dween 100 a nd 500 megabars press. The compressibility of liquid chlorine, says T. W. Richards,l2 the highest 01 all the elements, seems to be connected with its large at. vol., its great reactivity, and its volatility, since substances which already possess a large cohesive press. would be naturally less influenced by an external press. The compressibility of solid chlorine is probably less than 5 0x10-6 a nd may be as low as 25 x 10-6 ; a nd the compressibility of solid bromine is probably less than 30 X 10-6. T he surface tension.-The surface tension of liquid chlorine 13 a t -72" is 33'65 dynes per cm., 3 1-61 a t -61.5" ; 29.28 a t -49.5" ; 26.55 a t 35-3" ; a nd 25.33 a t -28.7. The temp. coeff. of the moleoular surface energy is 2-04, very near to the characteristic value for non-associated liquids, and hence it is supposed the molecules of liquid chlorine are present in the state of two-atom molecules, Clz. The surface tension, a, of liquid bromine 14 a t 0" is a =4290(1-0903810) dynes per cm. The values observed by W. Ramsay and E . Aston are 40'27 dynes per cm. a t 10'6 ; 34'68 a t 46'; a nd 29.51 a t 78.1". T he temp. coeff. agrees with the assumption that the molecules are not more complex than is represented by Brz. According t o R. Schiff, the atomic cohesion, a2, on the assumption that the capillary constant is an additive quality, are, in terms of hydrogen unity, 7, 13, a nd 19 for chlorine, bromine, and iodine respectively. d The viscosity m fluidity.-According to T. Graham,15 the coeff. of viscosity of chlorine gas is 1.287 x 10-4 a t 0 °, a nd 1-470x 10-4 a t 20". According to A. Campetti, the viscosity of chlorine is 1.328 x 10-4 a t 15", a nd it is not affected by the arc-light filtered through a dil. soh. of cupric sulphate to free it from the less refrangible rays. A. 0. Rankine found 1.297 x 10-4 a t 12.7", a nd 1.688 x 10-4 a t 99.1' for chlorine; and for bromine, 2.48 x 10-4 a t 223.4" ; 1-885 x 10-4 a t 99.8". A t To absolute, the viscosity of bromine vapour is 0.00002158T~/(1+4602'-1). A t the critical temp. the viscosity of chlorine is 1.897 x 10-4, a nd of bromine 2-874 x 10-4. According to T.E. Thorpe and J. W. Rodger, the viscosity of liquid bromine is 0'01245 a t 0.56' ; 0'01035 a t 16'16"; 0'00848 a t 35-86'; and0.00706 a t 56-48'. According t o E . C. Bing- THE HALOGENS ham, the fluidity of bromine-i.e, the reciprocaI of the viscosity-represented by $ is related with the absolute temp. by the formula T--0.79098$--1376.3t/-1 -i-227.16 with an error not exceeding 0 .03 per cent. The viscosity of liquid iodine is 2.252 a t the m.p. The intrinsic press., K, of the liquid halogens calculated by P. Walden from the relation K=a/u2, where a is van der Waals' constant, and v is the mol. vol. a t the b.p.-is 1060 a tm. for fluorine, 2135 a tm. for chlorine, 2530 a h . for bromine, and 2830 a tm. for iodine. The collision frequency, 6270 x lo6 per second ; a nd the coeff. of condensation from gas to liquid DgaE!Dliwid=0.00301. G . Jager's l6 estimates of the molecular d iameter of fluorine, chlorine, bromine, and iodine calculated fr'om the electrical conductivity of salt soln., are respectively 135 x 10-Q, 96 x 10-0, a nd 9 1 x 10-8 cm, The estimates of the diameter of the sphere of action of chIorine based on the kinetic theory of gases furnish numbers ranging from 3 .28 x 10-8 t o 4 .96 x 10-8 cm. A. 0. Rankine's estimates, based on the viscosities, are 3.15 x l O - - g cm. for the diameter of the chlorine molecule, and 3 -36 c ~ 1 0 - 8 m. for that of bromine ; a nd 1 -30x 10-22 C.C. for the volume of the chlorine molecule, and 1'59 x 10-22 C.C. for .that of bromine. G . J ager calculated for the mean free p ath 2 .9 x 10-8 em.-0. E . Meyer gives 4 %x 10-6 cm. ; a nd for the square root of the mean square of the molecular velocity, 3 .07 x 104 cm. per second ; and the arithmetical mean 2.86 x 104 cm. per second. The coefficients of thermal expansion.--The coeff. of cubical expansion of liquid chlorine follows from the determinations of sp. gr. a t various temp. According to A. Langa's data," the coeff. of expansion of liquid chlorine, a, is , The constant thus increases in magnitude as the temp. rises, until, a t abont 90°, it is as large as that of the gas. R . Knietsch gave a =0'001409 from -30" t o 0 °, 0'001793 from 50" t o 60" ; a nd 0 .003460 from 70" t o 80". T. E . Thorpe represents the expansion of liquid bromine at 6" by the einpirical formula 1 +0~0010621800 $0.0000018771482-0'00000000308583. J . I . Pierre gave 1 + 09010381862558 $0~00000171138085382+0~00000000544711863. The constants in the latter formula can be curtailed became after, say, the fourth significant figure the numbers are a 1 out of perspective with the accuracy of the measurements. According to 1 H.BilIet, the coeff. of cubical expansion of solid iodine is 0.0002350, a nd according ; to J. Dewar 0.0002510 between -38.85" a nd l iO t he coeff. of thermal expansion for liquid iodine is 0.000856 according to H. Billet. Thelique£action of chlorine,-The history of the liquefaction of chIorine is interesting. B. Pelletier in 1785 a nd W. J . G . K arsten in 1786 contested the view that chlorine is' a permanent gas because they showed that yellow crystals were formed when the gas is cooled. These crystals were regarded as solid chlorine. I n 1810, however, H. D avy 18 showed that these crystals were not formed a t -40" F. if dry chlorine be used ; t hat a soln. of chlorine in water freezes more readily than water alone ; a nd that the crystah contain water. He adda : " The mistake seems to have arisen from the exposure of the gas to cold in bottles containing moisture ; " and in 1823, M. Paraday showed that what chemists called solid chlorine about the end o the eighteenth and beginning of the nineteenth centuries, is chlorine hydrate. f On March 5, 1823, M. Faraday was operating with chlorine hydrate in a sealed tube. Dr. J. A. Paris l Q called a t the laboratory and noticed some oily matter in the tube Faraday was using ; he rallied Faraday " upon the carelessness of employing soiled vessels." Faraday started to open the tube by filing the sealed end; the contents of the tube suddenly exploded ; a nd the " oil " vanished. Faraday repeated the experiment, and Dr. Paris, next morning, received the laconic note : DEAR SIR,-The o il you noticed yesterday turned out to be &quid chlorine.-Yours faithfully,MIUHAEL FARAD AY. 52 1.NORGANIC AND THEORETICAL CHEMISTRY Chlorine can be condensed t o a golden-yellow liquid a t 0" and 6 atm. press. By sealing chlorine hydrate in one limb of a A-shaped tube, and pIacing that leg in warm water while the other leg is immersed in a freezing mixture (Fig. 15) of, say, ice and salt, yellow oily drops of liquid chlorine condense in the cold limb. M. F araday was much troubled with his tubes bursting, and refers to the personal damage he sustained in this way. He speaks of his eyes being fiIled as with broken glass, and of explosions so violent as to drive pieces ef gIass through window panes " like pistol shot." M. FIG, ld-Liquefaction of B araday was unable to freeze liquid chlorine by cooling it Chlorine. t o -40" ; b ut in 1884, K. Olschewsky obtained yellow crystals by cooling the liquid in a bath of evaporating ethylene. The critical ~ 0gtants.20-For the critical temp. of chlorine, J. Dewar gave 141" ; R . Knietsch, 146" ; A. Ladenburg, 148" ; a nd M. Pellaton, 144". The critical temp. of chlorine is about 145" ; t hat of bromine, 302" ; a nd of iodine, 512" (estimated). According to T. Andrews (1871) : I If a fine tube be hermetically sealed when one-half of the tube is filled with liquid bromine, and one-half with the vapour of bromine, and graduaIly heated until the temp. is above the critical point, the whole of the bromine becomes quite opaque, and the tube has the aspect of being filled with a dark red opaque resin. Even liquid bromine transmits much less light when strongly heated in an hermeticaIly sealed tube than in its ordinary state. R . Knietsch gave for the critical press. of chlorine 93.5 atrn., and of bromine, 131 atm. (estimated). J. Dewar gave for chIorine, 83.9 atm.? and M. Pellaton, 76'1 atm. The criticaI volume' of chlorine is 0.00615, and of bromine, 0.00605. M . Pellaton gave for the critica12densityof chlorine 0.573. The boiling and melting points.-H. V. Regnanlt 21 gave -33'6" a t 760 mm. for the boiling point of liquid chlorine ; a nd M. YeIlaton gave -34.5". R. Knietsch determined the vapour pressure of liquid chIorine a t different temp. ranging from -88" t o 146", which latter he regards as the critical temp. For temp. below the,b.p., R . Knietsch found the vap. press., in mm. of mercury, to be Vap. press. . 3 4'4" 7 10 -40' 498 -49'5O 365 -60" 217 -- 66" 155 -73" 100 -83" 50 - 88' 37.5 mm. a nd for temp. above the b.p. the vap. press., in atrn., were Vap. press. . -33.6' 1 .00 0" 3-66 20'85" 6-79 40" 11.5 70" 2 3.0 100" 41.7 120° 60'4 146" 93.5 a tm. IC. Knietsch represents his determinations of the vap. press. p of liquid chlorine a t a temp. 0°, between -33.6" and 0°, by the empirical formula: p=760 +32.9127(6+33.6)+Om810597(6+33'6)2 mm. ; those between 0" and 40" byp=2781 +82.301668+1.537029382 mm. ; a nd those between 40" and 146" by 11.5 +0.192966(0-40) +0.005365(6-40)2 atm. M. Pellaton uses log p=4'922232-BT-1 -C log T atm., where log B=2'9676491 ; and log C=1-8967405. A. Juliusberger gives loglc p =4-90847-393+3T-l+0*87093T mm. between 185" and 419" K. For liquid chlorine, M. Pellaton 22 found the value off in J. D . van der Waals' equation log (p,/p)=f(Tc/T-1) is nearly 2'5 ; t he ratio of the critical density to that calculated by the formula of a norma1 gas D=Mpd22412(1 +a6,), namely 0'15765 ; 3.635=Dc/D ; a nd an application of Trouton's rule, 67.5 x 70-92,238'5 gives 20'67. Each of these three values is characteristic of what is obtained with non-associated liquids. The reported b.p.23 of bromine are very discordant ; numbers ranging from 45" to 63" have been given ; A , G. Balard's number was 47". The best representative value is 59" a t 760 mm. If extreme trouble is taken to ensure accurate physical T HE HALOGENS 53 measurements on material which has not been prepared with extreme care so as to ensure the highest degree of purity, the work will be all out of perspective. The bromine of the early observers was, without a doubt, contaminated with x per cent. of chlorine. The vap. press, of liquid bromine in miIlimetres of mercury are, according t o W. R amsay and S. Young : H. W. B. Roozeboom's values are a little higher. W. Ramsay and 8. Young's and C. and M. Cuthbertson's values for the vap. press. of solid bromine are -800 +e3'0° - 4@0° --2KB0 - 14.0" -12'0" - 8.4" - 7'0 Vap. press. . 0 .13 0.66 1 - 83 7.74 25.0 30.0 40.0 45.0 mm. T. Isnardi observed 65-83 mm. a t 0" ; 35.37 mm. a t -10.9°; 24.95 mm. a t -15'8"; and 15.75 mm. at -21.1". The tripIe point is -7.3" and 46.4 mm. The vap. press. of solid bromine is given by log P=-7109.142T-1-43"33195 log I' +133'46929. The molecular rise of the b.p. of liquid br0rnine,~4 k, a t ordinary press. is 52-the calculated value is 49.5. Rough estimates of the b.p. of iodine were made by 5 . L . Gay Lnssac,25 and H. V. Regnault. Later determinat,ions were made by W. R amsay and S. Young, who found the vap. press. of the liquid, in mm. of mercury, to be 114'1" 1 20'4" 127'1" 8 0.8 113.4 142-9 Vap, press. 16&B0 475.0 169'4 174'5" 180'75" 1 84'4" 505.5 575.3 6805 764.2 mm. J. Dewar represents the vap. press., p, of liquid iodine by the formula : log p = 7.924-2316T-1. For solid iodine, G. P. Baxter, C. H. Hickey, and W. C. Holmes, between OP a nd 55", and W. Ramsay and S. Young, between 58.1" and 113.8", found the vap. press. : Yap. press. . 0" 15" 0 .030 0.131 30' 0.469 50" 64'5" 80'4" 102'7" 2-154 6-05 15-15 50.65 113'6" 87'0 mm. R . Naumann obtained a vap. press. of 0900004 atm. a t -21"; his other results are lower than those in the above table. J. Dewar 2G represents W . Ramsay and 8. Young's vap. press. p of solid iodine between 58.1 and 113.8" by log p=9.3635 W. Nernst -2872T-1; and between 85" and 114.1°, by log p=10.0392-3137T-1. represents the vap. press. of solid iodine by the expression log p=-3196T-1 t1.75 Iog T-0Q03128T+440, where 4.0 represents the so-called chemical constant of iodine. The formula agrees very well with R . Naumann's measurements. The'melting point of chlorine, determined by K. Olschewsky,27 is -102". The values for the m.p. of bromine were very discordant before those undertaking the measurement of physical constants realized the vital importance of carefully purifying their materials. The numbers which have been reported range from -7.5" A. J. Balard gave -18". The more recent determinations group themto -25'. selves about -713" as the best representative value for the m.p. of bromine.2S The m.p. of iodine is 113.6" according to H. V . Regnault ; 114" according to W. R ammy and S. Young; and 116.1" according to A. Ladenberg.29 A liquid can exist only when the presa, is greater than its vap. press. ; when the press. is less, the substance can exist only as a gas. If a fusible substance is under a lower press. than correeponds with its vap. press. a t the map.,it cannot melt when heated, but passes a t once into the gaseous state-this press. has been called the critical pressure of t he did.30 The principle is readily illustrated b y the following experiment : If a solid piece of mercuric chloride be placed in a glass tube closed at one end and connected with an air pump at the other, it is impossible to melt the salt when the press. is below about 4 00 m m., however great the temp. applied ; the solid merely sublim6s. If the press. rises above 4 50 mm. the solid fuses. The vap. press. curve of solid iodine is indicated by PO, Fig. 16; that of liquid iodine by OC ; a nd the effect of press. on the m.p. of iodine by ON. At the tiiple point 0 these curves meet. Fig. 18 shows a similar curve for water. The curve PO thus represents the sublimation curve or hoar-frost line ; OC, the boiling or vaporization curve, i .e. the effect of press. on the b.p. of the liquid. The same phenomenon occurs with water, iodine, etc., and the principle involved is the same as indicated in the law represented by Clapeyron-Clausius' equations with respect t o the lowering of the m.p. by an increase of press. Consequently, i f the vap. press. of iodine be less than that of t4e triple point, the soIid does not melt, but rather sublimes directly without melting a t the triple point a t 114.15' (89.8 mm.)' and A. von Richter a t 116.1" (90 mm.). According to R. W. Wood, if t he condensation of iodine vapour occurs above -60°, a black granular deposit is formed, but below that temp. a deep red film is produced. The melting of bromine or of iodine is attended by an expansion-with bromine, J. I. Pierre 31 found a 6 per cent. expansion. M . Toepler found an expansion of 0.0511 C.C. per gram of bromine, and 0.0434 C.C. per gram of iodine. Hence, by Clausius and Clapeyron's equation the m.p. of bromine is raised 0.0203" per atm. rise of press., and iodine, 0.0314" per atm. rise of press. The heats of vaporization and fusion.-According to R . Knietsch,32 the heat of vaporization of liquid chlorine is 67.38 cals. per gram, or 4.75 CaIs. per mol. a t -22". and 62-7 cals. per gram at So. T . Estreicher and A. Schnerr found a t -35.S0, 61.9 cals. per gram or 4.39 Cals. per mol. & I. PelIaton gets 64.7 cals. a t A -22", and 58.2 cals. a t So. T.' ndrews found the latent heat of vaporization of bromine a t its b.p. to be 45% cals. per gram, and H . V . R egnadt found that 50.95 &ds. were involved in condensing toa liquid one gram, N Cof bromine vapour a t its b.p. and cooling the liquid to 0"; if the sp. ht. of liquid bromine be 0'108, the heat of vaporization is 44.15 cals. per gram a t the b.p, The latent heat of vaporization of solid bromine, according to T. Isnardi, is 60'7 cals. per gram ; and for iodbe, according to K. Tsurata, 81 cals. per gram; and the mol. ht. of vaporization of liquid iodine is 10.57 Cals., or 10'65 Cals. according to J. Dewar. This is in good agreement with Trouton's rule, as is also the case with bromine and chlorine. According to T. Estreicher and M. Staniewsky, the heat of fusion of solid chlorine is, at --108", 22'96 cals. per gram ; H. V. Regnault's value for solid bromine is ~ernperd ures L 16.185 cals. per gram, or 2'6 Cals. per mol. The calFra. 16.-Vapour Pressure Curves of s olid and ~ i ~ culated heat of fusion of iodine is 2'92 cals. per gram, ~id I odine. or 3'29 Cals. per mol., and J. Dewar's vaIue for the difference betureen the observed heats of 'sublimation and vaporization is 3-78 Cals. per mol. The heat of sublimation of iodine a t its m.p. is 14'66 Cals. ; J . Dewar gives 14'43 CaIs.; R. Naumann, 14.96 Cals.; W . Nernst, 13-94 Cals. at 101" ; G. P. Baxter, 15.1 Cals. between 7.5" and 52.5". The heat of fusion of iodine a t 114' is estimated by G. N. Lewis and M . Randall to be 7.27 Cals., and the increase of free energy accompanying the change is 2000-5'17T, so that a t 295" K., t he increase in the free energy of iodine in passing from the solid to the Iiquid state is 4.60 Cals., and in passing from the solid to the gaseous state, a t 295" I<.,4.64 Cals. G . P. B axter, C. H. Hickey, and W. C. Holmes found for the increase in free energy in passing from the solid to t'he gaseous state a', T o, 16900+6*7T log T-0.01~20T2-78.73T; and G. N. Lewis and M. Randall, 26275+la6T log T-40a36T. l3.. Arctowsky found the speed of sublimation t o decrease with increase of press., being twelve times as great a t about 16 mm. press. a s i t is a t 760 mm. The specific heat and thermal conductivity.-A. Campetti 33 found the t h e r d THE HALOGENS 66 conductivity of chlorine gas is about 0.8 times that of air, and it is not altered by the insolation of the gas. The data for the sp. ht. of the halogens are somewhat meagre in comparison with some of the other gases ; t hey suffice, however, t o show that th0 mol. ht. are higher than is usually the case with diatomic gases; and that #b ratios of the two sp. ht. are lower. H. V. Regnault's values for the two sp. ht. of gaseous chlorine are Cp=0*1241 and C,=0.096 between 13" a nd 202", hence C,/C,=1.291. K. Strecker'svalues for Cp and C,, are rather smaller, and T. Martini's 1 values smaller still-he gives CP=Oml 1 ; C,=0*083 ; Cp/C,=1.336. The mol. ht. o chlorine, MCp=8.80 and MC,=6.81, are higher than is usual for diatomic gaaes, f M. Pier suggests that t h h for which Cp=6.885 ; MC,=4.900 ; MC,/MC,=l-41. is due to the dissociation of imaginary C14 molecules. The phenomenon has been discussed in connection with the sp. ht. of gases. M. Trautz found that the wave length in Kundt's tube alters when the gas is exposed t o light ; J. W. Mellor could detect no difference. M. Pier measured the mol. ht. of chlorine, MC,, a t temp. up to 1794", and found that up t o 1400" the results could be represented by MC,,=5"704 0.0005 cals., when the corresponding value for normal diatomic gases is MC,=4.900 0.00045 cals. The mol. ht. of chlorine rises from 6.194 a t 1288" to 6'317 a t 1365", to 6'677 a t 1490°, to 7'600 a t 1667", and to 8250 a t 1894". I t i s supposed that the dissociation of chlorine molecules, Clz, into atoms, 2C1, explains the large consumption of heat above 1450". M. Trautz found that MC, 'for chlorine a t one atm. press. between 25" and 100" is 5.22 ; between 25" and 150°, 5.35 ; a nd between 25" and 200°, 5.47 ; and he obtained a smaller value for C, when the gas is exposed to the light of a quartz lamp. On the contrary, A. Campetti found that illumination made no difference to any of the physical properties he measured. H. V. Regnault's two values for bromine gas are Cp=0.0555 and 0.05518 between 80" and 230" ; a nd for the mol. ht., MCp=8.80; MC,=6'80. K. Strecker found 1.292 for the ratio CP/C, between 20" and 388", where the result is not appreciably affected by temp. changes. Hence K. Strecker gives the values C,=0'0553 and C,=Om0428. M. Bodenstein and A. Geiger used the expression MCp=6.5+09O64T for the mol. ht., MCp of bromine gas a t the absolute temp. T, but there is a considerable amount of uncertainty about the accuracy of this expression. T. Estreicher and M. Staniewsky found the sp. ht. of bromine between -192" and -80" to be 00727 Cal. ; a nd the at. ht., 5-61. K. Strecker's value for the sp. ht. of iodine vapour between 250" and 377" is either C,=0.0349 or 0'0336, according as the vapour density of iodine be taken as 8.716 or 8,758. The value for C, is 0.0257, and the ratio of the two sp. ht., Cp/Cv=l.307. The mol. ht. MCp is accordingly 8.7 between the indicated temp. T. E~t~reicher M. Staniewsky found the sp. ht. and at. ht. of and iodinebetween -191" and -80" t o be respectively 0'0454 Cal. and 5-76 ; a nd between -90" and 17", respectively 0'04852 Cal. and 6.16. All the halogens have larger mol. ht. than the usual values for diatomic gases, and this the more the greater the at. wt. of the element. G. Starck and M. Bodenstein represent the mol. ht. a t the absolute temp. T, by the empirical formula MCp=6.5 +0'0038T ; a nd G . N. Lewis and M. Randall provisionally propose 6.5+0.004T for all three halogens. The sp. ht. of liquid chlorine between 0" and 24" is, according to R. Knietsch,s 0'2262 ; a nd between -80" and 15", according t o T.Estreicher and M. Staniewsky, 0.2230. H. V. Regnault's value for liquid bromine between 13.21" and 58-36" is 0'11294 cal. ; between 11.57" and 48.35", 0.11094 cal. ; a nd between -6.23" and 10m4", 0'10513. He also found that the value of this constant decreases as the temp. of the determination is lowered, so that between -7.3" and TO", the value is 0.106 cal., andbetween 6" a nd 14", 0'108 cal. T. Andrews also obtained the value 0.0171 cal. for liquid bromine between 11" a nd 45". According to R. Abegg and F. Halla, the sp. ht. of liquid iodine between 114" and 185", 0'0630, and the at. ht., 8'01. The sp. ht. of solid chlorine between -192" and 108", according to T. Estreicher end M. Staniewsky, is 0.1446, which makes the at. ht. of solid chlorine 5.13. Fot + + 86 INORGANIC AND THEORETICAL CHEMISTRY solid bromine, B. V. Regnault obtained 0.087 cal. between -77.75' and 0'19"; 0'082 cal. between 7-75' and -22L330-say, as a mean 0'084. This makes the at. ht. of solid bromine 6.71. H. BarschaIl found the sp. ht. of solid bromine between -77" and -183" to be 0.073 ; a nd F. Koref, 0.075 between -81'1" and -190.8". From a determination of the sp. ht. of antimony tribrornide L. von Pebal and H. J ahn obtained 6.52 for the at. ht. of bromine between 0" and 33" ; a nd 5.67 between -21" and -80". The sp. ht. of solid iodine between 9" and 98" is 0.05412 according to H. V. RegnauIt, and W. Nernst found the at. ht. a t -244'7", 3-78; -239.5", 3.97 ; -226.5", 4-17 ; --196", 5.38 ; -87", 5'92 ; -38", 6'36 ; a nd a t 25", 6-64. J. Dewar reported that a t --223", or 50" K. S pecificheat Atomic, heat . . 8 . . Chlorine. 0 .0967 3 .43 Bromine. 0.0453 3.62 Iodine. 0.0361 4.59 According to G. N. Lewis and-G. E. Gibson,35 the entropy of liquid bromine at 25" is 18.5 per mol., where the increase of entropy in passing from absolute zero to the m.p. T is +=JCp& log T=12.7 ; t he increase of entropy in passing from the solid to the liquid state a t To K. is 1290/266=4.85 ; a nd the increase in passing from the liquid state a t 266" K. t o that a t 298" K. is0.95. Similarly, the entropy of chlorine gas a t 25" is 27.8 per mol., where the increase in passing from absolute zero to the m.p. T of the solid is +=JC,d log T=9'1; the increase in passing from the solid to the liquid is 817/171=4V ; i n passing from the m.p. to the b.p., 2-63 ; i n passing from the liquid to the gas, 10.43 ; a nd in passing from the b.p. to 298" K., 0.83. The entropy of iodine a t 25" is 15'1 per mol., where the increase of entropy in pasaing from absohte zero to 298" K. is +=JCpd log T d 5 . 1 . R . C. Tolman's computations yield more than double these values. G. N. Lewis and M. R andall give for the free enern of formation a t 25" 6f solid I, 0 cals., and of soIid Br, 157 cals. ; of liquid Br, 0 cals. ; of liquid I, 460 caIs. ; of gaseous Br2, 755 cals. ; of gaseous J2, 460 cals. ; of gaseous Br, 22,328 3 cals. ; of gaseous I, 16,965 cals. ; of aq. Br2, 977 cals. ; of aq. 12, 926 cals. ; of a carbon tetrachloride soln. of Br2, 389 cals. E. Briner estimates the heat of formation 2CI=C12 to be 1130 Cals. a t 1670°, when the equilibrium constant is 0.01 ; f or bromine 2Br=Br2+57*0 Cals. a t 1050"-equilibrium constant 0.06 ; a nd for iodine 21=12+32-4 Cals. a t 1390"-equilibrium constant 0'66. The ratio of the kinetic energy of transIatory motion to the total energy of motion36 for molecules of gaseous oxygen, nitrogen, hydrogen, and the like is 0.607 ; for the three halogens this ratio is much smaller, being 0-48 for chlorine, 0.44 for bromine, and 0.46 for iodine. The refraction coefficients.--The index of refraction of chlorine 37 gas 1.000772 for white light was determined by P. L. Dulong in 1826, and confirmed by M. Croullebois in 1870. The indices for the C , D ,E, a nd G lines are 1.000699 (C), 1°000773(D), 1.000792,(E),a nd 1.000840 (G) ; a nd according to C. a nd M. Cuthbertson for light of wave length (pp) . W ave length Index of refraction 520'9 576'9 643.8 670.8~~ .480.0 1.00078135 1.0007703 1.00077563 l .00079166 1.00078651 J. H. Gladstone estimated the refraetion eq. of chIorine in its compounds to be 10'05 ; a nd later determinations by E. Conrady, 5. W. Briihl, and P. Eisenlohr, give for the H- a nd the Na-lines Ha, 5.933 ; D, 5.961 ; Hg, 6,043 ; H,, 6.101. The refraction eq. of the chlorine atom in the acid ch1orid.e~ rather higher, viz. is -6.3 to 6 -1. M. Croullebois' value for the dispersion of chlorine (pG-pC)/(pE-1) ==OS1780 t he atomic dispersion H,-Ha =OSl68. According to L, Bleekrode, ; the refractive index of liquid chlorine for the D-line is 1.367 a t 14"; or 1'385, according to J. Dechant, with a variation of 0.00098 per 1". This gives for the specific refraction of the liquid by the p-formula 0-27, and by the @-formula, 0.169 ; a nd for the gas respectively 0.24 and 0.16. 57 T HE HALOGENS The index of refraction of bromine gas a t O0 a nd 760 mm. for the D-line is 1.001132 according to E. Mascart. The atomic refraction of liquid bromine, according to J. H. Gladstone, is 15-3; a nd, according t o J. W. Briihl, 8-455 by the p2-formula. The specific refraction, according to A. Haagen, is 0.1918. The refractive indices of liquid bromine, selected from measurements by C. Rivihre, for rays of different wave lengths, a t different temp. as indicated in Table I. These data show that the Index of refraction. Wave length. 1. 5 1" 0 -- . . . 790.9 758.6 (A-line) 701.7 830.8 (Li-line) 631.5 692.5 639.0 (D,-line) 1 - P' O I 1 .6368 1 - 6394 1.6453 1'6495 1.6557 . . . - - refractive index of the liquid is greater the greater the wave length, and the lower the temp. At 20°, the dispersion between the A - a nd D-lines is 0.037, which is greater than the corresponding value for carbon disuIphide. The refractive index of iodine vapour for the red and violet lines from a cadmium electrode is 1.00205 for the red, 1.00192 for the violet a t 10" ; C . a nd M. Cuthbertson give for light of different wave lengths (pp), Wave length . Index of refraction . . 6 70.8 1 .001970 621.5 1.002 130 560-0 1.002 170 5 1 0.0 1.002210 600~~ 1.002120 According to P. P. le Roux, like all vapours with a large selective absorption, iodine has an anomalous dispersion since it increases with a fall of temp., being about 0.06 from A. Hurion's measurements-approximately as large a negative number as glass is positive. The atomic refraction of solid iodine is 24.5 by the p-formula, and 14.12 by the p2-formula. The refractivities of the four halogens<.e. the refractive index less unity multiplied by 106-are P, 195 ; Cl, 768 ; B r, 1125 ; I , 1920 (violet) and 2050 (red). According to C . Cuthbertson and E. B. R . Prideaux, if referred to fluorine unity, these constants are nearly in the ratio 1 : 4 :6 : 10. A similar ratio occurs with aeon, argon, krypton, and xenon. What C. a nd M. Cuthbertson call the dispersion, (p-l)lOG, for light of different wave lengths, is 775'63 for chlorine gas for light of wave length 670.8pp; 784'00 for 546.1pp ; a nd 791.66 for 48O'Opp ; or p-1='is313 x 1027/(9629.4x 102-n2). For bromine gas, the dispersion is 1152.5 for 670-8pp ; 1174'1 for 575'Opp ; a nd 1184.9 for 5 4 6 . 1 or ~ ~ p-1=4*2838 x 1027/(3919-2x lO27-n2) ; a nd for iodine gas, 1970 for 670b8pp 2130 for 618.Opp ; a nd 2120 for 500.0pp. ; The spectra of t he halogens.-In 1865, D. Forbes 38 showed that chlorine colours a Aame of a Bunsen burner or of a spirit lamp, green ; so do chlorine compounds after they have been treated with sulphuric acid. The line spectrum of the halogens obtained by the electric spark has been measured by J. Pliicker and W. H ittorf, G . Salet, A . J. Angstrom, and J. M. E der and E. Valenta. Most of the lines in the spark spectrum of chlorine fall between 27'6 and 6'i5.8pp, but the majority are a t the violet end of the spectrum. The spectrum of chlorine has been more particularly studied by J. M. E der and E. Valenta, who recorded about 400 lines, most of which were in the ultraviolet, although some extended into the blue, green, yellow, and red. Those in the violet and ultraviolet are sharper than thoae in the green or yellow, 58 INORGANIC AND THEORETICAL CHEMISTRY which latter are, for the most part, broad or indistinct. The line spectrum of chlorine can be observed by the discharge in a vacuum tube containing chlorine ah 50 to 100 mm. press. ; by sparking the gas a t ordinary press. ; a nd from fused chlorides, or minerals containing chlorine. The most pronounced lines of chlorine are the four in the yellowish-green ; a bright green line, and a group of lines-three of which are very bright-in the blue. The bromine spectrum is still richer in lines than that of chlorine-the brightest are a group in the blue and one group in the green. The spectrum of iodine in turn is richer in Iines than that of bromine-the brightest are a group in the yellow, and a group in the green ; t here are also many blue lines. If t he intensity of a spectral line be represented by its vertical distance from a datum line, the chief spectral lines of the three haIogens can be represented as in Fig. 17. E ach of the halogens gives two emission spectra-one with the continuous discharge, and the other with the oscillatory discharge. With iodine, if any of the solid be present in the tube during the oscillatory discharge, the vap. press. is so soon altered by the heat of the discharge ; a s a result, the discharge is damped and the nonoscillatory discharge appears. Hehce, with iodine, the oscillatory discharge can be obtained only for a few minutes. G. L. Ciamician specially studied the successive changes which variations of press. have upon the spectra of the halogens. He found that lines which are visible under one press, vanish a t another press., because, FIG. 17.-Chief Specihl Lines of Chlorine, Bromine, and Iodine. according to A. Schuster, there is " a mixture of several overlapping spectra." I n general, an increase of press. increases the intensity of the line spectrum, and causes new lines towards the red to become visible. Bromine and iodine also show line or spark spectra under the same conditions as the chlorine spectrum when the electric discharge in a vacuum tube passes through the vapours of these elements. J. At. Eder and E. VaIenta showed that bromine vapour a t a low press.-8 to 10 mm.-has a distinct and characteristic line spectrum. I t he press. is lowered, the spectrum f becomes faint, and the lines are broadened. Besides the line spectrum there is a continuous spectrum in the violet a t low press. and still a third spectrum a t a press. of 45 mm., which seems to correspond with the normal band spectrum of other elements. The flame of hydrogen containing bromine gives a continuous spectrum, so does bromine vapour heated a t low redness in a glass tube. The sat. vapour of iodine in a layer 0.1 metre thick, is opaque to daylight or to candle-light ; t he vapour appears a t the edges to be blue by transmitted light, black by reflected light ; a nd, according to C..!I Schonbein, it appears to be the blacker the higher the temp., owing to an increasing absorptive power for light. C. 6. Sellak says that a thin layer of solid or molten iodin'e transmits only the rays in the extreme red. The fine purple colour of iodine vapour is due to its transmitting freely the blue a'nd red rays of the spectrum, while it absorbs nearly all the green; but if the iodine vapour is in thick layers it absorbs the red rays, and the trans. T HE HALOGENS rnitted light is purely blue. A soln. of iodine in carbon disulphide exhibits the same phenomena since it appears blue or purple according to its density. The red alcoholic soln. does not show this phenomenon. The spark spectrum of a trace of iodine vapour in a vacuum tube shows bright characteristic lines. This spectrum does not correspond with the absorption spectrum; but, according to G. Salet, if a low-tension current is passed through a vacuum tube containing iodine, the spectrum shows a set of bands identical in position with the absorption spectrum of R. T. Thalen. The absorption or band spectrum of chlorine was examined by W. A. Miller in 1845 and E. Robiquet in 1859, but they failed to detect a line absorption spectrum with chlorine, although one had been previously noted by W. A. Miller in 1833 with bromine and with iodine. A. Morren (1869) and D. Gernez (1872) obtained the desired line spectrum by using a long tube-say, 2 metres in length-filled with the gas. G . D. Liveing and J. Dewar found that a smalI quantity of chlorine gave a wide absorption band stretching in the ultraviolet from 356 t o 302pp, which widened as the amount of chlorine was increased until they obtained a band stretching from 465 to 263pp (Fig. 18). J. Tyndall found that with the exception of air, nitrogen, and hydrogen, chlorine gas absorbed the long heat rays less than any gas he tried, while K. Angstrom and ITT. Palmaer found a single band in the infra-red spectrum stretching from 323 t o 607pp with a maximum a t 428pp ; E. R. Laird measured the complete absorption spectrum of chlorine a t ordinary temp., and found a very broad total absorption band in the violet region, a line absorption in the blue, green, and yelIow, particularly rich in the region between 545 and 480pp, and weakening a t both ends. The lines do not coincide with the known lines in the emission or spark spectrum of chlorine, although some lines are nearly coincident. With an increase of press., the absorption band in the violet region broadens out rapidly on the less refrangible side and more slowly on the more refrangible side ; a decrease of press. does not break the absorption band into lines. W. W. Coblentz has measured the ultra-red spectrum of thin layers of iodine, and found the vapour to be transparent for a wave-length 2 . 7 4 ~ ; and with thicker layers the absorption between 1 . 2 and 2"ip is constant. W. Burmeister found no infra-red absorption ~ bands in the absorption spectra of chlorine and bromine. According to R. W. Wood (1896), when iodine (or bromine) vapour is mixed with the vapours of carbon d i d phide, a portion of the iodine (or bromine) exists in a state of soln., and gives an absorption spectrum devoid of lines or bands, while another portion exists in the state o a gas, and gives a fine-line absorption spectrum. With a given density of f the vapour ot the solvent, a portion of the halogen can be vaporized without its s h o ~ n g gas absorption spectrum, but if a little more halogen be vaporized, the the fine lines of the gas absorption spectrum of iodine appear. The absorption spectrum of Iiquid chlorine is quite different from that of the gas. C. Grlirige 'found a n absor$tioli in the extreme red down t o about 697 or 686pp, and from there to about 512pp red, orange, yellow, and green light is transmitted ; absorption begins a t 512pp, and is complete in the blue and violet a t 503pp. Bromine and iodine vapours like chlorine show a characteristic absorption spectra with many lines. With decreasing atomic weight, an increasing amount of gas must be used to render the absorption lines visible and distinct. Thus, B. Rasselberg required a coIumn of iodine 10 cm. thick, bromine 75 cm., and chlorine 137 cm. to show the absorption lines. The absorption lines shift towards the red with increasing at. w t . This is usually characteristic of the behaviour of the emission spectra of a family group of elements. B. HasseIberg also measured about 3000 lines in the absorptionspectrum of iodine, 2500 in the bromine spectrum, and about 1000 in the chlorine spectrum. The number and sharpness of the absorption lines of-the halogens thus increase with increasing at. wt. With a higher dispersion, E . R . Laird has shown that the bands of the absorption spectra of iodine and bromine recorded by the early observers are composed of a number of lines. These bands appear as channellings in the spectra and make the spectra appear somewhat similar. Iodine 60 INORGANIC AND THEORETICAL CHEMISTRY shows these channellings most distinctly, and bromine and chlorine with diminishing distinctness. The absorption spectrum of iodine and bromine vapours disappears when the dissociation is high, presumably because the monatomic molecules give no absorption in the visible spectrum ;39 t he observed absorption spectrum is due to diatomic molecules. The temp. a t which the absorption spectrum disappears is higher with bromine than with iodine, and it is augmented by press. R. W. Wood estimated that there are between 40,000 and 50,000 lines in the absorption spectrum of iodine. G. D. Liveing and J. Dewar found that bromine vapour gives an absorption band FIG. 18.-Ultraviolet Absorption Spectra of the Halogen Gases. in the nltraviolet, which begins in the visible spectrum and extends to the solar L-line, when small quantities of bromine are present, Fig. 18, and to the solar P-line when more bromine is present. From this point to the line 250pp, the vapour is transparent, and, after that, the absorption increases with the refrangibility of the rays. C. Ribaud has aIso studied the ultraviolet spectrum of bromine up to 630'. Thin layers of iodine vapour are transparent for the ultravioIet rays, but there'is a strong absorption in the violet region of the visible spectrum ; with thicker layers of bromine the absorption extends nearly to the solar H-line, Pig. 1 9, b ut the vapour is still transparent for rays more refrangible than the H-line. The absorption band FIG. 19.-Ultravoilet Abeorption Spectra d l&pid &oPnine asld Iodine. in the ultraviolet spectra of chlorine, bromine, and iodine gases is thus shifted towards the less refrangible or red-end of the spectrum as the at. wt. of the element increases, until with iodine, the absorption band appears in the visible spectrum. A film of liquid bromine between two quartz plates has an absorption band, Pig. 19, which ends just where the transparency of the vapour begins, while the film is opaque for rays above and below this band. With soh. of iodine in carbon disulphide, the spectrum is also transparent for a certain distance, Pig. 18, b ut is shifted t o a Iess refrangible region lying between tile solar G- a nd H-lines. THE HALOGENS M. de Broglie, working with the zinc compounds of the elements, found the highfrequency spectra of iodine and tellurium to follow one another in accord with their chemical properties ; I . Malmer, working with the elements, found that order to be reversed. M. Siegbahn40showedthat M. d e Broglie and I. Malmer used the secondary radiations ; and with primary radiations he found the order to be that given by I. Malmer, ciz. tellurium-iodine. The wave-lengths for the al- and &-lines agree well with the series : At. numbers N . . Cd 48 1.364 1-460 In 49 1.406 1-485 Sn 50 1-435 1.521 Sb 61 1-461 1.551 Te 52 1.480 1 .573 I 63 1513 1.606 , . .. .. B 5a 6 ... ... 1.605 1,707 R . W. Wood 4 1 has shown that if a glass bulb with a few flakes of iodine be exhausted and sealed, a yellowish-green fluorescence appears if a beam of sunlight or arc-light is focussed on the centre of the bulb ; if the bulb contains air a t atm. press., no fluorescence occurs. Only when the press. is reduced to about 150 mm. does a feeble fluorescence appear. The intensity of the fluorescence gradually increases as the press. is further reduced; the most rapid change occurs when the press. changes from 10 cm. t o t hat of a high vacuum, With hydrogen the fluorescence appears when the press. is about 300 mm. higher than with air. If t he bulb is warmed the flnorescence appears a t higher press. R. W. Wood explains this by assuming that air at 15 cm. press. is able to dissolve all the iodine which vaporizes a t ordinary temp., but by a rise of temp. and consequent increased vapour press. of iodine, some iodine remains undissolved by the air, and it is this portion only which fluoresces. The fluorescent spectrum has a number of bands extending from the orange-red into the greenish-blue. R. W. Wood also found that when iodine vapour is heated in a sealed quartz bulb to about 700°, a luminous red cloud is formed-which in thin layers gives a banded spectrum resembling the fluorescent spect,rum but displaced a little more towards the red. Heated iodine vapour is probably luminous, as is shown by G. Salet's experiment : The room was made dark and when a hot gIass tube had cooled until i t was just barely visible, a fragment of iodine was thrown into the tube, which thereupon filled with luminous vapours. To obtain more brilliancy one heats the vapour of iodine i n a B ohemian glass tube by means of a n enameller's h p . The contents of the tube look like a r ed-hot bar of iron. One may also volatilize iodine around a platinum spiral brought to a v ivid inca.ndescence; the luminous vapour rises like a real flame about the spiral. I t is a case of &me without c o d u s t i o n . T he light fram the iodine gives a continuous spectrum, or rather a confused primary spectrum ; one perceives traces of characteristie channellings but no lines of the secondary spectrum. J. Evershed, and A. Smithells confirmed and extended these experiments ; t he former added : To sum up, then, i t appears that besides iodine, the vapours of bromine, chlorine, sulphur, selenium, and arsenic can all be made more or less incandescent by heating to the temp. at which the glass combustion tube softens, and the light Bmitted by each of these glowing vapours appeam to give a perfectly continuous spectrum ; while the corresponding absorption spectra are selective. Thus there is no such close relation between emission and absorption as is imphed by Kirchhoff's law of radiating bodies. There seems, however, to be a general relation between the total absorbing and radiating power for the visible rays ; those vapours which are highly coloured and absorb strongly i n t he visible spectrum also radiate conspicuously i n that art of the spectrum ; while colourless, non-absorbing vapours, such as phosphorus, emit no perceptible light when heated. H. Kijnen showed that the thermo-luminescence of iodine vapour begins at about 550°, and is stronger the denser the vapour, and he adds that the glow spectrum of iodine is specially interesting because it is one of the few cases where a visible spectrum ie obtained by merely heating a gas. According to W . Friederichs, the banded 62 INORGANIC AND T HEORETICAL CHEMISTRY absorption spectrum of iodine increases with temp. up to 500°, but decreases a t higher temp. until finally i t disappears. At 1250°, a continuous emission spectrum was obtained by C. Fredenhagen. Hence it is concluded that the band absorption spectrum is an effect of the diatomic molecules, and becomes weaker as the oneatom molecules begin t o form. When a substance is exposed t o a beam of radiant energy-say solar light-radiations of a certain wave-length are absorbed, and the energy is expended in inaugurating or stimulating intra-molecular or intra-atomic vibrations. The general effect is to raise the temp. of the body. Waves with a short period of vibration are absorbed, and emitted as heat radiations with a longer period of vibration. There is thus a degradation of radiant energy from waves of a short period t o waves of a long period. I n the case of fluorescent substances, light waves of short peri0d-e.g. those a t the violet end of the spectrum-are absorbed and emitted again as light waves of lower refrangibility. Violet or ultraviolet radiations may be absorbed and emitted again as green or red rays. This phenomenon is termed Jluorescence when the photo-luminescence is transient and shows only while the body is actualIy exposed t o the light stimulus--c.g. quinine scheelite, fluorspar, uranium glass, barium platinophosphorescence when the photo-luminescence persists after cyanide, etc.-and the stimulant light has ceased t o act on the body-e.g. Bologna-stone, Balmain's luminous paint, Canton's phosphorus, and other sulphides of the alkaline earths, some diamonds, etc. G. G. Stokes 42 illustrated the phenomenon by the following simile : Suppose you had a number of ships at rest on an ocean perfectly calm. Supposing now a series of waves, without any wind, were propagated from a storm at a distance along the ocean ; thoy wodd agitate the ships, which wodd move backwards and forwards ; but the time of swing of the ship would depend on its natural oscillation, and would not necessarily synchronize with the periodic time of the waves which agitated the ship in the first instance. The ship being thus thrown into a state of agitation would produce waves, which would be propagated from it in all directions. This I conceive to be a rough dynamical illustration of what takes place in t his actual phenomenon, namely, that the incidence of ethereal waves causes a certain agitation in the ultimate molecules (or atoms) of the body, and causes them to be in turn centres of agitation to the ether. I n fine, when the ultimate particles of a fluorescent substance are agitated by ether waves from, say, a source of light, they send out fresh waves of their own. The emitted radiations are of shorter wave-length than the absorbed radiations. The rate of transformation of the radiations from a high to a lower refrangibility is rapid with fluorescent substances, and much slower with phosphorescent substances. The absorbed energy may be dispersed i n o ther ways than i n developing fluorescent and phosphorescent effects-e.g. it may manifest itself in chemical action-which is utilized in photography ; a nd i t may be expended in augmenting the translatory motions of the molecules and be dissipated in the form of heat. I n the converse phenomenon, t.hermo-luminescence o r calorescence, the radiant energy supplied to the body as heat is so transformed that the body becomes luminous in the dark-e.g. some of the green varieties of fluorspar, scheelite, etc. G. G. Stokes' rule that $he exciting light is o shorter wave-length and greater f frequency than that o the excited l@ht is true only in certain cases. I n the case o f f bodies which do not obey Stokes' law, fluorescence is induced only when the incident light coincides in frequency with one of the sharp absorption bands of the substance, as R . W. Wood showed to be the case with sodium and iodine. P. Lenard's 4s hypothesis assumes that the incident light causes the liberation of electrons from the atoms of the phosphorescent substance, and that light is subsequentIy emitted when the ejected electrons return to the atoms. This return occupies an appreciable time with substances whichlave a small electrical conductivity, and, ex hypothesi, oppose a great resistance t o the motion of the electrons. The return of the ejected electrons is facilitated by a rise of temp. which increases the conductivity. The T H E HALOGENS 63 theory does not account for the extraordinary influellee of extremely small changes in the composition of the phosphorescent substance. S. Landau and E . Stenz examined the effect of low temp. and dissociation on the fluorescence of iodine vapour a t low press. Fluorescence decreases as t he temp. is raised, but does not cease a t 8W0. Dissociation destroys both fluorescence and the resonance spectra. It is therefore inferred that the complex vibrating system is not inherent in the atom, but in the molecule ; t hat the structure of the atom is relatively simple ; a nd that, in all probability, t h e absorption lines which are so characteristic of diatomic iodine and so sensitive to -the action of monochromatic light, do not belong to the absorption spectrum of monatomic iodine. W. Steubing found that the intensity of the fluorescence of iodine vapour is weakened between the poles of a powerful electromagnet. The result has nothing to do with the Zeeman effect, and has no connection with effects produced by admixture with gases, solvents, etc. It is produced by a d irect action of the magnetic field on the electrons causing the band spectrum weakening the individual vibrations. I n 1871, E . Budde 4 4 showed that when chlorine is exposed t o light rays of high refrangibility, an expansion--called the Budde effect-occurs and the temp. rises about lo. T his is not a direct heating effect of sunlight since the interposition of a screen between the source of light and the chlorine t o cut off the heat rays makes no difference to the effect. A. Richardson showed that the photo-expansion is proportional to the intensity of the more refrangible rays of light. E. B udde suggested that the expansion may be due t o the light loosening or actually decomposing some of the chlorine molecules into free atoms, C12 (2 vols.)+2C1(4 vols.), and that the slight rise of temp. is developed by the recombination of the gelochrten und zersetxen Chlmmolekule. He also showed that the expansion cannot be due to the direct warming effect of sunlight such as occurs, for example, when lampblack is exposed t o the red-heat rays, because the red-heat rays are least active in producing the effect. J. W. Mellor showed that the photo-expansion is directly proportional t o the rise of temp., and that it is unnecessary t o assume that the chlorine molecule is dissociated into atoms when exposed t o the actinic rays. The phenomenon is thus analogous with photo-luminescence. I n both cases, intramolecular vibrations stimulated by light are degraded into less refrangible vibrations which make themselves evident in the one case by a rise of temp., and in the other case by luminescence. According t o M. T rautz, the expansion in light is not the same a s when an equivalent amount of heat is applied to the non-illuminated gas. As indicated above, A. Campetti found Bhat ult.ra-violet light does not appreciably modifiv the viscosity and thermal conductivity of chlorine gas ; a nd, contrary to M. Trautz's opinion, this light does not aflect the sp. ht. a t constant vol. According to H. B. Baker and W. A. Shenstone, there is no Budde effect if the chlorine is dry, but it is produced by dry chlorine if platinum be present. I n this case the chlorine attacks the platinum in light not in darkness ; a nd in the case of moist chlorine, According to P. Caldwell, there is a chemical reaction, 2Cl2+2Hz0+4HC1+OZ. the effect does not occur with bromine-if so, i t may be that the active rays must be t h concerned in that portion of the absorption band of chlorine not covered ~ by t hat of bromine-Fig. 18. Electrical properties.-J. J. Thomson (1887) 45 reported that when electric sparks m e passed through iodine vapour between 200' and 230°, there was a considerable increase of press. which persisted for some hours. J. J. Thomson attributed this phenomenon t o what he estimated as a 47 per cent. dissociation: I2=I.+If. J. J. Thomson also studied the phenomenon with chlorine. According to E.P. Perman (18911, no perceptible change of density is produced by the discharge ; and according t o W. K ropp (1915), there is no evidence of such a phenomenon in quartz. J. J. Thomson could not detect the presence of free ions either in the ' pmliminary stage of the insolation of a mixture of hydrogen and chlorine, or in the later s t q e when the gases are aotively combining. The method employed was ' 64 INORGANIC AND THEORETICAL CHEMISTRY sensitive t o 1 i n 1014 of the molecules present. M. le Blanc and M. Vollmer could detect no signs of ionization when the mixture was illuminated by an osram lamp. The fog observed by P. V. Bevan t o be produced when a combining mixture of hydrogen and chlorine is expanded is not necessarily due t o ionization, for P. L enard has shown t h a t ionization, the formation of fogs, and chemical action are independent phenomena. G. Kiimmell's report that a m easurable ionization is produced by the insolation of chlorine, is thought t o be based on a m al-observation due t o defective insulation. E. R adel found that when moist chlorine is exposed to light from electric sparks repeated every hundredth of a second, a f aint cloud can be detected by ultra-microscopic methods. Radiations from polonium or radium bromide act similarly, but the effect is weaker. J. S. Townsend 46 showed that the chlorine gas liberated by the action of h ydrochloric acid on manganese dioxide has a strong positive electrification. The spark potential, V, t hat is, the lowest difference of potential between two electrodes required for sparking, is not constant for chlorine-air, nor for bromine-air.47 I n each case, there is a g radual increase with an increasing sparking distance and an increasing gas press., ultimately approaching a l imiting value. The converse phenomenon obtains with helium-air. There is an intense green coloration about the positive electrode with chlorine gas which with lower press. becomes very pale, and. it t hen gives the characteristic spectrum. The falls of potential at t he anode are very high with the halogens, presumably because the measurements are obscured by the action of the halogen on the electrodes.48 A. L. H ughes and A. A. Dixon give 8'2 volts for the ionizing: potential of chlorine gas, and 10.0 volts for that of bromine, while the value for chlorine calculated from K. T. Compton's formula V=O-194(K-l)-B volt, is 4-94 volts, where K d enotes the specific induction capacity, and V t he ionizing potential representing the least energy required t o ionize the gas by the impact of electrons. The calculated value for fluorine is 9'84 volts. C. G . F ound obtained for iodine vapour V =Bm5 volts. E . B . Ludham found chlorine is not ionized by ultraviolet light which is capable of ionizing air. The fluorescence produced in iodine vapour by light of comparatively long wave-lengths led t o experiments on the ionization of iodine vapour by exposure t o light. J. H enry, E. W hiddington, and J. P ranck and W. Westphal obtained negative results with light deprived of most of the ultraviolet b y passage through the glass of the apparatus. The latter, however, found that it is easier t o p roduce a glow discharge in fluorescing iodine vapour than through non-fluorescing vapour ; hence it is inferred that less work is required t o separate an electron from a m olecul? of the vapour when it i s fluorescing than when it is not. H. M. Vernon exposed chlorine in an ozone tube t o the action of a silent discharge, and obtained no other result than a slight expansion due t o the heating effect of the discharge. E . B riner and E. D urand also failed t o obtain a contraction--even z&5th volume under similar conditions. According to K . Kellner,49 when purified and dry bromine is exposed, in double-walled tubes like ozone tubes, to the alternating current of 250,000300,000 volts of a TesJa transformer, a s ulphur yellow crystalline deposit is formed on the walls of the tube, and if small quantities of bromine are used, the whole may be transformed into this product ; i t is stated that the glass takes no part in the change. I n the absence of some confirmatory evidence, it c annot be assumed that the bromine is here polymerized. Liquid chlorine is virtually a non-conductor of electricity ; according t o F. Lindej50 the conductivity is smaller than 10-16 reciprocal ohms ;. a nd according t o W. A. Plotnikoff, t h a t of liquid bromine is less than 10-8 rec. ohms ; a nd G. N. Lewis and P. Wheeler place the specific conductivity of iodine at a bout 3 x 10-5. T hey also find the conductivity of s o h . of potassium iodide in liquid iodine between 120" and 160' is equal t o that of the best conducting aq. s o h . Solid iodine is a.v ery bad conductor, and like other insulators, it develops electricity byfriction.51 K. Pajans found the heat of hydration of gaseous ions t o be : Cl', -23 ; Br', -32 ; I', -43 kgrm. cals, per gram-ion. P. L enard, W. Weick, and H. F. Mayer calculate that in soh. : 65 T HE HALOGENS Number of mols. of H,O per i o n . R adiusof complex . . F 11 3.9~ Br C l' G 3.2~ 6 3.0~ I 9 2.9~ rm. R . Lorenz, and P. Walden estimate the diameter of ions of chlorine in ay. soln. cm., and of iodine ions, to be 2-30 X I O - ~ cm. at 18" ; of bromine ions, 2-24 x 2.26 X I O - ~; in methyl alcohol soh. at 25O, the numbers are respectively 4-98X ~ O - ~ 4-78 x , a nd 4.84 X 10-s cm. ; t he diameters of the atoms are respectively 2.33 x 2'52,X and 2.60 x cm. ; while the molecular a nd 3-76 x 10- cm. diameters of bromine and iodine are respectively 3'42 x From the observations of E. C. Sullivan, A. A. Jakowkin, F. Boericke and I?, Haber, M, de K. Thompson 5 2 computed the electrolytic potentials of the halogens in soln. sat. with them, at 25O, to be -1.643 volts for chlorine ; -1'353 voks for bromine; and-0.817 volt for iodine. The e.m.f. of thegas-cellPt ] H [ HCl j C I P t 1 in which a hydrogen and a chlorine electrode are immersed in aq. hydrochloric acid &odd be the same as that required to electrolyze hydrochloric ac,idof the same concentration if the process were strictly reversible ; b ut the observed results are rather less than the theoretical. The gas-cell has therefore been investigated by P. J. Smale, E. Miiller, W. von Beetz, B. 0. Pierce, and others. According to E. Muller, the chief sources of error are due to the hydrolysis of chlorine according to A. A . Jakowkin's equation, and to the possible formation of perch lo ride^. E. Muller found that when measured against the hydrogen electrode, the electrode potentials of chlorine against hydrochloric acid i n t he same soln. are Concentration of acid. N -HCI OelN-HCI O.OlN.HC1 0.001N-HC1 . . . . . . . . 33.m.f. corrected for hydrolysis. Observed e.m.f. 1.366 1.486 1.599 1 .733 1 .366 1 .485 1 .546 1 .587 Calculated e.m.f.f ~ o m that of N-HC1. - 1.477 1.594 1 .712 The correction for the increase of chloride ions due to the hydrolysis of the chlorine has largely eliminated the deviations between the observed and calculated values. G . N. Lewis and F. F. Rupert find for the electrode potential of chlorine against the normal electrode to be -1.0795. F. Dolezalek measured the difference in the e.m.f. of two 5N- to l2N-hydrochloric acid cells of different strengths by the vap. press. method, and obtained satisfactory results. F. Boericke, G. N . Lewis and H. Storch found for the normal electrode otentials against hydrogen a t 25' Hz, -1.0872 volts; Brlllquid HBr ] H2, -1.0661 volts ; and B H z, -1.0824 volts. B. 0. Pierce gives for H2 1 NaBraq. I Br2, B r2gas B 1.252 volts; for H2 I KBraq. I B r2, 1'253 volts ; f or 0, I KBr,,. 1 Br2, 0'500 volt; for Hz [ KI,,. I I z, 0.861 volt ; and for O2 1 K Iaq. I 12, 0.057 volt. P. D Pooto and F. 1,. Mohler estimate the electro-Elfltinityof chlorine to be 4.8 volts. F. Linde 53 found the dielectric constant of chlorine a t -60°, with a wave-length about 104 cm., is 2.15 ; a t -20°, 2.03 ; a t 0°, 1-97; a t l o0, 2'08 ; and from 0" to the critical temp. a decrease of about 0.0044 per degree. For liquid chlorine W. D. Coolidge obtained a dielectric constant of 1.88 (14") and F. Linde, 1-93(14"). For bromine at l ofor wave-length of about 104 em., P. Walden found 4'6 ; a nd a t 23' for a wave-length 84 cm., H. Schlundt found 3.18 ; with iodine, with a wavelength of 75 cm., W. Schmidt obtained a dielectric constant of 4.00. Magnetic properties,-Chlorine, bromine, and iodine are diamagnetic.54 The magnetic susceptibility of chlorine at atm. przas. and 16" is -0.50 x 10-6 per unit mass ; and a t 15O, 0'007 X lo--= per unit volurne ; for bromine, the magnetic susceptibility st 18' is -0-38 x 10-6 ; a t 20°, 0'41 x 10-6 per unit mass, and a t lgO, -1.4~10-6 per unit volume. Por liquid iodine a t 115O, the magnetic susceptibility is - 0.4~10-6 and 180°, -0'3 ~ 1 0 - 6 unit mass, and for iodine crystals a t l o0, per -0.35 x 10-6. The atomic magnetism of chlorine in organic derivatives is 282, 249,218, and 194, according as there are 1, 2, 3, or 4 chlorine at,omsper molecule ; II VOL. 11. f F INORGANIC AND THEORETICAL CHEMISTRY with bromine, 413, 374, or 334, according as there are 1,2, or 3 atoms per molecule ; a nd with iodine, 642 or 577, according as there are 1 or 2 atoms per molecule. The molecular rotation of the plane of polarktion 5 5 of sodium light in an electromagnetic field with reference t o water unity is for a 1 0-1per cent. soh. of chlorine in carbon tetrachloride 4.344 a t 7.6' ; a nd the atomic rotation of chlorine (hydrogen unity) is 1.675 ; a nd of bromine, 3.563. I n studying the influence of a magnetic field on the optical behaviour of chlorine gas, A . H eurung found no appreciable increase in the region of the absorption for light of wave-length 518 to 640pp. R. W. Wood found that if iodine be placed between the poles of a magnet and parallel rays of arc-light be sent through a nicol prism, then through a tube containing iodine vapour, and finally through a second nicol, an intense blaze of emerald-green light appears when the magnet is excited. R. W. Wood (1906) and A. Heurung (1911) have measured the spectrum of the green light--several new lines make their appearance-and since the magneto-optical effect cannot be eliminated by the rotation of the analyzer, it is inferred that only a small portion of the light is polarized. The effect on the individual spectrum lines of chlorine could not be detected. Summm.-The gradation in characters which the halogens show with increasing at. wt. from fluorine to iodine, yields one of the most typical family series of elements. The best representative values of some of the physical constants of the halogens are sumn~ariied Table 1 . The family relationship of the halogens in 1 is illustrated by :(1) The s imilarity in the chemical and physical properties of the elements and their corresponding compounds, is such that the properties of any one Fluorine. At. wt. . S tate aggregation (0') Colour Sp. gr. (liq.) SD. voI. Ai. vol. Mol. vol. . .. . B .p. .. M.p. C ritical temp. Coeff.expansion Coeff. compressibility H e a t of fusion H e a t of vaporization Sp. ht. (0') S p. ht. gas, C,. Sp. ht. gas, C" R atio y R efractive index At. dispersion At. refraction Magnet. susceptibility Dielectric constant . . . . Chlorine. 35.46 19.0 Gas Pale yellow le108(-187') 0.9025 17-15 34.30 -187' -233" - Gas Yellowish -green le568(-33.6") 0-6635 23-52 47.04 -33.6' c .102O 147' 0.001 87 (0') 0.00095 22.96 caIs. 67'38 caIs. 0.2662 ( liq.) 0.1 155 0.0873 1.323 1.000768 0-50 10.05 5.9 X 1 -97 Bromine. Iodine. 2 6-92 79-92 Solid Liquid Violet Brownish-red 4.004 (107') 2.9283 (59") 0.2698 0.3392 34.23 27-13 68.46 54-26 183' 50' -7.30 114' 512O 302' 0.000025 (liq. ) 0-00106 (0") 0-000013 0.000052 2-9 cals. 16.185 cals. 81 c d s . 45.6 cals. 0,0515 (solid) 0*1071 (liq.) 0'0336 0.0553 0-0428 0.0257 1.307 1.292 1.001920 1-001125 1.22 3-65 15.3 24.5 -4.1 x 10" - 3.5 x 4.00 3.18 member of the family can be said to summarize or rather to typify the properties of all the other members although fluorine diverges a little in some of its properties. (2) The g radual transition of chemical and physical properties such that i the f elements be arranged in, order : I?, C1, Br, I, t he variation in any particular property in passing from fluorine to iodine nearly always proceeds in the same order, and that is the order of their at. wt. Similar family characteristics will be found wit4 the chemical properties of the T HEHALOGENS 67 halogens. Taking almost any property and comparing its magnitude in passing from the element fluorine t o iodine, or from the fluorides t o the iodides, a similar gradation will be observed : T hus, take the m.p. of the cadmium, calcium, barium, sodium, or potassium salts : . . . Sodium Potassium Calcium Barium Cadmium I .Fluoride. 80" . 1985' 8 , '. 1361" . 1000° 280" Chloride. 820" 790" 780" 960' 590' Bromide. 765' 750" 760" 880' 350' Iodide. 650° 705' 740' 740" 1000" The markedly greater jump in passing from the fluorides t o chlorides than with the other steps in the case of the'earths is supposed t o be explained by a difference in constitution. There is also a difficulty with the cadmium halides. REFERENCES. J. H. Kastle, Amer. Chem. Journ., 23. 500, 1909. a M von Pettenkofer and K. B. Lehmann, Sitzber. Akad. Mun.ich, 179, 1887. . ' W. Wahl, Proc. Roy. Soc., 88. A, 348, 1913 ; P. C. E. M. Terwogt, Zeit. anorg. Chem., 47 203, 1905; B. J. K arsten, ib., 53. 365, 1907 ; V. Kurbatoff, i b., 56. 230, 1908; H. Arctowsky, ib.,lO. 25,1896 ; S. M. Jorgensen, Ber., 2.465, 1860 ; E. S. von Fedoroff, Bzdl. Acad. St. Petershurq, (j), 2. 287, 1907 ; E. .Mitscherlich, Pogg. Ann., 98. 554, 1856 ; G. von Rath, i b., 110. 116, 1860 ; 2 L. Playfair, Proc. Roy. 8oc. Edin., 4. 241, 1862 ; J. C. G. d e Marignac, Mkm. Sciences Genhve, 14.208, 1865; 0. Lehmann, Zeit. Kryst., 1.460, 490, 1877 ; W. L. T. Addison, ib., 32. 298, 1900. J. L. Gay Lussac and Ti. 5. ThBnard, Recherche? physico-chimiques, Paris, 2. 128, 1811 ; R. Bunsen, Liebig's Ann., 86. 265, 1853; A. Leduc, Compt. Rend., 116. 968, 1893; 125. 571, 1897 ; 140. 642, 1905 ; H.M ois~anand A. B. du Jassonneix, i b., 137, 1198, 1903 ; Ann. Chim. . Phys., (8), 1 145, 1904. E. Ludwig, Ber., I. 232, 1868 ; F. P. Treadwell and W. A. K. Christie, Zeit. anorg. Chem., 47. 446, 1905; J. M. Crafts, Compt. Rend., 107. 301, 1888 ; H. Jahn, Ber., 15. 1242, 1882; M. Pier, Zeit. phys. Chem., 62. 385, 1908. * R. ICnietsch, Liebig's Am., 259. 124, 1890 ; J. Dewar, Phil. Mag., (5), 18. 210, 1884 ; Chem. News, 51. 27, 1885 ; I. Traube, Zeit. phys. Chem., 08. 289, 1909 ; A. Nadejdine, BdE. Acad. 8 t Petersbwg, 12. 290, 1885. V. a nd C. Meyer, Ber., 12. 1426, 1879 ; F. Seelheim, ib., 1 2. 2066, 1879 ; V. Meyer, i b., 12. 2202, 1879 ; 21. 2018, 1888;, V. Meyer and H. Ziiblin, ib., 13.405, 1880 ; C. Laager and V. Meyer, Pyrochemische Untersttchungen., Braunschweig, 1885; Ber., 15. 2769, 1882 ; J. M. Crafts, i b., 16. 457, 1883 ; Compt. Rend., 90. 183, 1880 ; 98. 1259, I884 ; A. Lieben, ih., 89. 353, 1879 ; 1. Troost and P. IIautefeuille, ib., 84. 947, 1877 ; M. Reinganum, Phys. Zeit., 6. 514, 1905. 1 8 R . P. Perman, Proc. Roy. Soc., 4 . 45, 1890 ; E. P. P erman and G. A. S. Atkinson, i b., 66. 8 10, 1900 ; J. J. Thomson, ib., 42. 343, I887 ; H. J ahn, Mowztsh,., 3. 176, I882 ; Ber., 15. 1238, 1882 ; C. Langer and V. Meyer, i b., 15. 2769, 1882 ; V. Meyer and H. Ziiblin, ib., 13. 405, 1880 ; 0. Brill, Zeit. phys. Chem., 57. 711, 1907 ; E. Mitscherlich, Pogg. Ann., 29. 217, 1833 ; J. M. Crafts, C m p t . Rend., 90. 183, 1880 ; M. Bodenstein and F. Cramcr, Zeit. Ekktrochem., 22. 327, 1916. V. Meyer, Ber., 13. 394, 1880; J. M. Crafts and F. Meier, ib., 13. 851, 1318, 1880; Compt. Rend., 90. 690, 1880 ; L. Troost, ib., 91. 54, 1880 ; 95.'30, 1882 ; H . St. C. DevilIe and L. Troost, . Ann. Chim. P h p . , (2), 58. 257, 1860 ; A. Krause and V. Meyer, Zeit. phys. Chem., 6. 5, 1890 ; H.Biltn, ib., 2. 934, 1888 ; E. P. Perman, Proc. Roy. Soc., 4 45, 1890 ; A. Scott, ib., 23. 356, . 1838 ; J. Dewar and A. Scott, B. A. Rep., 957, 1881 ; Cr. S tark and M. Bodenstein, Zeit. Elektrodiem., 16. 961, 1910 ; M. Bodenstein and M. K atayama, ih., 15. 246, 1909 ; Zeit. phys. Chem., 69. 26, 1909 ; 0. Brill, ib., 57. 71 1, 1907 ; I . Langmuir, Journ. Amcr. Chem. Sac., 37. 1139, 1915 ; G. N. Lewis and M. R anddl, ib., 34. 1128, 1912 ; 36. 2250, 1914. l o R. Knietsch, Liebig's Ann., 259. 100, 1890 ; A. Lange, Zait. a?egew. Chem., 13. 683, 1900 ; L. Bleekrocle, Proc. Roy. Soc., 37. 339, 1884 ; J. Drugman and W. Ramssy, Journ. Chem. Soc., 77.1228, 1900 ; F. M. G. J ohnson and D. McIntoe11, Journ. Amer. Chem. Soc., 31. 1138, 1909 ; M. Pellaton, Journ. Chirn. Phys., 13. 426, 1015. A. J. Balard, Ann. Chim. Phys., (2),32,337,1826 ;J. J. Pierre, i b., ( 3), 20. 5,1817 ; C. Lawig, D m Brom ~ t n dseine chemiden Verhdtnisse, Heidelberg, 1829 ; Pogg. Ann. , 14. 485, I828 ; T. E. Thorpe, Journ. Chem. Soc., 37. 141, 1880 ; W. Ramsay and D. 0. Mauson, Chem. New., 42281,1880 ; Ber., 13. 2045, 1880 ;L. W. Andrews and H. A. Carlton, Journ. Amer. Chem. Soc., 29. 688, 1907 ; J. D. van der Plaats, Rec. Trav. Chim. Pays-Bas, 5. 34, 1886.; H. Billet, Mkm. PZnst$td, 292, 1855 ; J. Dowar, Chem. hews, 91. 216, 1905 ; G. le Bas, ib., 115. 146, 1917 ; A. Lacleaburg, Ber., 35. 1256, 2902 ; J. S. Stas, Mkm. Acad. Belgique, 35. 3, 1865 ; J. Drugman and W. Ramsay, Journ. Chem. Soc., 77..1228, I ROO. l Z T. W . Richards, Zeit. phys. Chem., 49. 1, 1904 ; J ourn. Amer. Chem. SOC., 1643, 1916 ; 37. I NORGANIC AND THEORETICAL CHEMISTRY T. W. Richards and W. N. Stull, New Method for determining compessibility, Washington, 1903 ; A. h n g e , Zcit. angew. Chem., 13. 682, 1900. l3 L. Grunmach, Ann. PIysik, (4), 4. 374, 1901 ; A. A. Trussewitsch, Zeit. phys. Chem., 0. 360, 1890 ; F. M , G. Johnson and D. McIntosh, Journ. Amer. Chem. Soc., 31. 1139, 1909. l4 W. R a m ~ a y and E. Aston, Journ. Chem. doc., 65. 167, 1894 ; P. Walden, Zeit. phys. Chem., 65. 126, 1908; 66. 385, 1909 ; A. A. Trussewitsch, ib., 6. 360, 1890; G. Quincke, Pogg. Ann., 134. 356, 1868 ; 135. 621, 1868; R. Schiff, Gazz. Chim. Ital., 14. 368, 1884 ; Liebig's Ann., 223. 47, 1884. ' 6 T. Graham, Phil. Trans., 136. 573, 1846 ;139. 349, 1849 ; A. 0. R ankine, Proc. Roy. f loe., 86. A, 162, 1912 ; 88. A, 575, 1913 ; A. Campetti, Nuovo Cimento, (6), 17. 143, 1919 ; L. Kann, Sitzber. Akad. Wielz, 106. 431, 1897 ; T. E. Thorpe and J. W. Rodger, Phil. Tray., 185. A, 397,1894 ; 189. A, 71,1897 ; J ourn. Chem. Soc., 71.360, 1897 ; E . C. Bingham, Zeit. phys. Chem., 66. 238, 1909 ; K . Beck, ib., 48.641, 1904. l o C ;F. Jltger, Monatsh., 8. 498, 1888; A. 0. Rankine, Proc. Roy. SOC., A, 162, 1912; 86, l 7 H. Billet, Mkm. I'lnstitut., 292, 1855 ; T. E. Thorpe, Journ. Chem. Soc., 37. 141, 1880; J. Dewar, Chem. NEWS, 216, 1905 ; J. I. Pierre, Ann. Chim. Phys., (3), 20: 5,1847 ; A. h n g e , 91. Zeit. angew. Chem., 13. 683, 1900 ; R. Knietsch, Liebig's Ann., 259. 100, 1890. l 8 H. D avy, Phil. Trans., 101. 155, 1811 ; M. F araday, Quart. Journ. Science, 15. 71, 1823. l g J. A. Paris, Lije of Sir Humphry Davy, London, 1831 ; H. D avy and M. F araday, Phil. Trans., 113. 160,198,1823 ; K . Ohchewsky, Monatsh., 5. 127,1884. 2 0 J. Dewar, Phil. Mag., (5), 18. 210, 1884; Chem. NEWS, 51. 27, 1885 ; A. Ladenberg, Ber., 11. 822, 1878 ; R. Knietsch, Liebig's Ann., 259. 100, 1890; T. Andrews, B. A. Rep., 66, 1871 ; A. Nadejdine, Bull. Acad. St. Petersburg, 12. 299, 1885; M. Pellaton, Journ. Chim. Phys., 13. 426, 1915. 21 H V R egnault, Mkm. Amd., 26. 1, 1862 ; R . Knietsch, fiebig's Ann., 259, 100, 1890; .. F. M. G. Johnson and D. Macintosh, J ourn. Amer. Chem. Soc., 31. 1138, 1909; M. Pellaton, Journ. Chim. Phys., 13. 426,1915 ; A. Juliusberger, Ann. Physik, (4), 3. 618,1900 ; E . Marchand, ib.., 11. 573. 1913. . -- -2 2M. pellaton, Journ. Chim Phys., 13. 426, 1915. 23 J. I. Pierre, Ann. Chzm. Phys., (3), 20. 5, 1847 ; A. Landolt, Liebig's Ann., 116. 177, 1860 ; E. Mitscherlich, i b., 12. 159, 1834; E. P. Perman, Proc. Boy. Soc., 48. 45, 1890 ; E. P. Perman and G. A. S. Atkiison, i b., 66.10,1900 ; C. a nd M. Cuthbertson, ib., 85. A, 306,1911 ; B. G. Eggink, Zeit. phys. Chem., 64. 449, 1908 ; J. D. v an der Plaats, Bec. Trav. Chim. Pays-Bas, 5. 34, 1886; H. W. B. Roozeboom, i b., 3, 29, 59, 73, 84, 1884 ; H. Jahn, Sitzber. Akad. Wien, 85. 778, 1882; P. C. E. M. Terwogt, Zeit. anorg. Chem., 47. 203,1905 ; B. J. K arsten, z%., 53. 365, 1907 ; J . S. Stas, Mkm. Amd. Belgique, 35.3,1865; C. Lawig, Das Brom und seine chemischen VerhiiBnisse, Heidelberg, 1829 ; Pogg. Ann., 14. 485, 1828 ; W. Ramsay and S. Young, Journ. Chem. 80c., 49. 453, 1886; T. E. Thorpe, ib., 37. 141, 237, 1880 ;T. Andrew, ib., I. 18, 1847 ; T. Bolas and C. E. Groves, ib., 24. 773, 1871 ; A. Nadejdine, Bull. Amd, St. Petcrsburg, 12. 299, 1885; R, Knietsch, Liebdg's A m . , 259. 100, 1890; T. Isnardi, Ann. Physik, (4), 61. 264, 1920. 2 4 E . Beckmann, Zeit. phys. Chem., 46. 853, 1900. z s J. L. Gay Lussac, Ann. Chim. Phys., (I), 91. 5, 1814 ;J. Dewar, Chem. new,^, 78. 325, 1898; W. Ramsay and 8 . Young, Journ. Cl~em.Soc., 4 . 453, 1886 ; G. P. Raxter, C. H. Hickey, and 9 W. C. Holmes, Journ. Amer. Chem. Soc., 29. 127, 1907 ; W. M. Gorbenko, Chem. Zty., 30. 1303, 1906; H. V. R egnault, Ann. Chim. Phys., (3), 46. 257, 1857. z 6 J. Dewar, Chem. NEWS, 78. 325, 1898 ; W. Nernst, Nachr. GES. Wiss. attingen, 1. 1906; Applications o Thermodynamicsto Chemistry, New York, 1907 ; H. Arctowsky, Zeit. amrg. Chem., f 12. 427, 1896 ; R . T surata, P h k . Zeit., 1 417, 1900 ; R. Naumann, E x erimentelk Bestimmung . und theordische Berechnung kkiner Dampfdrucke oon Jod und Meta Eiodiden, Berlin, 1907; K. Stelzner and G. Niederschulte, Verh. deul, phys. G ES.,7. 159, 1905. . K. O lschew~ky, Monatsh., 5. 127, 1884. 2 8 J. I Pierre, Ann.. Chim. Phys., ( 3),3 0.5,1850; H. V. R egnanlt,ib.,(3), 26.274,1849; 3 .Baum. . 3 hauer, Ber., 4. 927, 1871 ; J. Philipp, ib., 12. 1424, 1879 ; 5. D. v an der Plaats, BEC.Trav. Chim. Pap-Eas, 5. 34, 1886 ; P. C. E. M. Terwogt, Zeit. czn.org. Chem., N , 03, 1905 ; W. Ramsay and 2 S. Young, Journ. Chem. Soc., 49. 453, 1886; H. C. P. Weber, Journ. Amer. Chem. Soc., a. 294, 1 1912; T. Tsnardi, Ann. Phlysik, (4), 61. 26Li, 1920. H. V. Regnault, Ann. Chim. Ph?ys., (3), 46. 257, 1857 1 W. Ramsay and S. Young, Journ. Chem. Soc., 49. 453, 1886; B. Lean and W. H. Whatmough, i b., 73. 148, 1898 ; A. Ladenburg, Ber., 35. 1256, 1902 ; J. Timmermans, -Journ. Chim. Phys., 4. 170, 1906 ; J. S. Stas, Mkm. Acad. Belgique, 35. 3, 1865; G. P. Baxter, C. H. Hickey, and W. C. Holmes, ,Journ. Amer. Chem. Soc., 29. 127,1907 ; G. P. Baxter, z%., 32. 1591,1910 ; G. N. Lewis and M. Randall, ib., 36. 2259, 1914 ; V. von Richter, Ber., 19. 1057, 1938, 1856. T. Carnelley, Proc. Roy. ~Soc., 1. 284, 1881 ; Chem. News, 42. 130, 1880 ;L. Meyer, Bw., 8. 3 1627, 1875 ; 13. 1831, 1880 ; 14. 718, 1881 ; 0, Pettersson, Ber., 13. 2141, 1880; R. W. Wood, Phil. Mag., ( 6), 32. 364, 1916. 31 H. Billet, Mkm. I'lnstitut, 292, 1855; J. 1 Pierre, Ann. Chim. Phys., (3), 20. 5, 1847; . K. T surata, Phys. Zeit., I. 417, 1900 ; M. Toepler, Wied. Ann., 53. 343, 1894. 3 2 R . Knietsch, Liebig's Ann., 259. 100, 1890 ; T. Estreicher and M. Staniewsky, Bdl. Acad. Cmcovie, 349,1910 ; T. Estreicher and A. Schnerr, ib., 345,1910 ; F. M. G. J ohnson and D. McIn. iosh, Journ. Amer. Chem. Soc., 31. 1138, 1909 ; G. P. B axter, C. H , Hickey, and W. C. H o h w , 0 P THE HALOGENS 8.,29. 127, 1907 ; G. N. Lewis and M. Randall, ib., 36. 2259, 1914 ; T. Andrcws, Journ. Chern. SOC.,1 18, 1847 ; H. V. Regnault, Mkm. Amd., 26. 701, 1862 ; N. Berthelot and J. Ogier, A m . . Chim. Phys., (51, 30. 410, 1884 ; K . Tsurata, Phys. Zeit., 1. 417, 1900 ; J. Dewar, Chem. News, ; 78. 325, 1898 ; 91. 21 6, 1905 ; P. A. Favre and J. T. Silbermann, Ann. Chim. Phys., (3), 37 461, 1853 ; Compt. Rend., 29. 449, 1849 ; M. B erthebt, Thermochemie, Paris, 2. 44, 1897 ; W. Nernst, Nachr. Gott., I , 1906 ; R. Naumann, Experimenklle Bestimmung und theoretische Berechnung kkiner Barnpfdrw:ke von Jod und Metalliodiden, Berlin, 1907 ; H. Arctowsky, Zeit. anorg. Chm., 12. 427, 1896 ; T. Isnardi, Ann. Physik, (4), 61. 264, 1920. 33 H. V. Regnault, Mkm. Acad., 26. 1, 1862 ; R ehtion des cxpCriemes pcur determiner les lois et Es donnkes physiques, Paris, 1862 ; K. Strecker, W i d . Ann., 13. 20, 1881 ; 17. 85, 1882 ; e T. Mwtini, Atti 1st. Veneto, (5), 7. 491, 1881 ; E . Mallard and H, le Chatelier, Compt. Rend., 93. 062, 1014, 1076, I881 ; Becherches expbrimentales et thkoriques sur la combustion des mhlunges gcbzeaux explo.sij8, Paris, 1883 ; A. Lan en, Mitl. Forsch. Geb. 11tg. Wesem, 8, 1903 ; M. Piel, Stevens, Ann. Physik, (4), 7. 285, 1902 ; W. Nernst, Z d . phys. Chem., 62. 385, 1908 ; E. Zebt. Elektr~~hem., 265, 817, 1911 ; W. Nernst, F. Koref, and F. A. Lindemann, Sitzber. pew$. 17. Akad., 13.261,1910 ; T. Estreicher and M. Staniewsky, Bull. A d . Cracowie, 834,1912 ; F. Haber, Themnodynumik technischer Ga~eaktionen,Miinchen, 1905 ; M, Bodenstein and A. Geiger, Zeit phys. Chm., 49. 70, 1904; G. Starck and M. Bodenstein, Zedt. Elektrochem., 16. 961, 1910; G. N. Lewis and M. Randall, Journ. Amer. Chem. Soc., 34. 1128,1912 ; J. Dewar, Proc. Roy. Sos., 89, A, 158, 1913; A. Campetti, Nuovo Cimento, (6), 17. i, 143,1919 ; M. Trautz, Zeit. Ekktrochem., 18.613,1912 ; J. W. Mellor, Journ. Chem. S oc., 81. 1280, 1002. 34 Knietsch, Liebig's Ann., 259. 100, 1890 ; T. Estreicher and M. Staniewsky, Bull. A d . Cramvie, 349, 1910; H. V. Regnault, Ann. Chim. Phys., (2), 73. 3, 1840 ; ( 3), 46. 257, 1856 ; (3),28.268,1849 ; P. A. F a a e and J. T. Silbermann, ib., (3), 37. 461, 1853 ; T. Andrews, Journ. Chem. Soc., I. 18, 1847 ; R. Abegg and F. Halla, Handbuch &r anorganzschen Chemie, Leipzig, 4. ii, 356, 1913 ; H. Bwechall, Zeit. Elektrochem., 17. 341, 1911 ; L. v on Pebal and H. Jahn, W i d . Ann., 27. 594,1886 ; F. Koref, Ann. Physik, (4), 36. 49,1911 ; A. Magnus and F. A. Lindemann, Zeit. Elektrochem., 16. 271, 1910 ; W. Nernst, ib., 17. 265, 817, 1911 ; Ann. Physik, (4), 36. 395, 1911 ; Sitzber. A k d . Berlin, 247, 1910; 65, 1911 ; J. Dewar, Proc. Roy. Soc., 89. A, 158, 1913. 36 G. N. Lewis and G. E. Gibson, Journ. Amer. Chem. Soc., 39. 2654, 1917 ; G. N. Lewis and bl, Randall, ib., 36. 2259, 1914; 38. 2348, 1916; R. C. Tolman, ib., 42. 1185, 1920; E. Briner, Joum. Chim. Phys., 12. 109, 1914. 36 R . Clausius, Pogg. Ann., 1 0 1353, 1857. 0. P. L. Dulong, Ann. Chim. Phys., (2), 31. 154, 1826 ; M. Croullebois, &., ( 4), 20. 136, 1870 ; J. Dechant, Monatsh., 5. 615, 1884; 5. H. Gladstone, Chem. News, 55. 300, 1887 ; 67. 94, 1893 ; F. Eisenlohr, Zeit. phys. Chem., 75. 585, 1910 ; E. Comady, ib., 3. 226, 1889 ; J. W. BruhI, ib., 7. 191, 1891 ; Liebig's Ann., 203, 1880; Ber., 13, 119, 1880; E. Mascart, Compt. Rend., 86. 321, 1878;C. RlviBre, ib., 131. 671,1900; F. P. le Roux, ib., 55.126,1862; L. Bleekrode, Proc. Roy. Soc., 37.339, I884 ; A. Haagen, Pogg. Ann., 131. 117, 1867 ; A. H. Borgesius, Wied. Ann., 54. 221, 1895; F. Zecchini, Cfazz. Chim. ltal., 22. 592, 1892 ; C. and M. Cuthbertson, Phil. Trans., 213. A, 1,1913 ; C. Cuthbertson and E. P. Metcalfe, i. 20'7. A, 135, 1908 ; A. Hurion, Ann. 1'Ecok b, Nw., (2), 6. 367, 1887 ; H. h n d o l t , Sitzber. Akad. Berlin, 64, 1882. 38 J, Plucker and W. Hittorf, Phil. Trans., 155. 24, 1865 ; J. Plucker, Pogg. Ann., 106. 83, 1859 ; 107. 528, 1859 ; V. w n Villigen, ib., 106. 624, I859 ; J. Moser, ib., 160. 177, I877 ; R. Thden, ib., 139. 503, 1870 ; A. Wiillner, ib., 120. 158, 1863 ; A. Mitscherlich, ib., 121, 474, 1864; G. Salet, Ann. Chim. Phys., (4), 28. 24, 1873 ; A. Morren, Compt. Bend., 68. 376, 1860 ; D. Gernez, ib., 74. 660, 1872 ; J. Chautard, ib., 79. 1123, I874 ;E. Robiquet, i b., 49. 606, 1859 ; W.A,MiUer, Phil. Mag., ( a),2. 381, 1833; (3), 27. 81, 1845 ; G. L. Ciamician, Sitzber. Akad. Wiem, 78. 872, 1878 ; J. M. Eder and E, Valenta, Denkschrift Akad. Wien, 1, 1899 ; A. Schuster, B. A. Rep., 258, 1880 ; A. J. Angstriim, i b., ( 4), 43. 398, 1872 ; E. Goldstein, Ve?*h. deut. phys. Ges., 38, 1886 ; H. Ebert, Wied. Ann., 53. 144, 1894 ; A. Ka'laline, ib., 65. 815, 1898 ; H. Konen, i b., 65. 257,1898 ; G. D. Liveing and J. Dewar, Proc. Roy. Imt., 10. 245, 1883 ( I a m indebted to thesc investigators for permission to use Figs. 18 and 19); E. R. Iaird, Astrophys. Journ., 14. 95,1901 ; J. Tyndall, Contributions to Molecular Physics in the domain o Radiant Heat, London, f 80, 1872; K. Angstriim and W. Palmaer, Ofvers. Akad. Fbrlland., 6. 389, 1893; C. Grrlinge, Lehrbuch der angewandten Optik, Braunwhweig, 217, 1886 ; H . E. Roscoe and T. E. Thorpe, Phil. Trans., 167.207,1876 ;B. Hasselberg, MCm. Acad. St. Peter-sburg, 26.4,1878 ; J. N. Lockyer, h. Soc., 23. 348, 1875; J. N. Lockyer and W. C. Roberts, ib., 23. 348, 1875; D. Forbes, Roy. Phil. Mug., (4), 1 . 302, 1865 ; C. F. Schanbein, Journ. prakt. Chem., (I), 55. 154, 1852 ; W. Frie1 derichs, Zed. wiss. Phot., 3. 154, 1905 ; C. Ribaud, Ann. Phys., (9), 12. 107, 1919 ; C. S. Sellak, Pogg. Ann., 140. 334, 1870 ; J. Conroy, Proc. Roy. Soc., 25. 46, 1876 ; Phil. Mag., (5), 3. 68, 1877 ; H. W. Vogel, Ber., 11. 913, I878 ; W. Burmeister, Verh. deut. ~ h y s .Ges., 15. 589, 1913 ; T. Aadrews, B. A. Rep., 66, 1871 ; W. W. CobIentz, Phys. Rev., ( l),16. 72, 1903 ; R . W. Wood, Phl. Mug., (5), 41. 423, 1896; J. Trowbridge and T. W. Richards, ib., (G), 43. 135, 1897. 9. J. Evans, Astrophys. Journ., 32. J , 391, 1911 ; R. W. Wood, Phil. Mag., (6), 35. 236, 1918. 66 M Siegbahn, Ber. deut. phys. G h . , 18. 39, 1916; M. d e Broglie, Compt. Rend., 158. 1785, . 1914 ; I. Malmqr, U ntersuchu~ fiber die Hochfrequempe ktra der Elemente, Lund, 1916. em R. W. Wdod, Phil. Mag., (6), 12. 329, 1906; J. Evershed, ib., ( 5), 39. 460, 1895 ; R. Nasiui . " INORGANIC AND THEORETICAL CHEMISTRY and F . Anderlini, Atti Accad. Lincei, ( 5 ) ,13. i i, 59, 1904 ; L . Puccianti, ib., ( 5 ) , 14. i , 81, 1905'; W . Friederichs, Zeit. wiss. Phot., 3. 154,1905 ; C. Fredenhagen, Phys. Zeit., 8. 89,1907 ;M. R einganum, ib., 8. 182, 1907 ; W . Steubing, Verh. dezit. phys, GES.,15. 1181, 1913 ; A nn. Physik, ( 4 ) , 58. 55,1919 ; A. S mithells, Phil. Mag., ( 5 ) ,37. 252,1894 ; ( 5 ) ,39. 132, 1895 ; G . S alet, Ann. Phys. Chim., ( 4 ) ,28. 34, 1873 ; H. Konen, Wied. Ann., 65. 279, 1898. ' 2 G. G. S tokes, Phil. Trans., 142. 463, 1853. 4 3 P. L enard, Ann. P h p i k , ( 4 ) , 15. 425, 1904; ( d), 28. 476, 1909; (4), 31. 641, 1910; R. W . Wood, Ph.ysica1 Optics, New York, 1914; J. Stark, Die Prinzipien der Atomdynamik, Leipzig, 1910-5 ; N. R. Campbell, Modern Electrical T k o r y , Cambridge, 1913 ; S . Landau and E . S tenz, Phil. Mug., ( 6 ) , 40. 189, 1920; R . W . Wood, ib., ( 6 ) , 35. 236, 1918; W . P. S p a s , ib., ( 6 ) , 27. 531, 1914; W . Steubing, Ann. Phys., ( 4 ) , 58. 55, 1920; Verh. deut. phys. Ges., 15. 1186, 1914 ; Zedt. P h y i k , I. 426, 1920. 4 4 E . B udde, W i d . Ann., 144. 313, 1871 ; Pogg. Ann. Ergbd., 6. 477, 1873 ; P hil. Mag., ( 5 ) , 42. 290, 1871 ; A. Richardcon, ib., ( 5 ) ,32. 277, 1891 ; J . W . Mellor, J o ~ r n . Chem. Soc., 81. 1280, 1902 ; Mr. A. Shenstone, ib., 71. 471, 1897 ; H. B. Baker, B. &4.Rep., 493, 1894 ; P. CaldweIl, Amer. Chem. Joztm., 31. 51, 1904; M. Trautz, Zeit. Elektrochem., 18. 513, 1912 ; M. Recklinghausen, Zeit. phys. Chem., 14. 494, 1894; A. Campetti, Nuovo Cimento, 17. i , 143, 1919. 4 5 J . J. Thomson, P ~ o cRoy. Soc., 42. 343, 1887; Proc. Cambridge Phil. Soc., 11. 90, 1901; . P. V . Bevan, Phil. Tram., 202. A, 91, 1903; E . P. Perman, Proc. Roy. Soc., 48. 45, 1891; W . Kropp, Zeit. Elektrochem., 21. 356, 1915 ; M . le Blanc and M. Vollrner, ib., 20. 494, 1914; P. Lenard, Sitzber. Aknd. Heidelberg, 31. 28,31,32, 1910 ; 32. 10,24, 1911 ; E . B . Ludham, Phil. Mrrg., ( 6 ) , 23. 757, 1912 ; G. K iimmeII, Zeit. Elektrockem., 17. 409, 191 1 ; E . R adel, Zeit. phys. Chem., 95. 378, 1920. 4 6 J . S . ohs send, Proc. Cambridge Phil. Soc., 9. 345, 1898 ; P hil. Mag., ( 5 ) ,45. 125, 1898. 4 7 A. L . Hughes and A. A. Dixon, Phys. Rev., ( 2 ) ,10. 495, 1917 ; K . T . Compton, ib., ( 2 ) ,8. 412, 1916 ; E . B. L udham, Phil. Mag., ( 6), 23. 757, 1912 ; J . H enry, Proc. Cambridge Phil. Soc., 9. 319, 1898 ; E. Whiddington, ib., 15. 189, 1910 ; J . Franck and W . Westphal, Verh. deut. php. Ges., 14. 159, 1.912; F. R itter, Ann. Physik, ( 4 ) ,14. 118, 1904; E . W arburg and F. R . Gorton, ib., (4), 18. 128, 1905; Ueber das Minimumpotential bei S itzenentladung in Chlor, Brom und lod und die Empflndlichkeit von Spitze72, Berlin, 1903; W . a tthies, ib., (4),18. 473, 1905 ; Phys. Zeit., 6. 729, 1905 ; G. B ode, ib., 6. 618, 1905; C. G. F ound, Phys. Rev., ( I ) , 16. 41, 1920. 0. R eichenheim, Verh. deut. phys. Ges., 11. 168, 1909 ; R. Raisch, Ann. Physik, (4),36.907, 1811. 4 9 K . ICellner, Zeit. Elektrochem., 8. 500, 1902 ; E . Briner and E . D urand, ib., 14. 706, 1908; H. M. Vernon, Chem. News, 63. 67, 1891 ; 5 . W . Mellor, Journ. Chem. Soc., 81. 1280, 1902. 6 0 F . L inde, Wied. Ann., 56. 563, 1895 ; W . A. P lotnikoff,Zeit. phys. Chem., 4 8. 220, 1904; G . N . Lewis and P. Wheeler, ib., 56. 179, 1907 ; F. E xner, Sitzber. Akad. Wien, 84. 511, 1881. &f J . L. Gay Lussac, Ann. Chim. Phys., ( I ) , 91. 8, 1814; P. Jolly, Pogg. Ann., 37. 420, 1836; P. Reiss, ib., 64. 52, 1845 ; W . Beetz, ib., 92. 452, 1854 ; C. J . K nox, Phil. Mag., ( 3 ) ,9. 450, 1836 ; ( 3 ) ,16. 188, 1840 ; J . Inglis, ib., ( 3 ) , 7. 441, 1835 ; K . Fajans, Ber. deut. p h y ~ .Gea., 21. 549, 1919; M. Trautz and F. A. Henglein, Zeit. anorg. Chem., 110. 237, 1920 ; F . W . Aston, Phil. Mag., ( 6 ) ,39. 611, 1920 ; P. Lenard, W . Weick, and H. F . Mayer, Ann. Physik, ( 4), 61.665, 1920; R. Lorenz, Zeit. phys. Chem., 73. 253, 1910; P. Walden, ib., 60. 807, 1908; Ion, I . 411, 1909 ; Zeit. a.norg. Chem., 113. 125, 1920. 6 WM. . T hompeon, Journ. Amer. Chem. Soc., 28. 731,1906 ; F. W . ICiister and F. Crotogino, de K Zeit. anorg. Chem., 23. 87, 1900 ; F. Boericke, Zeit. Elektrochem., 11. 57, 1905 ; E . C. Sullivan, Zeit. phys. Chem., 28. 523, 1899 ; A. A. J akowkin, ib., 29. 637, 1899; F. J . Smale, Zed. phys. Chem., 14. 577, 1894; 16. 564, 1895 ; N . T . M . W ilsmore, ib., 35. 317, 1900 ; E . Miiller, ib., 40. 158, 1902 ; F . Dolezalek, ib., 2 . 321, 1898 ; E. C. Sullivan, ib., 28. 538, 1899 ; 6 R . Luther, ib., 30. 628, 1899 ; W . Nernst, Ber., 30. 1547, 1897 ; Zeit. Ekktrochm., 4. 539, 1897 ; H . W ohlmill, ib., 5. 56, 1895 ; F . Haber, ib., 7. 1043, 1900 ; F. Boericke, ib., 11. 57, 1904 ; G . N . Lewis and F . F. R upert, Journ. A m r . Chem; Soc., 33. 299, 1911 ; G. N . Lewis and H . S torch, ib., 39.2544,1917 ; W . von Beetz, P q y . Ann., 77.493,1849 ; R . 0. Pierce, W i d . Ann., 8 . 98, 1879; P. D. Foote and F. L . Mohler, Journ. Amer. Chem. Soc., 42. 1893, 1920. F. Linde, W i d . Ann., 56. 546, 1895 ; W . D. Coolidge, ib., 69. 125, 1899 ; H. S chlundt, Joum. Phys. Chem., 5. 157, 503, 1901 ; W . Schmidt, Ann. Physik, ( 4 ) ,9. 919, 1902 ; ( 4 ) ,11. 114, 1903 ; P. Eversheim, d ., ( 4 ) ,13.492,1904 ; C. Riviere, Compt. Rend., 131. 671,1900 ; P. Walden, Zeit. phys. Chem., 70. 569, 1909. 6 4 K . H onda, Ann. Physik, ( 4), 32. 1027, 1910 ;P. Curie, Corn@. Rend., 115. 1292, 1892 ; 116. 136, 1893; P. Pascal, ib., 152. 862, 1911 ; Bull. Soc. Chim,., ( 4 ) , 11. 201, 1912; G. Quincke, Ivied. Ann., 24. 347, 1885 ; 34. 401, 1888; M. Faraday, 136. 41, 3846 ; S. Wleiigel and S . Henrichsen, Wied. Ann., 22. 121, 1884; S. Henrichsen, ib., 34. 180, 1888 ; 45. 38, 1892. 6 6 W . H. Perkin, Journ. Chem. Soc., 65. 2 0, 1894; H. Jahn, W i d . Ann., 43. 280, 1 891; 0. H umburg, Zeit. phys. Chem., 12. 401, 1893 ; A. Heurung, Ann. Physik, ( 4 ) ,36. 153, 1911 ; R . W . Wood, Phil. May., ( 6), 12. 329, 1906 d THE H ALOGEM 71 rj 11. Solutions of Chlorine, Bromine, and Iodine in Water, etc. Chlorine and bromine are fairly soluble in water ; iodine has a low solubility. Early determinations of the solubility of chlorine in water were made by J. L. Gay Lussao 1 in 1839, by J. Pelouze in 1843, and by F. Schonfeld in 1855. They noticed a maximum in the solubility curve in the vicinity of lo0, and at 100" the solubility is nil. Later determinations have been made by H. W. B. Roozeboom in 1885,and by L. W. Winkler in 1907. At temp. below 9.6", chlorine forms a crystalline hydrate, C1,.8H,O ; a nd this corresponds with the maximum in the solubility curve. The solubility curves of the gases chlorine and bromine 2 are indicated in Table 111. --- Chlorine. Temperature. Bromine. Absorption weff. Grms. C in 1 100 grms.water 760 mm. Temperature. Absorption coeff. Grms. 3 r in 100 grrnu. water 760 mm. --- -- 4.610 3.947 3-441 3.031 3-095 2-980 2.900 2.635 2.260 1.985 1.769 1'414 1- 204 1-006 0,848 0,672 0.380 0-000 0 3 6 9 9.6 10 12 15 20 25 30 40 50 60 70 80 90 100 0 1-46 1'25 1-08 Oq96 0-94 0.980 0-918 0-835 0.7 16 0-641 0.572 0-459 0-393 0.329 0.279 0.223 0.127 0.000 60.5 54.1 48.3 43.3 38.9 35.1 31-5 28-4 25.7 23.4 21.3 19.4 13-8 9'4 6.5 4.9 3-8 3.0 2 4 6 8 10 12 14 16 18 20 22 30 40 50 60 70 80 42.9 38.3 34.2 30-6 27-5 24-7 22.2 20.0 18.0 16-4 14'9 13.5 9.5 6.3 4.1 2.9 1.9 1.2 L. W. Winkler represents the absorption coeff. of chlorine in water at 8" by the formula: 3'0361-0~461978+0~000110782. The solubility curve really represents the increasing solubility of chlorine hydrate, and the decreasing solubility of chlorine gaa with rising temp. The maximum, about 9.6", occurs when the latter begins to predominate over the former. The solubility of chlorine in water, and the solubility when the gas is mixed with hydrogen or carbon dioxide, is greater, between 13" and a", corresponds with its partial press.3 According to L. W. Winkler, howthan ever, bromine vapour dissolves in water in accord with Henry's law. The soh. of chlorine in water is called chlorine water, or aqua chlorata ; nd the soh. of bromine a in water, bromine water, or aqua brornata. The solubility of liquid bromine in water, represented by the number of grams of bromine in 100 grms. of the soln., is water .. . Solubility. 0° 6 ' 10" 16" Zoo 26" 3' 0 4.17 3 -98 3-92 3-77 3-74 3-65 3 -52 3.58 3-46 3.48 3.36 3-44 3-32 3-61 40" 3-45 3-33 5' 0 3-52 3-40 The solubility of iodine in water has not been so closely investigated as that of chlorine or bromine,4 and the determinations are not in close agreement. H. Hartley and N. P. Campbell found that water dissolves 0" Iodine , 0.1620 18" 0:2765 2GP 3j0 46" 5' 6 0-3395 0.4661 0'6474 0.9222 6' 0 0-9566 grms. per litre I. The value at 0" is by G. Jones and R L. Hartmann, and that a t 60" by 72 INORGANIC AND THEORETICAL CHEMISTRY R . L uther and G. V. Sarnmet. The soln. require one t o two days to reach the point of sat. and they are brownish-yellow in colour. Crystals of chlorine hydrate or of bromine hydrate a re readily formed when aq. s o h of chlorine or bromine respectively acre cooled to about 0". According to J. Pelouze crystals of chlorine hydrate are formed when a few drops of hydrochloric acid are added to an aq. soln. of hypochlorous acid cooled to about 2" or 3". A. Ditte sealed chlorine hydrate, containing an excess of water, in a long bent A-tube. The leg containing the hydrate was warmed, and the chlorine was condensed to a liquid in the cold leg of the tube by the press. of its own vapou. When the apparatus was slowly cooled, fine fern-like crystals of the hydrate re-form on the surface of the water. On standing some time, these crystals disappear re-forming isolated greenish-yellolv crystals about 2 or 3 mm. long, These crystals are highly refracting, and belong to the cubic system. The crystals of either hydrate are formed by leading the gas or vapour through a tube cooled to about 4" and moistened on the inside with water. I n winter tiine, pipes 6 conveying chlorine gas may become clogged by the formation of the hydrate. Chlorine hydrate was formerly thought to be solidified chlorine, but H. n avy demonstrated that the crystals contained water. The yellow crystals of chlorine hydrate examined by M. P araday 7 (1832) contained 27.7 per cent. of chlorine and 72'3 per cent. of water ; t his is nearly eq. to C1.5Hz0, or to C12.10Hz0,and he added, " I have chosen it because it gave the largest proportion of chlorine of any experiment I made ; " a nd owing to inevitable losses of chlorine in drying, he said, " I t is even possible that this proportion of chlorine is under-rated, but I believe it to be near the truth." H. W. B. Roozeboom (1884) makes the proportion C12.8H20, and R. de Porcrand (1901), C1,.7H20. The existence of the hydrates, C1,.12H20 ; C12.7H20; a nd C12.4H20, reported by G. Maumenk, have not been confirmed. Hyacinth-red octahedral crystals of bromine hydrate, sp. gr. 1-49 (4"), were obtained by C. Lowig in 1829 ; their composition was determined by W. Alexejeff in 1876, and by H. W. B. Roozeboom (1884), whose analyses agree with the formula, Br,.10H20. H. Giran's thermal analysis of the binary system Br-H20, and his analyses of the crystals, agree with the formula Br2.8H20. According to M. F araday, the crystals of chlorine hydrate may be sublimed without decomposition in a sealed tube a t 15'5" (M. F araday), or a t 20' (E. Biewend) 8 ; b ut according to -F. Wohler a t 38") they form two liquids-one rich in chlorine, is a soln. of water in chlorine ; a nd the other rich in water, is a s o h . of chlorine in water ; according to W. Alexejeff, bromine hydrate also decomposes into two analogous layers a t 15". I n both cases the hydrate is re-formed on cooling to 0". T h e critical temp. of decomposition of chlorine hydrate in open vessels is 9.6", and in closed vessels, 28.7". According to H. W. B. Roozeboom, the dissociation press. of chlorine hydrate from -10" to -0.24" is between 156 and 248 mm. ; a nd between -0.24" and 28,7", from 248 mm. to 6 atm. There are turningpoints in the curve a t -0.24" and 28.7". F. I sambert's B measurements of the dissociation press. of chlorine hydrate between 0" and 14.5" agree fairly well with those of H. W . B. Roozeboom, According to the latter, the decomposition tension o f bromine hydrate from -10" to -0.3" is from 25 to 43 mm.; from -0.3" to 6'2O, 43 to 93 mm. Above 62", bromine hydrate crystals form an emulsion of water and bromine which slowly separates into two liquids-(i) a soln. of water in bromine and (3) a s ob. of bromine in water. I n open vessels, under atm. press., bromine hydrate decomposes at 6.8" ; a t higher press. the decomposition temp. is still higher -at 150 atm. of oxvgen press., it is stable a t 20°, according to P. Villard ; l o a nd at 200 atm. of hydroken press., it is stable a t +go. The increased stability of tho hydrate in compressed oxygen is connected with the great solvent power of compressed oxygen gas for bromine vapour, and the consequeut increased partial press. of the atm. of bromine on the hydrate. P. Villard has measured the influence of compressed gases on the vapour press. of liquid bromine, and he found that bromine vapour is fairly soluble in compressed oxygen 73 T HE HALOGENS such that a t 300 atm. press. compressed oxygen takes up six times the normal amount ; if the press. is released, the oxygen gives up the dissolved bromine in the form of small drops: The greater the press, the greater the amount o bromine dissolved by t he compresf sure oxygen, and the deeper the colour. Compressed nitrous oxide takes u p t he same amount of bromine a t 20 atm. press, as i t does a t 40 atm. Iodine also volatilizes in compressed oxygen and in compressed methane, and crystallizes out when the press. is relieved. Ethylene gas a t 300 atm. press. is coloured deep violet by t.he dissolution of iodine, and the colour gradually disappears on standing owing to the union of the iodine with the ethylene. A n interesting illustration o the dissolution of vapours by compressed gases is furnished f by methane which dissolves its own volume of ethyl chloride, C,H,Cl, a t 180 atm. ; a t 200 atm. press., the two substances are miscible in all proportions, and the surface of separation between liquid and gas disappears. P. Villard proposes to take advantage of the property for the distillation of substances decomposed by heat. The vapour is dissolved when the gas is compressed and rejected when the press. is relieved. H. W. B. Roozeboom has measured the solubility of chlorine and bromine hydrates in water and expressed his results in terms of " grams of chlorine per 100 grms. of s o h " He found that 100 grms. of soh. contain 0.492 grm. of chlorine at -0.24", a t which temp, the solid phase present is mixture of ice and chlorine hydrate, C12.8H20. Between 0" and 28.7" the solid phase is the hydrate alone, and the soh. has O0 Chlorine 0.533 2" 0'644 6" 4O 0.732 0-823 8" 0-017 9" 20' 0'037 1-85 28'7" 3-69 per cent. The binary system has not been completely explored. Similarly with bromine hydrate 100 grms. of s o h dissolve 2.17 grms of bromine a t -0-30°, and the solid phase is ice and the hydrate Br2.10H20. Between -0.3" and 6.2', the solid phase is tho hydrate, Br2.10H20,alone, and the soln. has Bromine .. -0 ' 2 O 2.25 O0 2.31 3O Go 2-97 3.50 6'2O 3.53 per cent, I n the latter case, there are two soln., one layer contains a soh. of bromine in water, the other, a s o h . of water in bromine. There is an unstable system with ice as the solid phase a t -3'7" with 3'03 grms. of bromine, which may form before the hydrate, Brz.10H20, separates. In the system C12+H20, there are two com- sdbm, ponebts, just indicated ; t wo solid phases-ice and chlorine hydrate, Cl,.8H20 ; two soh-one a soh. o water in an excess of chlorine, Sol. I, and a soh. f of chlorine in an excess of water, Sol. I1 ; a nd a gas phase-a mixture of chlorine and water vapour in varying proportions. The system has not been completeky studied, but sufficient is known to show - 0 . mpemtums . 7 ~ %~ 9 ~ ~ 8 that the equilibrium curves take the form shown 20.-E:quilibrium in t he diagrammatically in Fig. 20. The two invariant ~i~~~ system Wstrr-&lo: systems L and B have four coexisting phasesrine. L (28.7" ; 6 a tm.) denotes the point at which the hydrate decomposes-ice, Sol. I, Sol. 11, vapour ; B (-0'24' ; 244 mm.) is a eutectio point-ice, hydrate, s o h , vapour. The univariant systems with three coexisting phases comprise the curves : BL-hydrate, Sol, 11, vapour; DLhydrate, Sol. I, vapour ; LB-Sol. I, Sol. 11, v apour; LH-hydrate, Sol. I, 801. I1 ; CB-hydrate, ice, vapour ; BF-ice, Sol. 11, vapour ; BG-ice, hydrate, , Sol. 11. The bivariant systems, representing two coexisting ~ h a s e s are included in the areas bounded by the curves. A similar result was obtained with the system Br2+H20, where the hydrate is Brz.10H20,and the invariant systems occur a t the two quadruple points : L (6.2" ; 93 mm.) and B (-0.3" ; 43 mm.). Iodine forms no known hydrate with water. $ 74 INORGANIC AND THEORETICAL CHEMISTRY W. L . Goodwin's values for the specific gravity of chlorine water---presumably sat. a t the temp. indicated-are Temp. Sp. g r. . . . . 2-5O 1'00406 8.0" 0-00494 16.3" 1.00424 23' 1.000264 24" 1-00069 25-5O 0-09984 T he value for water a t 2-5" was 0.99980; and a t 25.5", 0'994247. The specific1 gravity of bromine water was measured by J. Slessor 11 i n 1858, and for aoln. containing w grms. of bromine in 100 grms. of water, he found w . Sp. gr. 1 .072 1.00901 1.205 1.00955 1.231 1.874-1.906 1.952-2.009 1.01223 1*01491 1.01585 2.089-2.155 1.01 807 3.102-3.169 1'02367 The last-named soh. was sat. According to T. J. Baker,l2 the molecular heat of solution of chlorine in water is 4.97 Cals., and according to J. Thomsen, 4-87 Cals. M. Berthelot places this constant between 3 and 7-5 Ca1s.-the indefiniteness is due to a reaction between chlorine and the water. H. le Chatelier estimates that a nlol. of liquid bromine dissolving in water develops 1-08Cals. Prom the effect of temp. on the absorption coeff., 8, the heat of s o h Q of bromine vapour in water is Q=-RT2(d log /3/dT) ; or Q=-4*571TZ(d log BldT). For bromine vapour, Q, at 0' is 8.35 Cals., and a t 60°, 6.5 Cals. If A denotes the latent heat of vaporization, the heat of s o h . of a mol. of liquid bromine in water will be A+&, where A=7696--8.480 Cals. The values of the heat of soh. of liquid bromine are very small, and become negative a t about 40". S. U. Pickering found a mol. of bromine absorbed 1.508 CaIs. when dissolved in 2700 mols. of water. H. H artley and N. P. Campbell estimated the heat of soh. of iodine to be 5.09 Cals. a t 21.5" and 7.38 Cals. a t 50". G. N. Lewis and M. Randall calculate the free energy of formation of a mol. of an aq. s o h . from the solid a t 25" o r 298" K. is 3926 cals., or -R T log 0.0132. The diffusion coefficient 13 of chlorine i n a q. soh. containing 0.1 mol. per litre is 1.22 'grms. per sq. cm. per day a t 12" ; 0 -8 for bromine ; a nd 0.9 for iodine. F. A. Henglein has made observations on the vap. press. of aq. soh. of chlorine, bromine, and iodine. The molecular conductivity of aq. soh. of chlorine were found by A.A. Jakowkin 14 t o be rather large. If v denotes the number of litres containing a mol. of chlorine, the molecular conductivity is, a t 0" and a t 26" At high dil., therefore, the conductivity is not far from that of hydrochloric acid. A. A. Jakowkin's explanation turns on the assumption, for which there is much confirmatory evidence, that in aq. soln., the chlorine is hydrolyzed forming hypochlorous acid, HOC1, and hydrochloric acid : Cl2+H2O+HC1+HOCl, and the high conductivity, with highly dil. soln, shows that the hydrolysis is nearly complete. According to A. A. Jakowkin, about one-tenth of the chlorine in a sat. soln. a t 0" is hydrolyzed. The partition law.-If a,, b,, c,, . . . denote the concentrations of a gas dissolved in a given liquid and a2, b2, c,, . . . .;::e corresponding concentrations of thv gas in the space above the fiquid, Henry's law means that This law, applicable for the distribution of a soluble gas between a liquid solvent and the space above, can be extended to include the distribution of a solute between 75 T H E HALOGENS two immiscible solvents-e.g. iodine or bromine between water and ether, water and benzene, and between water and carbon disulphide ; a s well as silver or gold between molten lead and zinc ; f erric chloride between ether and water, etc. This fact was indicated by M. B erthelot and E. Jungfleisch l5 i n their paper: Sur les lois qwi prisident au partage d'un corps entre deux dissolvants. I n illustration iodine or bromine in a mixture of immiscible water and carbon disulphide divides itself so that the ratio of the concentrations of the solute in each layer is always the same. This constant is sometimes called the partition coefltlcient. B erthelot and Jungfleisch's results in Table IV i llustrate the principle ; t he concentrations of the halogens Iodine. Water. CH, 0.004 1 0.003 1 OaO016 0~0010 Bromine. / I - .---. Quotient. 1 -74 1.29 1.66 0.41 Quotient. Water. CSa in the solvent indicated in t,he first two columns of each table refer to the amounts of solute in 10 C.C. of solvent. Although the constants are not uniformly the same, the deviations are within the range of experimental error. Later measurements, by A . A. J akowkin and others, between water and carbon tetrachloride, bromoform, amyl-alcohol, or carbon disulphide made under better conditions, give smaller deviations. Similarly, when an aq. soln. of hydrogen peroxide is shaken with amyl alcohol, the peroxide divides itself between the two solvents so that the ratio of the concentration of alcohol phase to the conc. of aq. phase has always the same numerical value a t a given temp. Expressing concentrations in milligram-molecules per litre, H. T.C alvert (1901) f ound a t 25" : Amy1 alcohol phase Aq. phase Quotient . . . . . 13.4 9 4-0 7.0 28.0 193.5 6.9 41 '9 296.7 7.1 65.0 460.0 7.1 94.5 670.0 7.1 130.2 9125 7.0 The facts are generalized in the so-called Berthelot and Jungfleisch's partition law : When two immiscible solvents are simultaneously in contact with a substance soluble in both, the solute distributes itself so that the ratio o the concentrations o the f f solute in each solvent is constant. Otherwise expressed : Concentration of solute in solvent A Concentration of solute in solvent B E Constant, say b . If the ratio be unity, the concentrations of the solute in each solvent will be the same; if t he ratio be far removed from unity, a correspondingky large proportion of the solute will be found in the one solvent which can be utilized to extract the soh. from the other solvent. E.g. ether will remove ferric chloride from its aq. soln., and since many other chlorides are almost insoluble in ether, the process is utilized in analysis for the separation of iron from the other elements; the solubility of cobalt thiocyanate in ether is utilized for the separation of cobalt ; p erchromic acid ia similarly separated from its aq. soln. by ether ; molten zinc extracts silver and gold from molten lead ; t he extraction of organic compounds from aq. s o h . by shaking out with ether or other solvent is much used in organic laboratories. If w o grrns. of iodine be dissolved in unit quantity of water, and this be shaken up with unit quantity of immiscible carbon disulphide, it follows that a quantity w , will remain 76 l NORGANlC AND THEORETlCAL CHEMISTRY i n the aq. layer and wo-w, will pass to the carbon disulphide, so that wl=k(w,-w,), or w,=wok/(l+k) remains in t he aq. layer, and w,-w, passes into carbon disulphide. A second extraction with the same quantity of carbon dlsulphide gives w, =k(w, w ,), or, substituting the previous value of w,, w, =w,kaj(l + k)2 ; a nd generally, a'fter the nth extraction - remains in the aq. layer-the greater the number, n, of extractions, the smaller the quantity of substance remaining in the aq. layer. I t can be shown in a similar way that with a given quantity of the extracting Iiquid, a better separation is obtained after many extractions with small quantities of the liquid than by few extractions with large uantities. The extraction can never be theoretically complete ; the smaller the value of k t e greater the efficiency of the process. Thus, with 10 grms. each of iodine and bromine in the aqueous layer the second extraction will leave lO(O-O0244+1d0224)" or 0-00006 grm. of iodine, and 1O(0-0125 +l*Ol25)5 or 0.0016 grm. of bromine. The same principles obtain in washing precipitates. % Suppose that the molecu2es of the solute ammonia remain normal, NH3, in one solvent, say water; and that the molecules of the ammonia partially or completely polymerize in the other solvent, say chloroform, such that 2NE3$N2H6. Then, for equilibrium, Co=klC21, where C1 represents the conc. of the molecules of ammonia, N&,in water ; a nd Co represents the conc. of the molecules of ammonia, N2Hs, in the chloroform. Again, i f t he molecules of ammonia are polymerized only in the chloroform, and not in the water, then, Co=K1C12, or C,=dKICo. W. Herz and M. Lewy (1906) tested this hypothesis by finding how the formula so deduced fitted the facts. The fit was quite satisfactory; this was taken to prove that the ammonia molecule doubles on itself in this solvent. The partition law thus furnishes a method of measuring the relative mol. wt. of a solute in a gii.en solvent. If t he partition coeff. be not constant, it is inferred that the assumed molecular conc. is different from that which actually obtains. The constitution of c hlorine water.-As already indicated numbers ranging from C12.10H20 to C12.7H20 have been given for the composition of the hydrate. C. P. Schonbein,16 shortly after the discovery of hypochlorous acid, suggested that the chlorine in chlorine hydrate is present as hypochlorous and hydrochloric acids : HCl.HOCl.9H20, although no experimental evidence was given in support of the hypothesis. H. E. Roscoe sought for oxy-chlorine acids in chlorine water by passing a stream of carbon dioxide through the liquid, but found no signs of these acids in the more volatile portion. C. Gopner assumed that the presence of hypochlorous and hydrochloric acids in chlorine hydrate was proved by the fact that mercuric chloride is formed on treating the hydrate with mercury-Wdter's reaction ; b ut R. Schiff argued that if chlorine hydrate contained hypochlorous acid, it should be rapidly decomposed in diffused daylight; but, unlike hypochlorous acid, chlorine hydrate is comparatively stable under these conditions. The smell of dry chlorine hydrate with its 28 per cent. of chlorine, is not so marked as that of a 0'7 per cent. soln. of chlorine. This shows that the vapour tension of the chlorine in chlorine hydrate is very small. R . Schiff 'also argues that the pale colour of chlorine hydrate is against the formula HCl.HOCl.9H20 for chlorine hydrate. The work of A . A. Jakowkin shows that a fractional portion of the chlorine in an aq. soln. is hydrolyzed a.s suggested by C. I?. Schonbein (1847) and by N. A . E. Millon (1849). A. A. Jakowkin I T employed the partition coeff. between chlorine water and carbon tetrachloride for estimating the concentration of free non-hydrolyzed chlorine in a soln. of that gas. I f t he hydrolysis be neglected, and the assumption be made that the whole of the chlorine in the aq. soln. be present as dissolved free chlorine, the partition coeff. of chlorine between water and carbon tetrachloride changes from 13'8 t o 5 .2. This is taken to mean that something is wrong with the assumption that the total chlorine is free in aq. soln. A. A. J akowkin calculated the degree of hydrolysis : C12+H20=HCl+HOCl, from the conductivity measurements of the aq. s o h , deducted the amount of hydrolyzed chlorine from the total chlorine, and T H E HALOGENS 77 the remainder, when assumed t o represent the chlorine in equilibrium with that in the carbon tetrachloride layer, gave a satisfactory partition coeff.-mean=20.0as illustrated in Table V , where the concentrations are expressed in milligrammolecules per litre. TABLE V.-PARTITION COEFFICIENT O F CHLORINE C HLORIDE A T BETWEEN Total chlorine. Aqrieoris lager CC1,- layer WATER ND CARBONTETRAA 0'. Hydrolyzed chlorine in aqueons layer, Chlorine ns Cia in aqueous layerCJ1-s) 2 . Partition coefficient k SC L C,(l-z) C , The partition coeff, rises from 2 0-0 a t 0" t o 30.5 a t 28-6",a nd to 35'2 a t 57.5". T he hydrolysis attains a maximum a t about 90". Hydrochloric and hypochlorous acids are inappreciably soluble in the carbon tetrachloride ; a nd the proof t h a t the molecules of chlorine dissolved in the carbon tetrachloride are present as C12molecules turns on the fact that the partition coeff. of chlorine between chlorine gas and carbon tetrachloride is constant in accord with Henry's law. Thus, representing concentrations in milligram-molecules per litre : , ChIorine in air Chlorine in CCI, Quotient . . . . 0 .1109 8 .908 0 '0124 0-2666 22-46 0'0119 0.5365 44-14 0.0122 0.8800 '75.00 0.0117 The mean value of the partition coeff. is therefore 0 ,012. Hence, the chlorine in eoln. has the same molecular state a s chlorine gas. These observations also show that the solubility of chlorine in water does not follow Henry's law if the total chlorine i n soln. be considered, because only that portion of the chlorine which i~ not hydrolyzed is partitioned between the liquid and the space above. If allowance be made for this, the partition coeff. between chlorine water and air is 2 0 ~ 0 . 0 1 2 =Oe24. Chlorineg as a t one atm. press. contains , - * mol. per litre ; hence, if the partition coeff. be 0 .24, a l itre of chlorine water in equilibrium with chlorine gas at one atm. press. and a t 0" will contain &4~,+4=0-186 mol. per litre, or 1.32 grms. of free chlorine and 146 grms. of hydrolyzed chlorine per 100 C.C. T he solubility of chlorine in water is 0.089 mol. per litre, and of this 0 .025 mol. is hydrolyzed. At 25", t here will be 0.081 mol. of non-hydrolyzed chlorine per litre. The presence of hypochlorous acid in the aq. s o h . is shown by leading a current of mixed air and chlorine through a flask of water between 90" a nd 95", a nd condensing the products in a cooled flask. Hypochlorous acid is volatilized and condensed. S. U. Pickering 1s also showed in 1880 t hat when an aq. soh. of chlorine is boiled in an open vessel, a dil. s o h . of hydrochloric acid remains after the expulsion of the free chlorine, and A . Richardson (1903) h as shown that when chlorine water is distilled, the hydrochloric acid is eq. to the hypochlorous acid in the distillate. The presence of chlorides, hypochlorous, hydrochloric, and other acids diminish the I n consehydrolysis in virtue of the mass action law :. C12+H20+HCl+HOCl. quence, the apparent solubility of chlorine will be diminished, since less will be the result needed for establishing equilibr'lum : C12+H,0+HC1+HOC1-unless be obscured by a reaction between the acid and the chlorine such as must occur when the soh. contains over about a gram of HCI per 100 c:c. of hydrochloric acid. The partition coeff. of bromine between water and carbon tetrachloride, is so INORGANIC AND THEORETICAL CREMISTRY 78 nearly constant that the results are not much affected by hydrolysis. The greatest deviations occur with the more conc. soln. of bromine (over 3 grms. per litre). The low solubility of iodine prevents highly conc. s o h . being employed, and the partition coeff.between water and carbon tetrachloride is fairly constant. The concentrations in Table YI are in grams per litre of soh. I Bromine. 1 I Iodine. In CC14 Coeficient. - In water. Coefflclent. 0.2913 0.1934 0.1934 0.0818 0.0616 In water. 87-91 85.51 85.30 85-15 85-81 The electrical conductivity of bromine water measured by W. C. B ray 19 is also referred t o hydrolysis : B r2+H20+HOBr+HBr, and a negligibly small part to He made the hydrolysis constant the ionization of bromine itself : Br,+Br+Br'. A. ~ of bromine,water 2.4 ~ 1 ; A. 0 Jakowkin, that of chlorine water 4.48 ~ 1 0 - 4 . ~ According to the measurements of G. J ones and M. L. H artmann, the hydrolysis constant for iodine in water a t 0' approximates 9 X IO-15 ; a nd according to W. C. Bray, 0% x 10-12 a t 25'. It will be remembered that the hydrolysis constant is represented by K i n the equation : q Br2]=[Ha][Br'][HOBr], where the symbols in brackets represent the concentration of the substances concerned in the balanced reaction, H20+Br2=HOBr+H.+Br', and the concentration of the water in the system is so large that the proportion which is concerned in the reaction is negligibly sma!l. The value of K for bromine a t 25' is nearly the geometrical mean of the hydrolysis constant of chlorine and iodine : Chlorine. 4 -48 x H ydrolysis constant, K (25') Bromine. 2.4 x lo-$ Iodine. 0.6 x lo-'% A t 0" the hydrolysis constant of iodine is 9 xIO-15 T he hydrolysis in iodine water The specific electrical conductivity of is assumed to be I2+ H20+HI+HI0. iodine water rises very rapidly to d o u t 4 ~ 1 0 - 6 then slowly to about 20 x 10-6; and this last effect is supposed to be due to a secondary reaction involving: 312+3H20=5HI+HI03. The photochemical decomposition of chlorine w ater, bromine water, and iodine water.-In 1785, C . L . Berthollet 20 noticed that chlorine water is gradually decomposed by exposure to light, forming aq. hydrochloric acid and oxygen. He said : I h ave filled a flask quite full with dephlogisticated muriatic acid (chlorine water) and connected the neck of the flask with a neumatic apparatus by means of a tube ; on ex osure to light, I saw a large number of bub les of gas collect on all sides of the liquid, a n 8 after some days, I f ound in t he vessel, connected with the flask by means of the tube, a quantity of an elastic fluid which was the urest vital air (oxygen). As the vital air was developed f rom the acid in the flask, so di the liquid lose its yellow colour, and appear as clear as water. The liquid did not bleach blue vegetable colouring matter, but only turned i t red, and retained veIy little of the smell of dephlogisticated muriatic acid ; it effervesced with alkalies, and in a word, the dephlogisticated muriatic acid was nothing more than common muriatic acid. I n another flask, similarly fllled with the same fluid, and covered with black paper, the liquid suffered no change, and no vital air was developed. i B T he explanation, on H , Davy's chlorine theory, is embodied in the equation: 2H20+2C12+4HC1 O2 ; and, in agreement with W. H. Wollaston's observation + T H E HALOGENS that t,he chemically active or actinic rays are the most refrangible, H. D avy (1812) also noted that rays a t the violet end of the spectrum are more active in promoting the decomposition of chlorine water than the other rays. N . T. de Saussure (1790) also found that the rate of decomposition depends on the intensity of the light. He said in his memoir : E fets chimiques de la Eumidre sur une haute montagne compar6s avec les plaines : ceua qu'on observe d a m l plaines : a ... As soon a4 light impinges on dephlogisticated muriatic acid, i t is decomposed. Since the decomposition takes pIace gradually, and its veIocity, within certain limitations, is proportional t o the intensity of the light, I h ad a n otion in 1787 of trying if the quantity of oxygen which is developed in this reaction cannot be utiIized as a kind of photometer to measure the action of Iight. The observations have been confirmed by T. Torosiewicz (1836),21 who noticed the decomposition proceeds more rapidly in white than yellow glass ; b y J. W. D raper (1845) ; W. C. W ittwer (1855), etc. According to J. W . D raper, warming the chlorine water facilitates the action of light, but does not itself provoke the reaction ; a nd he states that " t he decomposition of water once begun in the sunbeams goes on afterwards in the dark," for a bulb of chlorine water which has been exposed to sunshine was placed in the dark and the quantity of gas given off during the first six hrs. was Gasevolved . 1st 0'0162 2nd 0.0159 3d r 0.0086 4th 5th 0.0060 0.0038 6th hr. 0-0031cub.in. and thenceforth for four days in diminishing quantities. He says the evolution of gas is not altogether due t o the gradual escape of oxygen formed while the liquid was exposed t o the sun, and held in a state of temporary s o h , nor to the decomposition of hydrogen peroxide or chlorous acid formed in the liquid. T he quantity given off in the dark depends on the intensity of the light t o which it was originally exposed, and on the time of exposure. The cause of the decomposition of insolated chlorine water in darkness was attributed to a change in the nature of the chlorine induced by exposure to the sun's rays. He says : Chlorine is one of those allotropic bodies with a double form of existence-active and passive, As comrnonIy prepared i t is i n its passive state ; b ut on exposure to the indigo rays, or other causes, i t changes and assumes a n active form. I n this Iatter state, its &nity for hydrogen becomes so great t h a t i t decomposes water without difficulty. The decomposition of chlorine water when placed in the sunbeam, adds J. W. Draper, does not begin a t once, but a certain space of time intervenes, during which the chlorine is undergoing its specific change. R. Bunsen and H. E. Roscoe 22 d o not consider that the modification which chlorine undergoes by exposure to light is so perskitent as J. W. D raper records ; possibly, said the latter, because " t he insolation t o which the chlorine was submitted was not continued sufficiently long, or perhaps the light was not sufficiently intense." There is no doubt that chlorine prepared under ordinary conditions can assume an active and a passive state, ride infra. P. Wohler 23 found that when chlorine hydrate in a sealed glass tube is exposed to sunlight, it forms two liquids, but does not decompose, since, after a summer's exposure, the two liquids re-form chlorine hydrate when winter returns. The same phenomena occur if the chlorine hydrate is warmed and cooled under similar conditions. A conc. s o h . of chlorine water is f ar less prone to decomposition on exposure to sunlight than is a more dil. s o h . J. M. E der found the same to be the case with bromine water, but he also found that a conc. soh. of chlorine water lost 53.95 per cent. of chlorine while a dil. s o h . lost 41.87 per cent. under similar conditions, but he does not state the concentration very exactly. A. P edler further showed that soh. more conc. than one mols. of chlorine with 64 mol. of water had not decomposed perceptibly after a t wo months' exposure to tropical sunlight, and with that inoreasing dilution, the action became progressively greater, as illustrated in Table Y II. 80 INORGANIC AND THE0R.ETICA-L CHEMISTRY E. Klimenko 24 h as confirmed the observation that dil. soh. of chlorine water decompose more rapidly than conc. s o h . ; a nd he has studied the influence of hydrochloric acid and various chlorides on the action of sunlight on chlorine confined in sealed glass tubes. 5. Billitzer found that the presence of small quantities of Mol. of water per mol. chlorine. Percentage of chlorine which had acted on the water. Time of exposme t o actual sunlight. 2 mths. 132 hrs. 137.1 ,, nil 29 46 29 78 hydrogen chloride increased the speed of decomposition but with larger amounts ; E. Klimenko observed the speed of the reaction to be retarded by the presence of hydrochloric acid and chlorides, and this the more with the chlorides of the alkalies than with the chlorides of the alkaline earths ; a nd the retardation is also great,er with the elements of a given group the greater the at. wt. of the metal. For example, with soh. containing a mol. of t he given salt per litre, and the same quantity of chlorine, the amount of chlorine remaining after the same exposure when referred t o a tube with hydrochloric acid as unity, was as follows-HC1 unity : LiCl 0 .3079 NaCi 0.1732 KC 1 0.0900 O% ?h SrCI, 0 .3022 3aCI , Oa28a6 MgCl, 0.530 ZnCl, 0.2004 CdCl, 0.042 The quantity of chlorine which does not take part in the decomposition is independent of the time of exposure. If a beam of light of intensity I i s changed by an amount -dI in passing through . a layer of fluid of thickness da, it is assumed that -dl is proportional t o da, and to I,, so that --dI=aIodx, where a is a constant. By integration, the intensity I of the light after it has passed a layer of fluid of thickness x, is I=Ioehz. ' The constant a depends on the nature of the substance and the wave-length of the incident beam of light. It i s called the coefllcient of absorption of the substance for the light in f question. I a beam of light traverses a thin layer of chlorine wa8terof concentration Co,,W. C. Wittwer, in his paper Ueber die Einwirkung des Lichte auf Chlorwasser (1855), showed that the amount of hydrogen chloride, dC, formed in the time &, is proportional to the concentration s of the chlorine water, and to the intensity lo of the incident light ; or dC=CIdt, that is, the concentration C of the chlorine at the end of the time t is C=Coe-It. I f t he time is constant, and the layer of liquid have a thickness dS, the light changes dl=-ICodS, or the intensity of the light.after it has traversed a layer S is I=IocaW a nd Io(l-e-acos) ; represents the Ioss which the light has suffered in passing a distance dS. Hence, the change dC in the conc. of the chlorine water, initially Co, when light of intensity I traverses a layer . of thickness S i n the time d2' will be dC=C12(1-racos)dt, an expression which W. C. Wittwer found to be applicable to this react'ion. By keeping the experimental conditions constant, this equation reduces t,o dC=kCdt, where k i s a constant; if a denotes the initial concentration of t he chlorine, and Sdx the amount decomposed in the time dt, dx=k(a-a)dt, and W. C . Wittwer's equation assumes the integrated form . I. a k = - log t a-x . This equation is valid for pure chlorine water when the conditions are such that the field may be regarded as uniform ; t he constant k, however, increases as the T HE HALOGEN8 8'1 chlorine is consumed because the field of illumination becomes more uniform the more dil. the s o h . I chlorides are present in the soln., H. Tuchel assumed that f part of the chlorine in soln. is present as trichloride, and this portion of the gas k I denotes the f not directly concerned in the reaction : 2E120+2C12=4HC1+02. initial concentration of the chlorine directly concerned in this reaction, the preceding equation reduces t o the form : which is applicable to the decomposition of soln. of chlorides in water, when a-4' r epresents the amount of chlorine which does not take a d irect part in the decomposition, 2H20+2C12=4HC1+02. Only for pure chlorine water does a=,$. According to C. P. Barwald and A. Monheim (1835), the decomposition is accelerated by the presence of organic substances. J. Milbauer tried the effect of thirtytwo metal chlorides ; of sodium tungetate and molybdate ; of uranyl sulphate ; a nd o sulphuric, selenic, arsenic, and boric acids on the photo-decomposition of chlorine f water, and found. that none accelerated but that most retarded the action. Chlorine catalyzes the decomposition of bromine water; and bromine, chlorine water ; while iodine does not accelerate, but rather retards the reaction, probably by forming relatively stable iodine compounds. A. Benrath and R . Tuchel found the temp. coeff. of the velocity of the reaction with chlorine water between 5" and 30" increases in the ratio 1 : 1 .395 per 10". The hydrolytic reaction between watcr and chIorine results in the formation of hypochlorous acid : H2O+Cl2=HCI+HOC1 ; a nd in feeble diffused dayIight, the hypochlorous acid in decomposed into chloric acid and oxygen : 3HOC1=2HCI +HC103, according t o E. Klimenko ; or 8HC10=2HC103+6HC1+02, according to A. Pedler,Z5 so that chloric acid is always found among the produch of the decomposition. Aq. soln. of chloric acid alone are not decomposed by light ; b ut thin does not apply with chloric acid in the presence of hydrochloric acid. G. G-ore found the electromotive force of chlorine water decomposing in light gradually diminishes to a minimum value when the soln. contains hydrochloric, chloric, and hypochlorous acids ; on further exposure, the electromotive force increases slowly until the soln. contains nothing but hydrochloric acid and hydrogen peroxide. Hence, says G. &re, these are two essentially different phases in the reaction-(i) the formation of chlorine acids ; ( ij) t he decompo%itionof these acids into hydrochloric acid and hydrogen peroxide, but the presence of the last-named compound ha.^ n ot been conhmed. A . P opper could detect no perchloric acid in the products of the reaction ; b ut J. Billitzer has reported the presence of traces of this acid. According to A. PedIer, the more conc. the soln. of chlorine water, the more nearly do the products formed by the action of light approach the values indicated in the equation : 2H20+2C12=4HC1+02, but as the amount of water is increased, the proportion of chloric acid in the product increases. The side reactions which occur in the decomposition of insolated chlorine water, and the effect of the various products of decomposition on the speed of the change, render the reaction, 2H20+2C12=4HC1 +Oz, a n unreliable foundation for the construction of such a photometer as wan suggested by N. T. d e Saussure (1790) to measnre the intensity of light in terms of the volume of oxygen evolved ; similar remarks apply t o W. C. Wittwer's photometer which was based on the decomposition of a 0 .1 to 0'4 per cent. soln. of chlorine water. R. Bunsen and H . E. Roscoe also said that the many sources of error involved in the measurement of the intensity of light by the chlorine water photometer make the results of no value. According to C. Lowig,27 bromine water in light behaves in a similar way to that o chlorine water, but as J . M. E der showed, bromine water is much less sensitive f to light in that it decomposes with but one-sixth or one-twelfth the speed 0.f chlorine ' " VOL; TI. G 82 INORGANIC AND THEORETICAL CHEMISTRY water. The presence of tartaric or citric acid accelerates the decomposition of chlorine or bromine water in light. According to C. P. Cross and A. Higgin, water is not decomposed by bromine or iodine if heated eight days in a sealed tube a t 160°, b ut if a metal salt be present like lead acetate, the bromide or iodide is precipitated. H. W. Vogel states that iodine water is stable in light. According to H. Bordier, X-rays, like light, decolorize dil. aq. soh. of iodine or starch iodide ; a nd the X-rays effect a change in a few minutes which requires several hours with ultraviolet light. The solubility of the halogensinacid and salt solutions.-Thesolubilityof chlorine in various salt soln. has beendetermined by W. L . Goodwin,28E. G.Kumpf, C.A.Kohn and F. O'Brien, etc. W. L. Goodwin used the chlorides of lithium, sodium, potassium, calcium, strontium, barium, magnesium, cadmium, iron(ic), cobalt, nickel, and manganese, and found the solubility is increased by the presence of hydrogen chloride, but is in general decreased by the presence of the other chlorides. The decrease is very marked with the conc. soln. of sodium chloride. The solubility of chlorine in sulphuryl chloride, S02C12, has been studied by H. Schulze ; i n chromyl chloride, Cr02C12,by H. W. B. Roozeboom ; in soln. of sodium chloride, b y E. 0. Mandala ; a nd in arsenious chloride, AsC13, by B. E. Sloan, Some results a t atm. press. arc indicated in Table VIII. 1 Temperature. Water. Solubility coefficients. Hydrochloric acid. -- I n 20 - per cent. 8p. gr. 1 .046. Sp. gr. 1'125. I1 1 11'41 per cent. LiCl. Sodium chloride per cent.. --- - I - 0" 5 ' 1o0 15' 20' 25O 30' 40' 50' so0 -- 2 .8 2.0 2.7 2 .6 2-3 2-06 2 .4 2.4 2.0 1.6 1.4 1-3 1.8 1.35 1.0 - - - W. L. Goodwin carried the solubility curves down to temp. a t which chlorine hydrate was formed, and he noticed (1) t hat the solubility curve increased with rising temp. to about lo0,a s in the case of chlorine in water ; a nd (2) t hat the presence of chlorides in soln. lowers the temp. a t which the maximum appears, and also the f temp. a t which the hydrate separates. Otherwise expressed, the presence o chlorides in soh. hinders the formation of chlorine hydrate. The increased 80111bility of chlorine with an increase in the conc. of the hydrochloric acid may be noted, H. E. Roscoe 29 found that the coeff.of absorption of chlorine was lowered from 2.39 t o 1.98 a t 14" by ii$ih part of hydrochloric acid. M. Berthelot found that the absorption coeff. was raised i n presence of hydrochloric acid, for a 38 per cent. soln. of hydrogen chloride absorbs 17'3 grms. of chlorine per litre ; a 33 per cent, soln. 11 grms.:, and a 3 per cent. soln. 6 grms. of chlorine per litre. A t 21' t he solubility of chlorme in a dif. hydrochloric acid decreases with increasing conc. of the acid, to 1 .50 with a 0 .94 per cent. of soh. of HCI. This corresponds with the general behaviour of chlorine in aq. soh. of the other chlorides. After this, the solubility curve of chlorine almost doubles on itself, for the solubility of the chlorine now increases with increasing conc. of acid up to a solubility of 38.9 with a 3 1.2 per cent, soln. of HCI, The increase was attributed by J. W. Draper (1843) t o the possible THE H ALOGEX3 F3 formation of n bichloride of hydrogen, and by M.Berthelot (1881) to the formation of hydrogen perchloride, HC13. This was confirmed by the fact that the heat of soh. of a mol. of chlorine in hydrochloric acid is much greater than in pure water-with water, the heat of soln. is 7.5 Cals., and with HC1+4.5H20, 9.4 Cals. M . Berthelot's hypothesis is the most probable explanation of the phenomenon, although them is nothing in the experimental data t o show that particular polychloride is formed, This hypothesis is also in line with the behaviour of the other halogens in s o h . of the corresponding alkali halide, or haloid acid. The general effect of the presence of a salt in soln. is to reduce the solubility of bromine,sO and this the more the greater the conc. of the salt. For example, with a !EN-soh. of sodium nitrate, 3.374 grms. of bromine were dissolved per 100 LC., and with a N-soln., 2.88 grms., while the solubility of bromine in water a t 25" is 3-395 grins. per 100 c.c. The effect of a few typical salts, in grms. per 100 c.c., on the solubility of bromine-grams per 100 c.c.-in water is as foilows : Salt , Bromine NaNO, , , 9 .118 2 .48 8.509 2.80 NaCl (NH,),SO, 5.860 5.590 7.004 7.77 C H3.COONH4 7.709 34.05 H,SO, 4.903 2.936 The increase in the solubility of bromine in soln. of ammonium salts is very marked, as is also the case with the aIkali chlorides. The case with the alkali bromidej is specially interesting. The solubilities by F. P. Worley are indicated in Table I X. The marked increase in the solubility of bromine in s o h of potassium bromide was attributed by M. Roloff to the formation of moIecules of KBr,. He shook up n soln. o f bromine in carbon disulphide with water and with an aq. soln. of p o t a s s i p bromida, and measured the concentration of the bromine in the two layers. M. Wildermann has shown that the density of bromine vapour over a soln. of potassium bron~idcsat. with bromine is the same as over water sat. with bromine,, indicating that tho conc. of the free bromine in all the aq. soln. is the same, and a ny excess in the presence of potassiunl bromide must be united with the potassinm bromide. A11 t he bromine dissolved by a soln. of potassium bromide can bc removed (F. P. WORLEY). Bromine dissolved per litrs. Potassium bromide per litre of solution. Mcl. - -- - 0 0.02 0.04 0.06 0.08 0.10 0.20 0-40 0.60 0.80 Ou90 Grms. Grms. -- -- - .. ----- Grms. ---- . - 0 2.1 8 4-38 6-55 8-76 . 10.91 21 -82 43-82 65.46 87.64 98.19 2 6.50 38.56 41-91 45-31 4 8-44 52.23 69-69 104.90 141.60 178.70 198.70 by successive extractions with carbon disulphide, or by a stream of air. Hence it must be assunled that the polybromide formed is stable only in the presence of free bromine. For eq~dlibrium, KBrn+l+KBr+inBr2, and that there is a progressive dissociation of the polybromide, KBr,,+l as the conc. of the free bromine is diminished. Table I X shows that i f t he amount of bromine dissolved by the water 84 INORGANlC AND THEORETICAL CHEMISTRY be assumed constant, and the result be subtracted from that actually dissolved, the increase in the solubility of bromine is directly proportional to the amount of potassium bromide in the more dil. soh. Hence it is probable the whole of the potassium bromide in soh. unites with bromine to form the polybromide. The effect of a variation of temp. is negligibly small, and consequently it is inferred that the polybromide is not appreciably dissociated a t these temp., since it is not likely that if the polybromide is partly dissociated, the degree of dissociation will be independent of temp. With these assumptions, the amount of bromine dissolved by the more dil. soln. of potassium bromide corresponds with the amount required t,o convert all the dissolved potassium bromide, KBr, into potassium tribromide, KBr3. The still greater solubility of bromine in soln. containing over 0.1 mol. of potassium bromide per litre is taken t o indicate a tendency to the formation of a st,iIl more complex polybromide, say, the pentabromide, KBr5-vide 5 11in the next chapter. J. M. Bell and M. L. BuckIey found the solubilities of bromine (n gram-atoms per litre) in soln, of sodium brornidc containing w grms. per Iitre, a t 25" : w n . . Sp. gr. . 92.479 . 2.6 .. 319.7 359.0 4083 205'8 255.8 160.5 13-65 16-04 20-85 6.195 8.575 4.345 2.420 1,997 2.137 1.515 1.678 1.372 1 -213 A . P. Joseph gave for the solubility of bromine in aq. soln. of potassium bromide or rather the converse, in grams per 100 grms. of water a t 32.4" : w n . 0 Sp. gr. 750'2 704'3 771'1 801'3 8459 740'7 733.9 137.2 229-7 382.1 74.3 120.9 39.6 2 4.0 1.4753 1'5236 1.5980 1.4356 1.4633 1.4132 1.4063 13917 7 25.6 A . F, Joseph found that the solubility of potassium brornidc in water is increased by the addition of bromine. For moderate conc., about half a mol, of bromide is disqolved for each mol. of bromine added to the tvater. T his corresponds with that required for the formation of potassium tribromide. The limit of the soIvent capacit,y of watcr was not reached with s o h . containing over 2000 grms. per Iitre. There is, however, a maximum vol. conc. of bromide which corresponds with between 200 a nd 2000 grms of bromine per litre. F IG. 21 .--F. P. Worley's Apparatus, By assuming a definite formula for the polybromide attempts have been made to show that the relative proportions of the substance concerned in the reaction are in accord with the law of mass action. For example, with the polybromide KBr.nBr2 of conc. C, where the dissociation products are KBr of conc. cl, find nBr of conc. c2, the equilibrium condition is that KC=clc2fi, where K i s the equilibrium constant. I t he soln. are sat. with bromine, and the conc. of the free bromine in the soh. is f constant, KfC1=q, meaning that the conc. of the polybromide is proportional to the uncombined potassium bromide, and the constancy of the equilibrium constant is independent of the value of n. If t he soln. are not sat. with bromine, the equation I<C=clczn can be applied when the relative amounts of free bromine and of bromine combined as polybromide have been determined. No analytical process involving the removal of bromine from the system is applicable since the polybromide is decomposed as fast as the free bromide is removed. P. P. WorIey obtained the desired data by simuItaneously shaking, without mixing, two s o h of bromineone in pure water, one in a soln. of potassium bromide of known conc. in an apparatus, Fig. 21, with a common vapour space for the two soln. When the system is in equilibrium, it is assumed that the conc. of the free bromine in the two soln. is the same, since each is in equilibrium with the same vaponr space. The determination of the free bromine in the water is made by direct titration. T HE HALOGENS 85 The results agree better with the assumption that n = l t han n=2, although in neither case is the value of K constant. This is taken to mean that a small amount of a polybromide, more complex than KBr3, is also formed. G. J ones and M. L. H artmann's measurements of the electrical conductivity of aq, soln. of bromine in water and soln. of potassium bromide are interpreted to mean that bromine dissolves as Brz; that this is followed by a reaction B r 2 + H 2 0 + ~ B r 0 + He+Br' ; a nd simultaneously by KBr +Brz+KBr3 ; a nd by 2Brz+KBr+KBr5. A s at. soln. of bromine in water a t 0" has the composition in mols. per litre : 0'2539 Br2 ; 0'001085 H e; 0*000126Br' ; 0'00628 Br3' ; a nd 0 '00331 Br5'. The equilibrium constant K in K3[Br3']=[Br'][Brz] is O.C?51; a nd in K5[Br5']=[Brf][BrA2. The hydrolysis constant K ' in K[Br2] =[H*][Br'][HOBr] is 5.7 x 10-10. E . 0. Mandala measured the solubility of bromine in hydrochioric and hydrobromic acids The solubility of iodine in aq. s o h , of various salts 31 is somewhat similar to the behaviour of these salts on the solubility of bromine. The solubility is reduced by sodium and potassium sulphates and nitrates ; i t is raised a little by potassium and sodium chlorides, and considerably by potassium and sodium bromides, ammonium saltsalso increase the solubility of iodine ;boric acid has but little influence. C. K raus found the solubility of iodine is considerably augmented in the presence of hydrochloric acid ; phosphoric acid dissolves iodine slowIy in the cold and rapidly when heated, and similarly also in acetic, tartaric, citric, and tannic acids. The solubility of iodine in conc. sulphuric acid (over 83 per cent.) is about 6.6 grms. per litre; the colour is violet, which becomes yellow on dilution until, with 42 per cent. acid, the colour is brown. According to W. Vaubel the absorption spectrum of the dil. soln. shows red, yellow, and green, but not blue absorption bands ; with the yellow soh. there is no green band. A . H antzsch and A . Vagt explain this by assuming a brown hydrate of iodine is formed if sufficient water be present, and when the conc, of the acid is great enough no brown hydrate is formed and the anhydrous iodine shows its characteristic violet colour. Soln. in nitric acid behave similarly. The solubility of iodine in potassiunl iodide soln. follows the same general character as that of bromine in soh. of potassium bromide, but the effects produced are more marked. This is illustrated in Table X. Potassjum i o d l d o p s r lltrc. Piiiigram-molecules. 0 0 .830 1.861 3.322 6 .643 13'29 26.57 53.15 106.30 1- Iodins-per Grams. Miljigram molecules, 1, I litre. - Grams. . 0 f . 37 2-75 6 '5 1 1 1-'03 2 2-07 4 4.15 8 8.30 176-60 . - A soln. of " twenty-two grains of iodine and thirty-three grains of iodide of pota.ssium, in one ounce of distilled water " forms the liquor i odi of the British Pharmacopoeia. The effectsproduced by the ammonium salts are attributed to their hydrolysis into ammonium hydroxide, and the consequent, formation of ammonium iodide or polyiodide. The effects I;roduced by soln. of the halide salts are doubtless due to the formation of polyiodides as in the analogous case with bromine and potassium bromide. A . A . J akowkin allowed carbon disulphide to remain in contact with aq. soln. of iodine and potassium iodide until equilibrium was attained ; aiid 86 INORGANIC AND THEORETICAL CHEMISTRY assumed that any free iodine always divides itself between the water and carbon disulphide in the ratio 1 : 410, a s is the case if potassium iodide be absent. Assume I x mol. of free f that in the aq. soln. there is the balanced system : K13+KI+12. iodine be contained in unit volume of the potassium iodide soh. which contains a mol. of potassium iodide, there wiIl be present 1-x mol. of the complex K&, and a-(1-x) mol. of potassium iodide. Consequently, by the law of mass-action (1-x)K=(a-l+x)x, A. A . Jakowkin was able t.o calculate t.he equilibrium constant K, and found the results in harmony with this assumption, fcr K varied between 1577 x 10-6 and 1808 x 10-0 ; whereas with the assumption t,hat the system is K15=KI+212, K varied from 10,180x 10-0 t o 461 x 10-6. The argument dues not apply to conc. soln. of potassium iodide, where there are indications of higher polyhalides than K13. With iodine and potassium bromide, mixed polyhalides are formed-c.g. potassium iodobromide, KBr12. The effect of chlorine on the alkali chlorides is thus quite different from the e8ect.s of the other two halogens on their alkali salts ; a nd less than the effect of the other two potassium halides on the eolubility of iodine. The solubility of iodine in water is 0.340 grm. per litre a t 25"E. 0 . Mandala and A . Angenica give 0.0334 per cent., or 0'00131 mol. per litre a t 25" ; with a normal soh. of pot~ssium chloride, the solubility is 0'658 ; with a normal soln. of potassium bromide, 3.801 ; a nd with a normal soh. of potassium iodide, about 14 grms. per litre. E. 0. Mandala and A. Angenica measured the solubilities of iodine in aq. soln. of various concentrations of hydrochloric acid, hydrobromic acid, and hydriodic acid, and found it to be equal to that in the soln. of the corresponding potassium salt of the same conc. Hence, the solubility of the iodine is specific to the halogen ion, and is independent of the nature of the positive ion of the halide. The f.p. of soh. of hydriddic and hydrobromic acids are not altered by the dissolution of iodine. Iodine and bromine are fairly soluble in arsenic chloride, AsCI3 ; 32 100 grms. of this compound dissolve 8-42 grms. of iodine a t 0°, 11-88grms. a t 15", itnd 36'89 grms. a t 96". A l ittle iodine and bromine dissolve in sulphur$ chloride, and the soln. conduct electricity. P. Walden explains the conductivity of, say, iodine by assuming the molecule of iodine is ionized into a positively and negatively charged iodine atom : 12=I,+I'. Iodine and bromine also form conducting soln. in liquid sulphur dioxide, and the soln. of iodine is violet ; according to J . Inglis, the soln. of iodme in sulphur chloride, S2C12,is a conductor of electricity, and according t o E. Solly, a non-conductor. Iodine dissoIves in liquidammonia forming, according toE. C. Franklin and C. A. Kraw, a series of substitutionderivatives, N&I and N13.nNH3,where n stands for 1, 2, or 3. T he colour changes according to C. Hugot from black to red, to pale yellow. According t o U. Antony and G. Magri Iiquid hydrogen sulphide forms a dark red soln. without perceptible reaction. F. Sestini found iodine soluble in liquid sulphur B trioxide to the extent of about 200 arms. per litre ; a nd E. 1%. iichner that iodine dissolves in liquid carbon diozide to the extent of 5 per cent., wlzil e bromine is slightly soluble in the same.menstruum. According t o B. D. Steele, D. McIntosh, and E. H. Archibald, bromine dissolves in liquid hydrogen chloride and raises its electrical conductivity. The solubility of the halogens in organic solvents.-L. Bruner 33 h as measured the solubility of iodine in mixturcs of ethyl alcohol and water a t 15" : Alcohol Iodine . . 10 0 .05 20 0.06 40 0.26 60 1-14 80 4-20 DO 7.47 100 15.67 peroent. ,, ,, N. Schoorl and A . Regenborn say that owing to the ready formation of hydrogen iodide, accurate determinations of the solubility in aq. alcohol can be made only by making a sat, soh. of iodine in absolute alcohol, diluting the soln., and determining the iodine a t once. The sohbility follows a fairly regular course : Percent.alcoho1 Per cent. iodine . . 1 00 20 95 14.8 90 11.4 80 7-2 60 2.3 40 0.55 20 0.08 10 0.045 0 0.025 87 T HE HA.I,OGENS When the amount of ethyl alcohol exceeds 18 per cent., the addition of watcr causes a precipitation of iodine, at lower a!cohol conc. there is no precipitation. The maximum precipitation occurs when j ust enough water is added to bring the alcohol conc. to 18 per cent. The tinctura iodi of the British Pharmacopoeia is a soln. of " half an ounce of iodine, and a quarter of an ounce of potassium iodide in a pint of rectified spirit." P. Wantig found the mol. ht. of soln. -1,941 Cals., and S. U. Pickering -1.714 per 880 mol. of ethyl alcohol. C. Lowig found that alcoholic tincture of bromine is slowly decomposed in darkness, rapidly in light. Alcoholic soh. of iodine, according to H. E. B arnard, are stable in light and in darkness, but according to J . M. Eder they decompose 1000 times more slowly than chlorine water under similar conditions; T . Budde has shown that hydriodic acid, acetic ester, and aldehyde are formed, and the electrical conductivity of the soln. increases. J , 8 .Mathews and E. H. Archibald and W. A. P atrick found a freshly prepared N-soln. to have an electrical conductivity of 2.4 x 10-6 reciprocal ohms ; a nd a sat. soln., 1'61 x 10- +eciprocal ohms at 25". The decomposition is accelerated by the presence of platinum. The heat of s o h decreases with concentration from -7.92 to -7.42 cals. respectively for dilute and sat. soln. in methyl alcohol, and likewise from -4.88 to -5'22 cals. for similar soln, in ethyl alcohol. The solubility of iodine in aq. soln. of propyl alcohol is not very different from that in ethyl alcohol. H. Arctowsky has measured the solubility of iodine in benzene, chloroform, ether, and carbon disulphide-in some cases a t very low temp. At -83", 100 grms. of a sat. soln. of iodine in ether contain 15.39 grms. of iodine, and at -1OSo, 15.09 grms. 8. U. Pickering found the mol. ht. of soh. -1.536 per 800 mol. of ether. The following results are expressed in grams of iodine dissolved per 100 grms. of the sol?. P. Walden has measured the electrical conductivity of ethereal soh. of iodine ; t he conductivity increases with time, probably owing to chemical changes. According to J. Traube, iodine reacts with both ether and alcohol if heated in sealed tubes ; and according to D. McIntosh with ether, forming C4Klo0.C12 at -51". The solubility of iodine in benzene is : Iodine .. 4 '7" 6'6' 10'5O 8 -08 8-63 9.60 13'7" 10.44 16'3' 11-23 per-cent. and the mol. ht. of soln. is -4,681 Cals. according to P. Wantig, or -6.114 CaIs. per 1200 mol. of benzene according to S. U. Pickering. The viscosity of a soln. of bromine in benzene is 0.00737 a t 12". The diffusion constant of iodine in benzene is 1.41grm. and of bromine in benzene 1-75per sq. cm. per day. L. Bruner noticed that soln. of iodine in nitrobenzene conduct electricity, but not if all is thoroughly dried with phosphorus pentoxide. H. M. Dawson found that 100 c.c. of nitrobenzene dissolved 5'077 grms, of iodine. A . F. Joseph measured the solubility of potassium bromide in soln. of bromine in nitrobenzen?. Inchlorofornz under similar conditions: at -73.5") the solubility of iodine is 0.080; at -60°, 0'129 ; a t -49", 0.188 ; and, according.to W. Duncan, a t lo0, 1.77, or one part of iodine required 5-66parts of chloroform for soh. According to E. Beckmann and P. Wantig, in the proximity of the f.p, of chloroform, -61°, almost a11the iodine separates from the soln. and only Om0164per cent. remains. A . Kantzvch and A. V agt give for the solubility in chloroform a t 0°, 1.96 ; a t IS0, 3.78 ; a nd at 30°, 5.56 grms. of iodine per 100 grms. of soln. The mol. ht. of soln. is -5.484 Cals, according to E. Wantig, or -6.014 Cals., according to 8. U. Pickering, per 740 mol. of chloroform. A. A. Jakowkin gives for bromoforrn a t 25") 18.955 grms. of iodine per 100 c.c. of the solvent. The solubility of iodine i n carbon disulphide, which didifies at --116", is : - 100" -20' 0" l o0 20" 25" 30" 42' Iodine , 0'32 1-14 7.89 10.51 14.62 1 6-92 19'26 26'75 per cent. A. A. Jakowkin's number a t 25' is 23-0 grms. of iodine per 100 C.C. of sat. soln. corresponding with 15'4 grms. of iodine per 100 grms. of soh. H. Arctowsky found INORGANIC A ND THEORETICAT, CHEMISTRY 100 grms. of a sat. soln. of bromine in carbon disulphide contained 45.4 grms. of bromine a t --95", 39.0 grms. a t -110-so, and 36.9 grms. a t -116". 6 . U. Pickering, P. Wantig, and J. Ogier obtained the values -5.008, -5.241, and -4.8 Cals. for the mol. ht. of soln, of iodine in 12.780 mol. of carbon disulphide. The heat of soh. of iodine in carbon disulphide decreases from -17.65 to -16.65 cals. respectively for dil. and sat. soln. J. C. G. de Marignac found the sp. ht. 0.219 and 0.228 respectively for soh. of a gram-atom of iodine in 10 and 20 mol. of carbon disulphide. The soh. of bromine follow the additive law. The diffusion constant for iodine in carbon disulphide is 2.55 and of bromine 3-11grms. per sq. cm. pcr day. The viscosity of a sat. soln. of bromine in carbon disulphide is 0.00378 (16"))and of iodine O.00378 a t 16". According to A. Karion, the refractive index of a 0.2 per cent. soln. is 2'074 for the D-line, and 1.982 for the G-line, if the soln. follow the additive rule. There is anomlous dispersion. The absorption spectrum of a soln. of bromine in carbon disulphide is not that which characterizes liquid bromine. The soln. are nonconductors of electricity. The colour of the soln. of iodine like the vapour is purple in thin layers, blue in thick layers. A s at. soln. of iodine in methyl iodide has a sp, gr. of 3.548 a t 23" ; t he solubility increases with rise of temp. and the mixture appears to be completely miscible above the m:p. of iodine. The absorption coeff. of chlorine in carbon tetrachloride, is 83 a t 0°, and as measured by W. J. J ones a t 15", 51.7. According to L . B runer and H. Arctowsky, the solubility of iodine in carbon tetrachloride a t 14*B0is 20'6 grms. per litre, and a t 25") 30'33 grms. per litre. The soln. do not conduct electricity appreciably, and according to 8. U. Pickering the mol. ht. of s o h . is -5.782 per 700 mol. of water. According to W . Herz and M. Knock, the solubility of iodine in an aq. s o h of glycerol of sp. gr. 1-2555a t 25" (water a t 4" unity) and containing about 1.5 per cent. of impurities, is : Glycerol Iodine Sp. gr. . .. .. 0 0-0304 0.9979 7 -15 0.0342 1.0198 20.44 0.0482 1.0471 40.5 5 0.0875 1.0995 69.2 0,278 1.1765 100 per cent. 1.223 , ,, , 1'2646 P. JValden3.l has studied soh. of iodine in acetaldel~yde,Aydrazine hydrate, and ncetonitde; J . R.Mathews, soln. of iodine in pyritdine, ethyl, allyl, and phenyl isothiocyanic esters, and phenyl isocyanate; while H. 'A. Allen has studied soln. of bromine and iodine in various oils. REWERENCES. J. L. Gay Lussac, Ann. Chim. Phys., ( 2), 70. 426, 1839 ; (3), 7. 113, 1843 ; J. Pelouze, ib., (3), 7. 176, 1843 ; F. Schonfeld, Liebig's Ann., 95. 8, 1856 ; H. W. R. Roozeboom, Rec. Trav. Cl~im. Pays-Bas, 3. 59, 1884 ; 4. 69,1885 ; Zeit. phys. Chem., 2. 452, 1888 ; L. W. Winkler, Math. Tcrmeszettadomanyi Ertesito, 25. 86, 1907 ; H. Landolt and R. Bornstein, Physi!caIisch-chemische Tabellen, Berlin, 097. a L. W. Winkler, Chemn.. Ztg., 23. 687, 1800 ; Zeit. phys. Chem., 55. 344, 100G ; M. Wildermann, ib., 11.413, 1893 ; flr. H. McLauchlan, ib., 44. 617, 1003 ;K. W. B. Roozeboom, i b., 2. 452, 1888 ; Rec. Trav, Chim. Pays-Bas., 3. 73, 1884 ; 4. 71, 1885 ; W. Dancer, Jozwn. Chem. Soc., 15. 477, 1862 ; F.'Dietze, Pharm. Ztq., 43. 290, 1898. H. E. Roscoe, Joum. Chem. Soc., 8. 15, 1856 ; Liebig's Ann., 95. 357, 1885 ; R . Runsen, db., 93. 1, 1855. F. Dietze, Pharm. Ztq., 43. 290, 1898 ; L. Dossios and W. Weith, Zeit. Chem., 12. 378, 18G9 ; J. C . Wittstein, Pharm. Vierteljahrsschr., 6. 201, 1857 ; W. H. McLauchlan, Zeit. phys. Chem., 44. 617, 1903 ; A. A. Jakowkin, i b., 18. 590, 1895 ; A. A. Noyes and J. Seidensticker, ib., 27. 359, 5 1898; W. Meyerhofer, ih., 2. 594, 1888; W. Herz and M. Knoch, Zeit. awrg. Chem., 4 . 269, 1905; P. P. Fedoteeff, db., 69. 22, 1911 ; J. Meyer, db., 30. 113, 1902 ; H. Hartley and N. P. Campbell, Journ. Chem. Soc., 93. 741, 1908 ; R . Luther and G. V. Sammet, Zeit. Elektrochem., 11.293,1905 ; W. C. BraypJourn. Amer. Chem. Soc., 32.932,1910 ;G. Jones and M. L. Hartmann, 6. 37. 241, 1915. 3. J. Pelouze, Ann. Chim. Phys., (3), 7. 185, 1843 ; M. Cournaux and P. Guigues, J O U ~ P fimm. Chim., f5),23. 238,1891 ; A. Ditte, Compt. Rend., 95. 1283. 1832. G. Lunge, Handbuch der Soda-lndwtrie, Braunschweig, 3. 271, 1896. H. Davy, Phil. T ~ a m . ,101. 155, 1811 ; M. Faraday, Quart. Journ. Science, 15. 71, 1823 ; 11. W. b. Roozeboom, Rec. Trav. China. Pays-Bas, 3. 29, 59, 73, 84, 1884; 4. 216, 1886; 5. 350, THE HALOGENS 1 886; Zeit. phys. Chem., 2. 452,1888 ; R. d e Forcrand, Compt. Bend., 133.1301,1901 ; 134. 991, 1901 ; H . G iran, ib., 159. 246, 1914; C . Lowig, Pogg, Ann., 14. 114, 1828; 10. 376, 1 829; G. MaumenB, Bull. Soc. Chim., ( 2 ) ,39. 397, 1883. E . R icwend, Journ. p a k t . Chem., ( I ) , 1 5. 440, 1838; F. W ohler, Liebig's Ann., 8 5. 374, 1853 ; W . A lexejeff, Ber., 9. 1025, 1876. F. I sambert, Compt. Rend., 86. 481, 1878. l o P. Villard, Jourfl. Phys., ( 3 ) ,5. 453, 1897 ; Chem. News, 78. 297, 309,1898. l 1 W . L. G oodwin, Trans. Roy. Soc. Edin., 30. 597, 1882 ; J . Slessor, Edin. P M . Jou.m., 7 . 287, 1858. l2 T . 3 . R aker, PTOC. Roy. SOC.,68. 9, 1901 ; J . Thornsen, Thermochemiscl~e Untersuchungen, Leipzig, 2. 26, 1882 ; H . le Chatelier, Compt. Rend., 100. 50, 441, 1885 ; H . Hartley and N . P, Campbell, Journ. Chem. Soc, 93. 741, 1908 ; S. U. P ickering, ih., 53. 865, 1888 ; M. Rerthelot. Ann. Chim. Phys., (51, 5. 318, 1875; G. N. L ewis and M . R andall, Journ. Bmer. Chem. Soa., 2.. 2259. 1914. -6 I S H . E uler, Ivied. Ann., 63. 273, 1897 ; G. H iifner, ib., 60. 134, 1897 ; Z ed. phys. Chem., 2'7. 237, 1898 ; F . A. H englein, Zeil. Eleklrochem., 26. 431, 1920. l 4 A A. J akowkin, Zeil. phys. Chem., 29. 613, 1 890 ; W . C . B ray, Journ. Amer. Chem. Sac., . S2.932,1910 ; W. C. Bray and E . L. C onnolly, ib., 33. 1487, 191 1 ; U. J ones and M . L . Hartmann, ib., 37. 241, 1915. l6 M. Rerhhelot and E . J ungflejsch, Ann. C'him. Phys., ( 4 ) ,26. 396, 1872 ; H . T . C alvert, Zeit. phys. Chem., 38. 513, 1901 ; A . A. J akowkin, ib., 18. 585, 1895 ; W . H erz and M. Lewy, Jahrb. Schles. Vnierland Kztltm, I , 1906. 'W. Schonbein, Pogg. Ann., 72. 450, I847 ; N . A. E . Millon, Compl. Rend., 28. 4 2, 1849 ; F. A. J. Ralard, Anq~.Chim. Phys., ( 2 ) ,57. 225, 1934 ; H . E. Roscoe, Journ. Chem. Soc., 8. 15, 1856 ; Liebig's Ann., 95. 357, 1855 ; C. G opner, Ber., 8. 287, 1875 ; R. S chiff,ib., 8. 410, 1875; W . W olters, Dingier's J ourn., 210. 362, 1873 ; J ourn. prakt. Chem., ( 2 ) ,7 . 468, 1873. l7 A. A. J akowkin, fitit. phys. Chem., 29. 613, 3899. . l 8 S. U P ickering, Jour?~. . Chern. Soc., 37. 139, 1880; A. R ichardson, ib., 83. 380, 1903. I @ W . C. Bray, Journ. Amer. Chem. Roc., 32. 932, 1910; G . J ones and M . L. Harhruann, &6., 37. 241,1915 ; A. A. J akowkin, Zeil. phys. Chem., 29. 613, 1899. 2 0 C. L. R e~thollet, t m . dcad., 276, 1785 ; W . H . Wollaston, Gilbert's Ann., 31. 416, 1809; M 39. 291, 1811 ; H . Davy, Elements of Chemical Philosophy, London, 1812 ; N . T . de Saussure, Nbm. Acad. Roy. T u r i n , 4. 4 41, 1790 ; CrelE's A m . , 1. 356, 1796. 21 T. T orosiewicz, Repert. P h r m . , 57. 335,1836 ; W . C. Wittwer, Pogg. A nn., 94. 597, 1855 ; 97.304, 1856 ;J . W . D raper, Phil. Nag., (31, 27. 327, 1845 ; H . Tuchel, Ueber die Einwirkung des Lichfes auf Cldorwnsser, Konigsberg, 1. P 1913. r., 2 2 R. R unsen and H . E . R oscoe, Pogg. Arm., 96. 373, 1865 ; 97. 304, 1855 ; P hil. T ~ T M 46. 1. : 3E5,601,1857 ; J . W . Draper, Phil. Naq., ( 4 ) ,14. 321, 1857 ; D. L . Chapman and G . 13. Burgess, Proc. C hem Soc., 20. 52, 1G4, 1904. F . U T6hler,Liebig's Ann., 85. 3 75, 1853 ; A. Pedler, Joum. Chem. Sac., 57. 613, 1890; J . M . Eder, Sitzbcr. Akad. Wien, 02. 340, 1885. E. Klimenko, Ber., 28. 2558, 1895; E . K limenko and G . P ekatoros, Z d . phys. C'7Lem, 4. 483,1889; 1: K limerlko and W. R udnizky, Journ. Buss. Phys. Chem:Soc., 18.220,1895; C. F. Riir; wald and A. M onheim, Repert. Pharm., 54. 355, 1835 ; J . Rillitzer, Chem. Ztg., 30. 1 133, '1906 ; J . Milbauer, Zeil. phys. Chem., 88. 564, 1914 ; A . B enrath and H. Tuchel, Zeil. wise. Phot., 13. 383, 1914. 2 5 A. P opper, Liebig's Ann., 227. 161, 1883 ; 231. 137, 1886 ; A. P edler, Journ. Chem. SOC.,57. 613,1890 ; G . Gore, Proc. Roy. Soc., M. 362, 1889 ; Chem. News, 57. 184, 1888 ; 6 0. 271, 1889 ; J . Billitzer, Chem. Ztg., 30. 113J, 1806; E. K ljmenko, vide supra. le . C . W ittwer, Pogg. Ann., 94. 097, 1885 ; LiebZg's Ann. SuppX, 4. 6 3, 18G5 ; R. R unscn W and H . E . Roscoe, Pogg. Ann., 96. 373, 1856; N . T . d e Saussure, vide supra. 2 7 C. L owig, Pogg. Ann'., 14. 485, 1828 ; J . M . E der, Sitzber. Akad. Wien, 92. 304, 1885 ; H. W. V ogel, Phot. Corrcsp., 62, 1866 ; C. F . Cross and A. H iggin, Journ. Chem. BOG.,35. 249, 1879; T . W . Richards and W . N. S tull, Zeit. phys. Cl~em.,41. 547, 1902; H. Rordier, Compl. Rend., 163. 291, 191G. W. L. G oodwin, Trans. Roy. Soc. Edin., 30. 587, 1882 ; Ber., 15. 3039, 1882 ; C. A. K ohn and F . O'Brien, Journ. Boc. Chim. I d , 17. 100, 1898 ; E. G . K u m p f , U d e r die Absorptio?a eon Ghlwdwch C l~.l~rnatriu??zMsuqe?~, 1881 ;M . D ettmar, Liebig's Ann., 38.35, 1841 ; H . Schulze, Graa, Journ. prakb. Chem., ( 2), 24. 108, 1881 ; H . W . B. R oozeboom, Rec. Trav. Chim. Pays-Bas, 4. , 279,1885 ; R . E . S loan, Chem. News, 44. 203, I881 ; E . 0.M andala, G azz. Chin&.Ilal., 50. ii, 89, 1920. a 8 H. E. R oscoe, Jousn. Chem. Soc., 8. 14, 1856 ; J . W . Mellor, ib., 79. 216, 1901 ;M. Rerthelot, Ann. Chim. Phys., ( 5 ) ,22. 462, 1881 ; Compt. Rend., 91. 194, 1880; J. W . Draper, Phil. Mag., ( 3 ) , 23. 401, 1843; E . 0.M andala, GaC(zz. Chim. Ilal., 50. ii, 8 9, 1920. ao A. A. J akovkin, Zeil. phys. Chem., 20. 38, 1896 ; W . H . McLauchlan, ib., 44. 600, 1903 ; M. Wildermann, ib., 11. 421, 1893 ; M. Roloff,ib., 13. 327, 1894 ; F. Roricke, Zeil. Ekklrochem., 11. 57, 1905; F. P. W orley, Journ. Chem. Soc., 87. 1107, 1905 ; G . A. L inhart, Journ. A m r . Chem.SOC., 0 158, 1918 ; G . J ones and M . L. H artmann, T r a m . Amer. Electrochem. Soc., 30. 4. 295, 1916; J . M . Bell and M. L . Ruckley, Journ. Amer. Chem. Soc., 3 4. 14, 1912; A. F. Joseph, Jwron. Chem. Soc., 1 17. 377, 1920 ; E 0.Mandala, CTazx. Chim. Itccl., 50. ii. 89, 1920. " 90 INORGANIC AND THEORETICAL CHEMISTRY A. A . Noyes and J . S eidensticker, Zeit. phys. Chem., 27. 359, 1898; A. A. Jakowkin, ib., 13.539,1894 ; L. B runer, ib., 26. 147, 1898 ; W . H . M cLauchlan, ib., 44. 61 7,1903 ; E . Brunner, ib., 58. I , 1906 ; R. A begg and A. Hamburger, Zeit. anorq. Chem., 50. 403, 1906 ; C. K raus, Neues Rep. Pharm., 21. 385, 1872 ; T . Koller, N e w s Pharm. Jahrb., 25. 206, 1886 ; M. S ocquet and A. G uilliermond, Journ. Phawn. Chim., (31, 26. 280, 1854 ; M. Debauquc, ( 3 ) ,20. 34, 1851 ; W . V aubel, Journ. prakt. Chem., (2), 50. 349, 351, 362, 1884; E . 0.Mandala and A . Angenica, Gctzx. C him. Ital., 5 0. i, 573, 1920. 32 B . E. S loan and J. W . M allet, Chem. News, 4 6. 194, 1882 ; W . V aubel, Journ. prakt. Chem., ( 2 ) ,63. 381, I901 ; F. S estini, Bull. Soc. Chim., ( 2 ) ,10. 226, I868 ; E . C . F ranklin and C. A . Kraus, Amer. Chem. Journ., 20. 821, 1898 ; U.A ntony and G. Magri, Gazz. Chim. Ital., 35. i, 206, 1905 ; E. H. Biicher, Zeit. phys. Chem., 54. 665, 190G ; A. H antzsch and A. V a g t , ib., 38. 705, 1901 ; P. Walden, ib., 43. 385, 1903 ; E . S olly, Phil. Hag., (31, 8. 130, 400, 1836 ; J I nglis, ib., ( 3 ) , 0. . 4 50, 1836 ; ( 3 ) ,7 . 441, 1835; ( 3 ) , 8. 12, 130, 191, 1836; C . H ugot, Ann. Chim. Phys., (71, 21. 6, 1900 ; 13. D. Steele, D. M cIntosh, and E . H . Archibald, Proc. Roy. Soc., 74. 321, 1905. 3 5 L . B runer, Anz. Akad. Cracow, 731, 1907 ; Z eit. Elektrochem., 16. 264, 1910 ; Z e4. phys. Chem., 26. 147, 1898 ; A . Hantzsch and A. V a g t , ib., 38. 7 05, 1901 ; A . A. Jakowkin, ib., 18. 586, 1895 ; W . H . McLauchlan, ib., 44. 6 00, 1 W 3 ; P. Wgntig, ib., 88. 513, 1909 ; P. Walden, ib., 43.'385, 1903 ; T . Budde, VercifS. Geb, b.ilitarsanitdtswesens, 52, 1912; Chem. Centrb., ( 5 ) , 16. i , 1856, 1912 ; H . E . B arnard, Pharm. Rev., 26. 308, 1908; H. Arctowsky, Zeit. anorg. Chem., 11. 2 76, 1896 ; 6 . 404, 1894 ; E . Bechmann and P . W gntig, ib., 67. 17, 1910 ; J W . R etgers, ib., . 5. 3. 343, I893 ; W . H erz and &I.K noch, ib., 4 269, 1906 ; H . Euler, Widd, Ann., 63. 273, I897 ; J . M . E der, Sitzber. Akad. Wien, 92. 340, 1885 ; H . M . V ogel, Phot. Corresp., 62, 1866 ; W . D uncan, Pharm. Journ., 22. 544, 1892 ; J. C . G. de Marignac, N e w s Arch. P h a m . , 29. 217, 1870; M . B erthelot and E . Jungfleisch, Ann. Phys. Chim,, ( 4 ) , 26. 396, 1872; J. Ogier, Compt. R e d . , 91. 924, 1880 ; S . U . P ickering, Journ. Chem. Soc., 53. 865, I888 ; 11. M. Dawson, z'b., 93. 1308, 1908 ; W . J. Jones, ib., 99. 392, I911 ; P. W alden, Zed. Elektrochem., 12. 77, 1906 ; A . Harion, Ann. Z'&cole Norm., ( 2 ) ,6. 367, 1877 ; M , von Wogan, Ber. deut. p h y ~ Ges.,. 6. 562, 1908; J. H. Mathews, Journ. Phys. Chem., 9, 641, 1905; E. H . Archibald and W . A . Patrick, Journ. Amer. Chem. Soc., 34. 369, 1912 ; D. McIntosh, ib., 33. 71, 1911 ; J. T raube, Verh. d e d . Baturforsch., 63. 103, 1892 ; C . Liiwig, Lieiebiq's Ann., 3. 288, 1832 ; N. Schoorl and ?. A. Regenborn, Pharm. Weekb., 56. 538, 1919 ; A . I J oseph, Journ. Chem. Soc., 103. 1554, 19133 4 P. W alden, Zeit. phys. Chem., 43. 3 85, 1903; J . H . M athews, Journ. Phys. Chem., 9. 641, 1905 ; Arch. Pharm., ( 3 ) ,23.43 1 , 1885 ; A. H . A llen, Journ. Soc. Ch,em. I d . , 5. 65,282, 1886. § 12. Chemical Reactions of Chlorine, Bromine, and Iodine Chlorine unites directly with most of the elements. The inert gases of the argon family, nitrogen, oxygen, and some of the platinum metals, resist attack by free chlorine, although compounds with all but the inert gases can be obtained indirectly. Chlorine unites directly with hydrogen in light, but not in darkness ; t he union is also induced by the silent electrical discharge, or by the mere presence of some catalytic agents. Chlorine is not combustible in air. A jet of burning hydrogen lowered into a jar of chlorine continues burning with the formation of hydrogen chloride. Conversely, chlorine gas may be burnt in an atm. of hydrogen. Hydrocarbons are decomposed by chlorine; for instance, a piece of cotton wool soaked in warm turpentine (CIOHIG)will inflame when placed.in a jar of chlorine. The burning of the turpentine in chlorine gas is accompanied by the separation of den~e clouds of free carbon ; t he chlorine combines with the hydrogen forming hydrogen chloride. A wax candle burns in chlorine with a very smoky flame ; t he hydrocarbon-wax-is decomposed in a similar manner. Hence, chlorine may be regarded as non-combustible, and a supporter of combustion. Hydrogen and bromine unite under the influence of the silent discharge or when heated, but not when exposed to sunlight. Hydrogen and iodine unite when heated; a t ordinary temp. the reaction with hydrogen and iodine is endothermal, being -0.8 Gals., but the reaction is exothermal above about 500". I n 1842, T . Andrews 1 pointed out that although moist chlorine-combines energetically with zinc, copper, and iron-filings, perfectly dry chlorine " has no action whatever a t ordinary temp., . . . a nd the same remarks may be applied t o the behaviour of dry bromine in contact with dry metals." Indeed, t.horoughly dry chlorine is somewhat inert chemically, and it has no appreciable action upon bright metallic sodium, copper, etc. Dried chlorine scarcely acts a t all upon dry silver ; T HE HALOGENS 91 moist chlorine readily attacks the metal, more rapidly in light than in darkness. Moist chlorine is particularly reactive towards these solids. No one has succeeded in drying mercury a nd chlorine so thoroughly that the two elements do not react at ordinary temp. Chlorine does not combine with oxygen directly, although several compounds of chlorine and oxygen can be obtained indirectly forming i l series of chlorine oxides or hydroxy-chlorine compounds. While the affinity of the halogens for hydrogen decreases with increasing at. wt., the general tendency with oxygen is in the reverse direction, but not in so marked a way as with hydrogen. Thus, although fluorine forms no known compound with oxygen, numerous compounds of oxygen with chlorine have beeli obtained; and, judging by the know11 compounds with oxygen, the affinity of bromine for oxygen appears to be less, not greater, than is the case with chlorine. Bromine seems to occupy an anoma'lous position with respect t,o oxygen, although it must be remembered that bromine has not been investigated so much as chlorine. Bromine and iodine form an unstable series of compounds analogous with hypochlorites and chlorates, but the bromine analogue of perchlorates has not been prepared. To summarize : Oxidizing action Oxy-acids . Oxide8s b , . . . F LUO~INE. S tronge~t None None CHLORINE. Very strong HCIO HClO, HC10, HC1o4 CI ,O C 10, - Cl,Oi BROMINE. St,rong HBrO - HBrO, - - - IODINE. Weak HI0 - HIO, HJO, - 1204 LO6 - P. Hautefeuille and J. Chappius 2 claim to have made a compound, N2Cl2OI8,by the action of the silent electrical discharge ou a mixture of chloriile and oxygen with a trace of nitrogen. The affinity of bromine for oxygen seems to be even less than that of chlorine or iodine f ~ oxygen. No bromine oxides are known, but r several oxides of chlorine and iodine have been made. According to M. Berthelot,3 no change could be detected after exposing iodine in an atm. of oxygen t o sunlight for five months : b ut J. Ogier obtained a compound of the two elements under the influence of the silent discharge. The rapidity of the action of the halogens on water, previously discussed, is slower thegreater the at. wt. of the halogen. Moist chlorine, or chlorine water, is a powerful oxidizing agent. The decomposition of chlorine water in sunlight has already been discussed, whereby oxygen gas is given off and hydrogen chloride is formed : 2H20$2C12+4HC1+02. This reaction is reversible in light, and even a conc. aq. solu. of hydrochloric acid is oxidized by air in light, so that such a soh. acquires a yellow colour owing to the formation and soh. of free chlorine formed by reaction with the absorbed oxygen. W. Henry 4 also found, in 1800, t hat when a mixture of hydrogen chloride and oxygen is passed over platinum black, the mixture is decomposed with the formation of chlorine and water : 4HC1+02=2H20-+-2C12, and it has also been shown that the metals, iridium, palladium, xuthenilun, and osmium act in a similar way. The heat of formation of a molecule of water is 68.4,Cals. and of a molecule of hydrogen chloride in aq. soln., 3 9-3 Cals. Hence, in the reaction H2O,,.+ClEW, =2HCl2,,.+0, the amount of heat required to decompose a molecule of water is nearly 10 Cals. less than that evolved in the formation of two molecules of hydrogen chloride in aq, soln. Hence, the instability of a s o h of chlorine in water might have beeu anticipated from the thermochemical data. If? however, the gases be in C question : 2HClgaa+O=HzOgas+C12gaB+1330als., and this agrees with observation, for the reaction once started will proceed to a certain limiting value without furthor heatiug if sufficient precautions be taken to prevent loss of heat by radiation. This reaction is the basis of Deacon's procew for chlorine. I NORGANlC AND THEORETICAL CHEMISTRY I a piece of coloured litmus paper, coloured petals of a flower, or a piece of cloth f dyed with turkey red or indigo blue be placed in a jar of dry chlorine no appreciable change occurs ; b ut i moisture be present, the colours are bleached by the chloriue. f The actiou appears to be due to the formation of a colourless oxidation product. Ordinary oxygen will not do the work of bleaching. One school of chemists therefore assumes that the oxidation is effected by the nascent oxygen. A s o h . of indigo is decolorized by chlorine and bromine ; a nd a s o h . containing a milligram of diphenylamiue in 10 C.C. of sulphuric acid is coloured an intense blue. The direct union of many of the metals with chlorine is attended by incandescence -for instance, powdered antimony, arsenic, and bismuth when shaken into a flask containing chlorine. Since the chlorides of antimony, etc., so formed are poisonous, the experiments are best made in a closed vessel, or in a well-ventilated fumechamber. When the bulb tube containing the powdered element is raised, it is easy to shake the contents through the flask of chlorine. to illustrate the incandescence which attends the combustion, without a n escape of the poisonous chlorides into the atm. of the room. Copper, o r brass foil-Dutch metal-phosphorw, boron, a nd silicon also ignite spontaneously in chlorine. Molten sodium, hot brass wire, and iron wire also burn in chlorine. Liquid chlorine a t its b.p., -33.6", reacts with arsenic with iucandescence, but it does not react with antimony or bismuth. Arsenic even reacts with liquid chlorine a t as low a temp. as -80" when potassium and sodium retain their metallic lustre without reaction. Aluminium t akes fire in chlorine a t -20°, G . J ust and P. H aber found that there is an emisbut is not attacked a t -33.6". sion of electrons a s evidenced by .electrical conductivity when the vapour of iodine acts on the heated metals, aluminium (160°), copper, and silver. Yellow phosphorus unites with liquid chlorine with explosive violence, and red phosphorus is also in attacked. T a nd gold a re also attacked by the liquid. According to G . Lunge, liquid chloriue does not attack iron below 90°, and i t is on account of this inertness of iron that it is possible to transport large masses of liquid chlorine in iron bombs. Chlorine gas attacks fine.1~ divided gold a t ordinary temp.,6 bromine quickly dissolves gold ; d ry iodine does not attack gold, but between 50" and the m.p. of iodine, gold forms crystals of aurous iodide, AuI. Platinum i s attacked by chlorine under certain conditions. According to F. H aber, if a platinum electrode be dipped into hydrochloric acid containing chlorine, and then thoroughly washed with water and hot alkali, a blue coloration appears in contact with starch and potassium iodide, which is much more intense than that produced by the catalytic action of the metal on the oxygen of the air, and i t is considered that the metal has been attacked. Platinum does not react with bromine either in aq. or in hydrochloric acid soln. Chlorine, bromine, or iodine, moist or dry, rapidly attacks mercury.6 Sodium is indifferent towards bromine or iodine. According to V. Merz and W. Weith 7 no reaction occurs even a t 300". A piece of sodium was kept in contact with bromine for sixteen years without chemical action. With potassium a nd broinhe or iodiue the reaction occurs with explosive violence. Magnesium foil is not attacked after lying for five years in contact with bromine, but aluminium i s rapidly attacked by bromine with incandescence. Sulphur r eacts with liquid chlorine in the proximity of its b.p., and i t f o r m various chlorides of sulphur ; s indarly with selenium a nd tellurium ; a t lower temp. the liquid does not react with sulphur, while i t does react with selenium and tellurium. Liquid chlorine reacts with sulphur dioxide forming sulphwyl chloride, SOzClz;carbon disulphide i s mjscible in all proportions with liquid chloriqe without chemical action. All hydrogen compounds, excepting hydrogen fluoride, are decomposed by chlorine with the formation of hydrogen chloride--e.g. hydrogen sulphide, hydrogen phosphide, arsenide, iodide, etc. The metal bromides, iodides, a nd sulphides a re likewise decomposed either a t ordinary or a t higher temp. According to M. B erthelot,s bromine is absorbed by conc. hydrochloric or aq. soln. of b arium or strontium chlorides with the development of much heat--owing, he assumes, to the formation of a pcrchlorobromide, R"(Cl.Br&. N. N. Beketoff observed T HE HALOGEN8 93 no action between iodine ajld cmsium chloride after standing for 50 days a t room temp. I n their study of the action of iodine on sodium thiosulphate i n 1842, M. J. Bordos and A . Gklis fouud that a soln. of iodine is decolorized by this salt, owing to t+he formation of sodium tetrathionate, Na2S406, and sodium iodide : 12+2Na2S203 =2NaI+Na2s406. H. Hertlein translates this reaction in terms of the ionic where hypothesis : 4Nan+2S203ff+12=4Na.+21f+S40~', the two negative S20{anions transfer half their charge to the un-ionized iodine atoms forming two If-ions, and the two S203"-ions simultaneously form one S40,Y,"-ion. According to C . A. R . Wright (1870) and S. U. Pickering (1880), there is a secondary reaction resulting in the formation of sodium hydrosulphate. The latter represented the reaction : Na2S2O3+4l2+5H2O=8HI+2NaHSO4, a nd the soln. becomes distinctly acid ; C. A. R. Wright attributed the formation of the sulphate to the oxidizing action of iodineon the tetrathionate, say: Na2S406+712+10H20=14HI+2H2S04+2NaHS04, and not to the direct oxidation of the thiosulphnte. S. U. Pickering noted that ' a greater proportion of sulphate is formed the higher the temp.-at 0 a bout 1-84per cent. of the iodine is consumed in forming sodium hydrosulphate ; a t lo0, 1-94per cent. ; a t 20°, 2.10 per cent. ; a t 30°, 2-35 per ceut. ; a t 5.2", 3'90 per cent. -the degree of dilution, the excess of alkali iodide or the amount of hydrochloric acid, and the time occupied by the reaction do not affect the result. According to R. H. Ashley, the oxidation of the thiosulphate in alkaline soh. progresse? further than the tetrathionate. G. Topf, J. P. Batey, E. Abel, C. Friedheim, and others have also noted the formation of sulphates during the action of iodine on sodium thiosulphate in alkali or sodium hydrocarbonate s o h I. M. Kolthoff thus summarized his observations : The tetrathionate reaction occurs in neutral and in strongly or feebly acidic s o h , while in weakly alkaline s o h some sulphate is simultaneously formed, and in strongly alkaline soh. all the thiosulphate may be' converted into sulphate. The decomposition of the thiosulphate in a strongly acid soln. is slow in comparison with the formation of tetrathionate. According to J. Bougault, in alkaline soh. the iodine forms a hypoiodite : 2NaOH+12+NaI +4NaIO +H20 +NaOI+H20, which reacts with the thiosulphate : Na2S203 =2NaHS04+4NaI ; M . J. Fordos and A. Gblis, and G. Lunge have shown that the hypochlorites, hypobromites, and bypoiodites also oxidize the thiosulphate to tetrathionate. Some blkali iodate is also formed : 6NaOH+312= +3H20. The oxidation to iodate i~ much retarded by using carbonate, and still more by using normal carbonate in place of. t and E. J. Maument5 noticed that with barium thiosulphate, BaS20 thionates are formed--e.g. H2S204 and H2S609-as well as the te Chlorine and bromine act differently from iodine on the thiosulphate H. Hertlein, the ionizing tendency of the two first-named halogens i the salt is oxidized to sulphate. According to M. Berthelot, the oxidation of sodium thiosulphate to the sulphate by bromine liberates 150 Cab. M. J. Fordos and A. GBlis noted that the sulphate is formed when chlorine or bromine acts on the thiosulphate : Na&3,O3+4Cl2 +5H20 =Na,S04+H2S04 +8HC1. As shown by G. Lunge, the reaction takes place in two stages ; t here is first a considerable precipitation o sulphur : Na28203+H,0+C12=2NaCl+H2S04+S ; a nd the turbidity then f clears owing to the oxidatiou of the sulphur to sulphuric acid : 8+4H20+3C12 =6HCl+H2S04. G . Lunge also noted that a little tetrathionate is formed: 2Na2S2O3+Cl2=2NaCl+Na2S1O6 ; a nd some trithionate is formed a t the same time. According t o L.W. Andrews,10 when potassium iodide is t itrated with chlorine water in neutral soh. with chloroform or carbon tetrachloride as iudicator (these immiscible aolvents become violet, owing to the liberation of iodine, and they become colourless when titration is complete), the reaction is represented : KI+3C12+3H20 =KCl+HI03+5HC1. This reaction is the basis of A. and I?. DuprB's process for titrating potassium iodide (1855). If a large excess of hyd~ochloric acid be present, INORGANIC AND THEORETlCAL CHEMISTRY t,he reaction stops a t an earlier stage for the ensuing reactiorl is symbolized : R1 +Cl,=KCl+ICl, and the iodine chloride colours the liquid yellow. I a s oh. f of potassium iodate be used in place of chlorine water in t itrating potassium iodide the stage a t which the reaction stops depends upon the conc. of the acid ; with lcw conc. the reaction goes no further than the liberation of the iodine : 5KI'-/-KI03 +6HC1=6.KC1+312+3H20; with an excess of acid, 2KI+KI03+6HC1=3KC1 4-31CI+3H20. I n J . L. Gay Lussac and L. J. T hhard's early attempt 1 1 t o decompose oxymuri.atic acid (chlorine) by passing it over red-hot c h a r d , t hey reported : Tho first portions o the oxygenated muriatic gag wore completely converted into ordinary f muriatic gag. This &ect diminished gradually in spito of a very great elevation of temp., and soon the gas passed without alteration, mixed only, towards the end of the experiment, with one thirty-third of an inflammable gas, which we believe to be carbonic oxide gag. This result clearly showed us that oxygenated muriatic gas is not decomposed by charcoal, and that the muriatic gas which we had obtained a t the commencement of $he operation was 4ue to tho hydrogen of the charcoal. I n fact, on taking ordinary charcoal without igniting it, muriatic gas was disengaged during a lengthened period oven a t a temp. only slightly elevated. According as the charcoal lost its hydrogen, howovor, the quantity of muriatic acid went on diminishing, and finally nothing was obtained but oxygenated muriatic gas ( i.e. chlorine). ... ... En passant, t his is one of the best ways of purifying charcoal. H. D avy (1814) 1 2 a nd H. Ziiblin (1881) obtained no evidence .of a reaction between iodine or bromine and carbon a t a white heat. Porous charcoal absorbs chlorine with the evolution of 6-18 Cals. per 35.5 grms. of chlorine. According t o W. G. Mixter strongly campressed sugar charcoal absorbs about 4 per cent. of chlorine a t a red heat, and does not give i t up in vacuo a t that temp. About 1.5 per c.ent. of bromine and traces of iodine are also absorbed b y porous charcoal. The absorbed gas can be displaced by hydrogen but not by nitrogen ; bromine and iodine are less easily absorbed and more easily lost than chlorine. Lampblack absorbs more chlorine than charcoal, and both lose the adsorbed gas a t about 1000". Chlorine seems to act with most energy on those forms of carbon which are contaminated with hydrogen. .Gas carbon, graphite, and the diamond are not affected by chlorine. M. Meslans 1 3 found that if hydrogen be passed over the charcoal sat. with this gas, hydrogen chloride is formed in darkness ; a nd according to M. Berthelot and A. Guntz, much chl0rine.i~ t the same time displaced by the hydrogen with an absorption of heat. a f The net result of the process is therefore an absorption of heat. I water be poured on carbon sat. with chlorine, both hydrogen chloride and carbon dioxide are formed. R . Lorenz found +hat chlorine is completely converted into hydrogen chloride and carbon monoxide by passing a mixture of steam and chlorine through a tube fiHed with coke a t s dull red heat. The reaction, C12+H20+C=CO+2HC1, is complete. If t he hydrogen chloride be absorbed by water, the residual carbon monoxide is contaminated with but a little carbon dioxide. According to P. Pischer, coal rapidly absorbs 30-36 per cent. of bromine. The brominated coal gives off hydrogen bromide. Carbon unites with chlorine directly when the carbon forms an electric arc in an atm. of chlorine ; hexachlorobenzene, C6H6C16,is formed ; a nd if the gas be confined in a vessel with a construction so that the upper compartnlent contains the arc, and the lower part is cooled by a freezing mixture, hexachloro-ethane, C2C16,is formed. The positive pole appears to be the active agent. Similar results were obtained with bromine and iodine. According t o E. B arnes, dry chlorine has no perceptible action on calcium carbide after two months' treatment at ordinary temp. According to H. Moissan, and F. P. Venable and T. Clarke, bromine has no actibn in the cold on calcium carbide, but when heated the products are calcium bromide and carbon. E. Barnes, however, found that liquid bromine attacks calcium carbide at ordinary temp. producing carbon hexabromide and calcium bromide : CaC2+4Br2=CaBr2 +C2Br6 ; a t 100" in sealed tubes, carbon and calcium bromide but no carbon hexabromide were formkd. E . and H. E rdrnann found that iodiue acts on calcium T H E HALOGENS 95 carbide a t 170°, producing tetra-iodoethylene and di-iodoacetylene. R . L ucion (1889) could find no evidence of a reaction between dry carbon dioxide a nd chlorine a t a red heat, but in the presence of moisture, hydrogen chloride is formed. Chloriue unites with carbon monoxide f orming phosgene, or carbonyl chloride, COC12. Ammonia, r eacts vigorously with chlorine, forming ammonium chloride and nitrogen gas ; if t he chlorine be i n excess, oily drops of violently explosive uitrogen chloride, NC13, are formed. Some ammonium hypochlorite, NH40Cl, is simultaueously produced. Nitrogen bromide does not appear t o be formed by the action of bromine ou aq. ammonia. The ppeed of t h e reaction, 2NH3+3Br2=N,+6HBr, is more rapid with free ammonia than with the ammonium salts, and 8 . Raich 1 4 a ttempted to get a comparison of the affinity coeff. af the acids forming the various ammonium salts in terms of the rate of decomposition of the different ammonium salts by bromine. It was assumed that the stronger the affinity of t.he acid for ammonia, the slower the rate of the daconiposition of bromine. It was found that with the ammonium salts of the organic acids,--e.g. oxalic aud formic acidsside reactions occurred owing to oxidation or substitution. With 10 C.C. of a 0105N-soh. of the ammonium salts, and 25 c.c. of a OS075N-soh. of bromine water a t 25", 2'5 C.C. of bromine was consumed in Time Hydrochloric . 162 Nitric 148 Sulphuric 118 Arsenic Monochloracebic Tartaric 98 29 20 Acetic acid. 4 hrs. This was taken t o represent the relative order of the affinity of these acids for ammonia. When a n alcoholic s o h . of iodine is treated with ammonia; substitution products and nitrogen iodide are formed. According to A. W. Browne and F. P. Shetterley,l5 a trace of azoimide, HN3, is formed b y the action of chlorine on acidic or a1kaliue soh. of hydrazine sulphate ;a ccording to E. E bler, hydrazine snlphate or chloride in acid s o h . is completely decomposed by bromine-for example : N 2H4 +2Br2=4HBr+N2. A soln. of iodine in potassiumiodidc acts in a similar manner. T. Curtrus and H. Schulz found that tincture of iodine acts on an alcoholic soh. of hgdrazine hydrate a s symbolized by the equation : 5 N z Q . H 2 0+2T,=4N2&HI f 5H,O+N,. Nitric oxide u nites with chlorine forming nitrosy1 chloride, NOCl ;t he same product is formed when chlorine acts ou nitrogen tetroxide, N20,. Bromine also forms nitrosyl bromide, NOBr ; b ut iodine suspended in water forms nitric and hydriodic acids. Wheniodine is warmed with conc. nitric acid, iodic acid and nitrogen peroxide are formed. According t o J. B. Senderens, chlorine acts on silver nitrate formingsilver chlorate and silver chloride, while according t o H. Moissan,lB chlorine free from hydrogen chloride precipitates silver chloride from a s o h . of silver nitrate, and the corresponding amount of oxygen is set free. Iodine and silver nitrate form silver iodide and iodic acid. Solid iodine acts ou silver nitrate in the darkin accord with the equation: 5AgN03+312+3H20=5AgI+5HN03+HI03 ; a nd bromine, AgN&+Br2+H20=AgBr+HOBr+HN03, a s indicated by J. Spiller, and C. I?. Schonbein. According to K. Briickner, mercurous oxide, Hg,O, and iodine, 12, in the proportions 1 : 2 mol, react; 6Hg20+1212=11Hg12+Hg(I03), ; t he same transformation occurs slowly in the presence of w ater, more rapidly in hot water. With a dry mixture of mercuric oxide, HgO, and iodine, 6HgO+612 =Hg(I03)2 +5HgT. The action of the halogens on the alkaline hydroxides is discussed in connection with theoxy-chlorine acids. I n the presence of hydrogen peroxide a nd an alkali hydroxide, ths alkali chloride is formed : Clz+H202+2KOH=2KC1+02+2H20. Similarly with iodine and bromine. When near its b.p. liquid chlorine reacts with arsenic forming arsenictrichloride, but a t lower temp. arsenic and antimony a re not affected. Alkali amenites a re oxidized t o arsenates : K 3As03+H20+C~=KH2As04+2KC1. k o w salts a re oxidized to ferric salts : 6PeS04+3C12=2Fe2(S0,)3+2FeCI, Alcohol, ether, chloroform, and carbon disulphide are better solvents for bromine or iodine than water, so that if a soln. containing free bromine or free iodine is shaken up with one of these solvents, the former colours these solvents yellow or brown, and t he latter rose to reddish-violet, according to the concentration. Liquid chlorine a t 96 INORGANIC AND 1THEOR.ETICAL CHEMISTRY its b.p. does not react with potassium permanganate; thallous chloride is converted into T12C13by liquid chlorine, and into TlzC&by gaseous chlorine a t ordinary temp. Uses of the halogens.-Chlorine is largely employed in the preparation of bleaching powder or chloride of lime, bleaching liquor or hypochlorites, chlorates, and various chlorinated chemica1s-e.g. 3000 t ons of chloracetic acid are said to be used per annum in the manufacture of synthetic indigo ; chloral ; carbon tetrachloride ; various chlorinated ethanes are used in the extraction of fats, etc., with the advantage of their being nou-inflammable ; chloro-derivatives benzene and naphthalene, in the manufacture of dyes. Liquid chlorine was formerly employed in the chlorination process for gold, but this has been largely displaced by the cyanide process. The gas was introduced by the Germans as an agent of destruction in warfare in the second battle of Ypres, on 23rd April, 1915. This was in contraventl:on of International agreement. The Germans thus obtained a temporary advantage until respirat,ors with sodium thiosulphate and carbonate had been supplied t o the Allied Forces, and in self-defence, the Allies had retaliated on the Germans with interest in kind. I n consequence, this and other gases even mdre deadly were used on the battlefields. For chlorine poisoning the inhalationof hydrogen sulphide, or of thevapour of alcohol, ether, chloroform, or steam has been recommended. There is a form of chlorine poisoning to which the workmen dealing with electrolytic chlorine are subject ; '7 i t is attended by swellings in various parts of the body, giddiness, and coughing ; t he effects are, supposed to be due not to the direct action of chlorine, but rather to the action of chlorine oxides contained in the gas. According to A. Leymann, the workmen may also be affected by a peculiar skin disease which has been traced to chlorinated products derived from the action of chlorine on the tar coating of the electrolytic cells used in making the gas. At the beginning of the nineteenth century, chlorine was recommended as a disinfectant by W. Cruickshank ;1s a nd towards the middle of the century, the efficiency of chlorinated lime as a disinfectant and deodorant was generally recognized, and in 1854 a Royal Commission recommended this substance for deodorizing the sewage of London. At that time, the action of disinfectants was generally supposed to be effected by the arresting or preventing of putrefactive changes. The work of T . Schwann (1839), Ij. P asteur (1862), etc., showed that bacteria were responsible for putrefaction and fermentation, and that specific organisms were responsible for certain specific diseases. Chlorine was successfully used in the ~anitary work connected with the outbreak of .puerperal fever in Vienna in 1845. J, R ace estimated that in 1918 over 3000 x 106 gallons of water were being treated per diem in North America ; a nd over 1000 cities and towns employed the process. I n 1889, W. Webster proposed the use of electrolyzed sea-water as a disinfectant, and this liquid was introduced by E. Hermite (1889) as Wermite's,Ruid for domestic purposes as well as for flushing sewers and latrines. The objections to this liquid were due to the unstable character of the magnesium hypochlorite forined a t the same time as the sodium hypochlorite in $he electrolysis of sea-water. The magnesium salt readily hydrolyzes, forming the hydroxide which deposits in the electrolytic cell, and leaves a soh. of unstable hypochlorous acid. A conc. soh. of sodium chloride was soon substituted for sea-water, and the liquor has been called electroxone, chloros, etc. Various preparations of hypochlorites were also employed on a large scale for the purification of water ; H. Bergk used chlorine peroxide ; a nd in A. C. Hou~ton's process water was chlorinated with chlorine itself. S. Rideal noticed the peculiar effect of ammonia on the germicidal value of hypochlorites ; h e said that the first rapid consumpfion of chlorine is succeeded by rt slower action which continues for days, and that the germicidal action continue3 &er free chlorine and hypochlorite have disappsared ; i t appeared that ammonia substitution products are formed yielding compounde more or less germicidal. J. Race, attempting to make ammonium hypochlorite by the double decomposition of ammonium oxalate and calcium hypochlorite, obtained a soh. with its germicidal action greatly enhanced; and it mas assumed that the unstable ammonir~rn THE HALOGENS hypochlorite passed into monochloroamine, NH40Cl=NH2Cl+H20. T he use of chloroamine for chlorinating water was tried on a large scale a t Ottawa (1917), and a t Denver, Col. (1917), with good results. N. D. D akin and co-workers (1917) tried other chloroainines-e.g. sodium toluene-p-sulphochloramide ; a nd several aromatic sulphodichloroamines, e.g. C12N.SOz.C6H4.COOH,were found s uitable for use in t abloid form for the sterilization of small quantities of water by c avalry and other mobile troops. The preparation has the trade name halaxone, and H. D . D akin and co-workers have shown t h a t three parts of halazone per million suffice t o sterilize heavily polluted waters in 30 m in., a nd this concentration is effective in destroying pathogenic organisms. The effect of ammonia i n d estroying t,he bleaching activity and the property of oxidizing organic matter by h ypochlorite soln. is taken to prove that the nascent oxygen hypothesis fails to expla'in the retention of the bactericidal power of such soln., and it is attributed to the direct toxic action of chlorine or chloroamines. Bromine is used i n t he preparation of various chemicals-bromides, etc.employed in the manufacture of aniline dyes, and in' photography. The alkali bromides are used medicinally, and free bromine is used as a n oxidizing agent in analytical chemistry ; i n t he manufacture of ferricyanides, permanganates, et c. Bromine is also used as a d isinfectant or steriljzation agent, largelv i n t he form of hornurn solid$cutum, which is a m ixture of kieselguhr with a binding a,gent which is moulded i n t he form of rods, baked, and eat. with liquid bromine. The product contains up t o 75 p er cent. of its weight of bromine.19 Iodine is used'in various forms i n m edjcine--e.g. tincture of iodine, liquor iodi, iodized cotton, iodized wine, iodized water, oils and syrups ; iodides of potassium, mercury, iron, arsenic, lead, etc. ; a nd as methvl iodide or di-iodide ; iodoform, CH4 ; e thyl iodide, C 2H51; iodole, C&NH ; h istole ; etc.-largely for external application as a n a ntiseptic. Some iodides are used in photography, and in analytical operations ; a nd a considerable amount of iodine is used in the preparation of aniline dyes. REFERENCES. 1 J. A. Wanklyn, Chem. News, 20. 271, 1869 ; R. Cowper, Journ. Chem. S oc., 43. 153, I883 ; W. 9 Shenstone, ib., 71. 471, 1897 ; A. Gautier and G. Charpy, Compt. Rend., 113. 597, I891 ; . U. Kreusler, Ber., 24. 3947, I891 ; A. Lange, Zeit. angew. Chem., 13. 683, 1900 ; A. von Perentzky, Chem. Ztg., 32. 285, 1908 ; T. Andrews, Trams. Roy. Irish -4cad., 19. 398, 1842. a P. Hautefeuilloand J. Chappius, Compt. Rend., 98. 273, 1884. a M. Berthelot, Compt. Rend., 127. 795, 1898; J. Ogier, ib., 85. 957, 1877 ; G. Just and F. Haber, Zeit. Elektrochem., 20. 483, 1914. ' W. Henry, Phil. Tram., 90. 108, 1800 ; 99. 430, 1809. ' J. Thornsen, Journ. pakt. Chem., ( 2), 37. 105, 1888 ; E. Petersen, ib., (2), 46. 328, 1892; (2), 48. 88, 1893 ; G . Kruss and F. W. Schmidt, ib., (2),47. 301, I893 ; Zeit. anorg. Chem., 3. 421, 1893; F, Haber, ib., 51. 356, 1006 ; F. Meyer, Compt. Rend., 139. 733, 1905. R. Cowper, Journ. Chem. Soc., 43. 153, I883 ; W. A. Shcnstone, ih., Yl. 471, 1897. V. Merz and W. Weith, Ber., 6. 1518, 1873 ; V. Merz and E. Holzmann, ib., 22. 867, 1889. M. Berthelot, Bull. Soc. Chim., (2),43. 545, I885 ; Compt. Rend., 100. 761, 1886 ; T. S. Hum1 pidge, Ber., 17. 1838, 1884 ; N. N. Beketoff, ib., 14. 2052, 1881 ; P. Blau, Momtd~.., 7. 547, 1896; P. Lazareff, Journ. RussianPhys. Chem. Soc., 22. 383, 1890. H. Hertlein, Zeit. ~ h ~ Chem., 19. 289, 1896; S. U. Pickering, Journ. Chcm. Soc., 37. 128, a. 1880; R. H. Ashley, Ame~. Journ. Science, ( 4), 19. 237, 1904 ; ( 4), 20. 13, 1905 ; J. Bougault, Journ. Phat-m. Chim., (7), 16. 33, 1917 ; Compt. Rend., 164. 919, I917 ; E. J. MaumenB, i6., 89. 422, 1879; M. Berthelot, ih., 100. 773, 925, 971, 1888; M. J. Pordos and A. GBlia, i6., 15. 920, 1842 ; Ann. Chim. Phys., (3), 8. 349, I843 ; G . Lunge, Ber., 12. 404, I879 ; C. A. R. Wright, Chem. News, 21. 103, 1870 ; G . Topf, Zeit. anal. Chem., 26. 137, 277, 1887 ; J. P. Batey, Analyst, 36. 132, 1911 ; E. Abel, Zeit. anorg. Chem., 74. 395, I912 ; C. fiiedheim, Zeit. angew. Chem., 4. 415,1891 ; I. M. Kolthoff, Pharm. Weekbl., 56. 572, 1919. L. W Andrew~, . Journ. Amer. Chem. Soc., 25. 756, 1003 ; A. and P. DuprB, Liebig's Ann., M. 465, 1855. l1 J. L Gay Lussac and L. J. ThBnard, M km. Soc. Arcueil, 2. 295, 1809 ; Alembic Club Reprints, . 13, I897 ; IT. Ziiblin, Liebig's Ann., 209, 277, 1881. 0 l a H. Davy, Ann. Chim. Phys., (I), 92. 89, 1814 ; ( I), 9 . 289, 1815; H . Ziiblin, Liebig's Ann., 209. 277, 1881. VOL. 11. B 98 I NORGANIC AND THEORETICAL CHEMISTRY l 3 M. Bleslans, Compt. Red., 7 6. 92, 1873; 77. 781, 1873 ; H. Moissan, ib.,-118. 601, 1894; M. Berthelot and A. Gumt, ib., 99.7, 1884; Ann, Chirn. Phys., (6), 7. 138, 1886 ; W. G. Jixter, Amer. Journ. Science, (3), 4 363, 1893 ; R. Lucian, Chem. ZQ., 4 32, 1889 ; F. Fischor, Zeit. 6. . angew. Chem., 12. 764, 787, 1899 ; Journ. Gasbeleuchtung, 49. 419, 1906; J. Haussermam, Zeit. Eleldrochem., 8. 203, 1902 ; W. von Bolton, ib., 8. 165, 1902 ;9. 209, 1903 ; R. fiorenz, Zeit. angew. Chem., 6.313, 1893; Zeit. anorg. Chem., 10. 74, 1896; E. Barnes, Chem. News, 119. 260, 1919; F. P. Venable and T. Clarke, Journ. Amer. Chem. Soc., 17. 306, 1895 ; E. and H. Erdmann, Ber., 38. 237, 1905. 1. " Raich, Zed. phys. Chem., 2 124, 1888. . l6 A. W. Browne and F. F. Shetterley, Journ. Amer. Chem. Soc., 30. 53, 1908 ; E. Ebler, Analytische Operationen mit Hydroxylamin- u d Hydrazin-salzen, Heidelberg, 1905 ; Zeit. awy. Chem., 4 371, 1905 ; T. Curtius and H. Schulz, Yourn. pakt. Chem., (2),4 521,1890 ; R.Stall&, 7. 2. i b., (2), 66. 332, 1902. l6 H. Moisuan, Le$uor et ses curnposks, Paris, 1900 ; J. B. Senderens, Compt. Red., 1 04. 175, 1987 ; K. Bruckner, Monalsh., 2 . 341, 1906 ; A Duflos, Schweigger's Journ., 6 496, 1831. 7 2. l7 K. Herxheimer, Zeit. angew. Chem., 1 310, 1899 ; A Leymann, Concordia, 7, 1906. 2. I s W. Cruickshank, Nicholson's Journ., ( I),5. 202, 1801 ; T. Schwann, Milcroskqische Undersuchungen aber die Uebereimtimmung in der Textur u d dem Wachtum der Tiere und Wnzen, Berlin, 1839; E. Hermite, Brit. Pat. No. 5160, 1883 ; W. Webster, Engineer, 67. 261, 1889; H. BergB, Rev. d'Hyg., 2 . 905, 1900 ; A. C. Houston, F ifth Report o Royal Commission on Bewage 2 f Sanitary Imt., 31. 33, 1910 ; H. D. Dakin, Disposal, London, 1905 ; S. Rideal, Journ. ROW. J. 33. Cohen, M. Duafresne, and J. Kenyon, Proc. Roy. Soc., 89. B, 232, 1916 ; H. D. Dakin and E. K. Dunham, Brit. Med. Journ., 682, 1917 ; J. Race, Journ. Amer. Waterworks Assoc., 5. 03, 1918 ; Chdorination qf Water, New York, 1918 ; S. DelBpine, Journ. So,c. Chem. I d . , 29. 1350, f 1911 ; A . B. Hooker, Chloride o Lime in Sanitation, New York, 1913. ' 9 A . Frank, Dingkr's Jotwn., 2 47.514,1883 ; B. Merck, Phrm. Ztg., 50. 1022,1906. $ 13. Colloidal Iodine and Iodized Starch W. Harrison 1 p repared unstable hydrosols of iodine b y adding hydriodic acid t o a d il. s o h . of iodic acid ; a nd by adding alcohol t o s oh. of iodine i n glycerol. The hydrosol is a t f ist blue, b n t i t quickly flocculates forming a g rey a ggregate which, if the conc. be small, can scarcely be recognized. J. A mann h as p repared s oh. in s ulphuric acid a nd i n propylamine. The r elation between t h e colour and t he n ature of the solvent has already been discussed. D ry iodine does not colour d ry s tarch blue, b u t rather turns it brown, a.nd the brown gives place t o blue on c ontact with water. When i n c ontact with starch, i odine forms an intense blue-coloured product, iodized starch. U nder similar conditions, bromine forms a n i ntense yellow. According t o C. Meineke, t h e reaction with iodine is delicate enough t o reveal the presence of 0*0000003 grm. of iodme per c.c., while P. Mylius and G. J ust say 0.0000001 g rm. The great sensitiveness of this reaction makes starch a v aluable indicator for t h e presence of free iodine in v olumetric analysis. According t o F. C. Accum, starch w as recommended for the detection of iodine by P. S tromeyer. F. Mvlius, I?. E. H ale, a nd C. L ounes found t hat h ydrogen iodide and iodine are present i; iodized starch i n t h e r atio 1 : 4, a nd t he was formula (C241&00201)4HI assigned t o t h e blue product where the iodine and i o h d e a re r elated a s H-TZ14, etc. I?. S eyfert a nd J. T6th o bjected to F. Mylius' contention t hat a n iodide or hydriodic acid is necessary f or t h e production of t he b lue coloration. A n excess of p otassium iodide colours iodized starch brown, the brown colour becomes blue on dilution with water. T he r eddish-brown colour produced by t he a ction of iodine a nd a conc. s oh. of p otassium iodide on starch is also turned blue b y w ater. I Mylius suggested t hat t h e reddened starch contains ?. no potassium iodide, b u t twice a s m uch iodine a s t he blue product, b u t I?. E. Hale suggested t h a t since a conc. s oh. of p otassium iodide turns t h e blue starch red, and t he reddened product becomes blue when treated with water, it is m ore likely that t h e red product c ontains more iodide t han the blue. According t o E. W. Washburn, t h e presence of a lkali chlorides, magnesium sulphate, etc,, does not interfere. The colour o btained in t h e presence of potassium iodide is not t h e same i n t int as if other salts are present ; a nd t h e amount of iodine required t o produce a blue colour of T H E HALOGENS 99 the same intensity is smaller with soln. of potassium iodide than with s o h . of other salts of the same conc. F. DIylius maintains t h a t the presence of a n iodide, i .e. 1'-ion of a conc. a t least 10-6 grm. per litre, is necessary for the development of the blue colour, so that the starch s o h . is usually mixed with a n iodide-say iodide of potassium, zinc, etc. C. T o~nlinson found many other iodides served the same purpose. According t o W . H arrison, the addition of alcohol changes the blue colour of iodized starch to violet, red, orange, and yellow, and E. B audrimont found that the same colour changes are produced by heat, provided not too much iodine is present. The blue colour begins to leave aq. iodized starch a t about 40°, and disappears between 60" a nd 30". A s o h . of iodized starch heated above 70" is colourless if dil., and straw yellow if conc. On cooling, the original colour is restored if the s o h . be rapidly heated and quickly cooled, but with a prolonged heating, t h e intensity of the colour is feebler, and maybe does not return a t all. The colourless product is supposed b y L. Bondonneau, J. L. P . D uroy, and P. G uichard t o be a colourless organic iodide because the colour is restored by the addition of nitric &id or an iodate ; a nd with chloroform, it becomes violet. L. W. Andrews and H. M. Cottsch say that clear starch soh. prepared a t 150°, takes up in the cold iodine eq. to (CGH1005)121 ; a starch soln. heated with an excess of iodine to 100" for a short time forms ; (C6H1005)1212 a nd a starch s o h . heated to 100' for a long time gives a colourless soI11. containing most of the iodine as an organic iodide; some hydriodic acid and glucose are also formed. F. Mylius and F . W. K iister say that dried biue-iodized starch -forms a brown powder which becomes blue again in presence of water. The colour of iodized starch is so different from that of other iodine compounds that it h as attracted some attention. N. Blondlot, A . BBchamp, J . J. Pohl, R. Fresenius, E. D uclaux, B. Briickner, and F. W. K iister have supported the hypothesis that iodized starch is a mixture of starch and iodine, or a s oh. of t he latter in the former, while E. G. Rouvier, A. Payen, J. F ritzsche, L . Bondonneau, P. G uichard, E. S onstadt, H . Pellet, and F. Mylius hold it t o be a chemical compound of starch and iodine, and hence arose the term starch iodide. There is, however, no agreement as to the composition of the alleged compound since the amount of iodine is variously reported to be from 3 .2 t o 19'6 per cent., and the formulz accordingly extend over a wide range. A . Coehn showed that iodized starch is a negative colloid. F. W. K iister found the amount of iodine absorbed by solid starch is wholly dependent on the conc. of the soln., and he concluded that iodized starch is neither a compound nor a mixture, but is a welldefined solid s o h . of iodine in starch. M. Katayama also showed that, in dil. soln,, the conc. of t he iodine in the starch is proportional to the conc. of iodine and starch respectively, but varies in a somewhat complicated manner with the iodide conc. In very dil. s o h , it is proportional to the second or third power of the iodide conc., but the effect is proportionately smaller as the conc. increases, and depends also on the starch conc. He also draws a similar conclusion t o P. W. Kiister, L. W. Andrews, and H . M. Gottsch. Soln. of iodized starch give up considerable amounts of iodine to chloroform, but subsequent additions of chloroform give no sign of a partition coeff. as would be anticipat'ed if t,he iodine were lncrely dissolved in the starch. The vap. press. of iodine in iodized starch is very small after the removal o the first portion of iodine. M. Padoa and B. SavarE!measured the conductivities f o soh. of iodized starch, and although it was not possible to obtain a product with f a constant ratio between iodine and hydrogen iodide, they say that one definite additive compound, analogous with F . W . Kiister's assumption, is formed. W. Biltz also argued that other substances are coloured blue by the adsorption of iodine. For example, A. Dainour, R . J. Meyer, and N. A. OrlofT found this to be the case with lanthanum acetate, and basic praseodymium acetate ; C. G raebe also found euxanthic acid, cholic acid, and narceine behave similarly. Iodine is also adsorbed by a number of other substances forming brown instead of blue solid soln.--e.,q. J. W alker and S. A . K ay found such t o be the case with magnesia ; W. A . R . Wilks I NORGANIC A ND THEORETICAL CHEMISTRY I 00 f ound t h a t with slaked lime bromine and iodine dissolved in carbon tetrachloride forin adsorption products, while chlorine forms a chemical compound. K . E strup and E. B. Anderson studied the adsorption of iodine by precipitated barium sulphate ; a nd H. Siegxist and E. F ilippi found tannin and a number of organic bases t o behave similarly. G. C . S chmidt studied t he p artition of iodine between alcohol and animal charcoal, and found that t his is n ot in accord with Henry's law C/C,=constant, where C d enotes the quantity of iodine adsorbed by the charcoal, and C, t he quantity which remains in 10 C.C. of the solvent. 0. C. M. D avis studied the partition of iodine with different solvents and with different forms o charcoal, and found t h a t f with constant surface area, t h e amount of adsorption is specific, depending both on t he n ature of t h e solvent and of the adsorbent. I x denotes the amount of iodine f adsorbed when w grms. of solid are employed, the amount adsorbed per gram is represented fairly well by H. F reundlich's empirical formula : A mount adsorbed per gram, x/w=P/p, where /3 a nd p a re constants. 0. C. M. D avis found t hat Solvent. Toluene Benzene Ethyl acetate Alcohol Chloroform 1 Animal charcoal. 8 P I Sugar carbon. P P 1 Cocoanut carbon. . T he adsorption of iodine is approximately the same with animal and sugar charcoal, but cocoanut charcoal adsorbs only a fraction of this amount. The adsorption consists of a surface condensation, and a diffusion, owing t o solid soln., into the interior ; t he surface condensation is complete in a few hours, the diffwion may occupy weeks or months. The product obtained by saturating animal charcoal w ith iodine has been used as a medicament under the trade name iodccrztraco, and L. Corridi has investigated which solvents extract the iodine most readily-water is slow, dil. acids are slower, but dil. a lkali s o h . extract most of the iodine. K. Scheringe and P. Guichard have studied the adsorption of iodine vapour by sand, silica, opal, agate, alumina, magnesia, beryllia, and charcoal. E. F. Lundelius has also studied the adsorption of iodide from soln. in carbon tetrachloride, carbon ' disulphide, and chloroform by c,harcoal. REFERENCES. W. Harrison, KoEl. Zeit., 9. 5, 1911 ; W. A. R. W&, i b., 1 . 12, 1912 ; K. Est,rup and 1 E. B. Andersen, ib., 10. 161, 1912 ; J. Amann, ib., 6. 235, 1910 ; 7. 67, 1910 ; KoEl. Beiheft, 3. 337, 1912; S. U.Pickering, Jozcm. Chem. Soc., 42:311,1882; H. B. Stocks, Chem. News, 56.212,1887; 57. 183, 1888 ; E. Sonshdt, ib., 25. 248,1872 ; A. Vogel, Repert. Pharm., 22.349, 1873 ; 25. 568, 1875 ; E. Schar, Phurm. Centrh., 37. 540, 1895 ; C. Meineke, Chem. Ztg., 18. 157, 1894 ; J. Tbth, ib., 15. 1523, 1583, 1891; N. A. Orloff,ib., 31. 75, 1907; Ber., 37. 719, 1904; F. Mylius, Ber., 20. 688, 1887; 28. 388, 1895; A. L. Potilihin, i b., 13. 240, 1880; W Biltz, ib., 37. 719, 1904; . C. Graebe, ib., 33. 3360, 1900 ; F. W Kiister, ib., 28. 753, 1896 ; Zeit. anorg. Chem., 28. 360, . 1895 ; M. Katayama, ib., 56.209, 1907 ; It. J. Meyer, ib., 33. 31, 1903 ; G. Just, Zeit. phys. Chem., 63. 577, 1908 ; G. C. Schmidt, ib., 15. 5 5 1894 ; F. C. Accum, A Practical Esssay on Chemical 1, Reugents and Tests, London, 287, 1818 ; P. Goppelsroder, Pogg. Ann., 109. 57, 1863 ; C. M. v an Deventer, Maad. Naturweten., 14. 98, 1888; F. E. Hale, Amer. Chem. Journ., 28. 438, 1902; J. L. Lassaigne, Journ. Chim. Ned., ( 3), 7. 180, 1861 ; C. Tomlinson, Phil. Mug., (5), 20. 168, 1885 ; C. F. Roberts, A w . Joum. Science, (3), 47,422,1894 ; M.Padoa and B . SavarB, .4tli Acead. Lincei, (Fi), 14, i, 467, 1905 ; 17. i, 214, 1908 ; l . W Washburn, Journ. Amer. Chem. &x., 3. 30.31,1908 ; L. W. Anhews and H. M. GCttsch, ib., 24.865,1902 ;J. J. PoN, Journ. prakt. Chem , ( l), 83. 35, 1861 ; A. Lenssen and J. Lowenthal, ib., (I),86. 216, 1862 ; C. F. Schijnbein, ib., 385, I (I), & .1861 ; C. von Niigeli, L'Inst., 265, 1863 ; 33. Baudrimont, Compt. Rend., 51. 825, 1860 ; J. L. P. Duroy, i b., 51. 1031, 1860 ; E. Ducbux, ib., 74. 633, 1872 ; Ann. Chim. Phys., (4), 25. 264, 1872 ; N. Blondlot, ib., (3),43. 225, 1855 ; A. Girard, ib., (6), 12. 275, 1887; Journ. I'harm. Chim., (3),39. 45, 1861 ; A. BBchamp, ib., (3),27. M6, 1855 ; (3),28. 303, 1855 ; J. L. Soubeiran, ib., (3), 21 389, 1852 ; A. Seput, ib., (3), 21. 202, 1852 ; M. Magnes-Lahens, ib., (3), 19. INORGANIC AND THXORl3TICAf; CHEMISTRY chloride and silver. J. S. Goldbaum electrolyzed a soh. of sodium chloride with cu, weighed silver anode and mercury cathode. Sodium as cation forms an amalgam, and thc chlorine attaches itself to the silver of the anode, and the increase in weight represents the halogen content of the salt. A correction was maae for the silver which dissolved from the anode. He obtained C1535.456, when referred to sodium =23.00. F. W. Clarke's calculation of the ratio Ag : KC1=100 : x furnished as a general mean 69-1138f 0.00011 for the amount of potassium chloride which unites with 100 parts of silver to form silver chloride ; or 52.0163f0.00018 parts of potnsuium chloride furnish 100 parts of silver chloride. The first ra.tio gives the relation between the mol, w t. of oxygen as standard and the mol. wt. of potassium chloride ; t he second gives the relation between potassium chloride and silver ; a nd the third the relation between silver and chlorine. 3. S. Stas' work on at. wt. has been deservedly eulogized. For many years it was considered to be so near perfection as was possible to man. J. S. Stas seemed to have taken the most subtle precautions to exclude errors of manipulation, and to ensure the purity of his materials. He also followed the advice of J. J. Berzelius, for, in order to eliminate constant errors, he used materials from different sources, and followed many different paths in arriving at his results. Only when consistent values were obtained by different methods did he assume that the results were reliable. The following outline will give an idea of the plan of J. 8 . Stas' work : 1. Determination o the rela'tion between sodium chloride and silver.-Ten f samples of a known weight of purified silver were dissolved in nitric acid, and each was treated with purified sodium chloride. The excess of silver was determined volumetrically and the precipitated silver chloride was reduced to metallic silver and weighed. The best representative value of these determinations was : 54.2078 parts o NaCl were required for 100 parts o silver. f f 2. Determirtatwn of the relation between chlorine and silver.-Silver cbloride was prepared from purified silver in several different ways : (1) By burning the metal at a red heat in chlorine gas; (2) by dissolving the silver in nitric acid, and ~recipitating silver chloride (a) with hydrogen chloride, (b) with hydrochloric acid, and (c) with ammonium chloride (in this process some silver chloride was lost in washing the precipitate). The best representative value of this work was : 132.8445 parts of silver ch.bride was obtairlzedfrom 100 parts o s ilu~r. By calculation it follows (1)t hat 100 parts of silver combine with f 132.8445 less 100=32'8445 parts of chlorine ; (2) that 100 parts of silver combine with 52.2078-32.8445=21a3633 parts of sodium. Or, 3. Determination o the reluiion between silver chloride and oxygen.-This f gives the relation between silver, sodium, and chlorine. I t he relation between f hydrogen or oxygen and any one of these elements be known, the at. wt. o f these elements with respect to hydrogen or oxygen as standards follows directly. This relation can be determined by converting silver chlorate into silver chloride: 2AgC103=2AgC1+302. A known weight of purified silver chlorate was decomposed by sulphurous acid : AgC103+ 3H2S03=AgCI +3&S04. The chlorates were also decomposed by ignition. As a result it was found that 74'9205 parts o silver chloride zacre equivalent to 100 parts of silver c7hlornte ; or 25.0795 parts f of oxygen were cq. to 100 parts of silver chlorate ; or 25'0795 parts of oxygen are eq. to 74-9205 parts of silver chloride. Bince A gCIO,=AgCl+ 3 0 , if 0= 16, i t follows that the eq. w t. x of silver chloride must b e 2 5.0795 : 8 + 16=74-9206 : x, where x= 1 43.395. B ut 100 p arts of silver are eq. to 3 2.8445 p arts o chlorine, and hencc 1 32'8445 p arts of silvcr chloride w ill f correspond with 1 32'8445 : l 43.395= 100 : x . or x = 1 07.942 p arts of silver. Similarly for chlorine, since 1 32-8445 p arts of chlorine unite w ith 1 07.942 parts of silver, 1 43.395 p arts of silver chloride will c ontain 3 5 4 5 3 p arts of chlorine. Hence, i f 0 = 16, C l= 3 6.463. Again for sodium, 3 2-8446 pasts of chlorine combine with THE HALOGENS 2 1.3633 parts of sodium, and hence 3 6.463 parts of chlotino unite with 23.0599 parts of sodium. Hence, if 0=16, the combining weights of silver, sodium, and chlorine are related as Similar experiments were made with halogens bromine and iodine, and the remarkable accuracy of J. 8. Stas' work may be judged by comparing his 1865 values with those employed a t the present day : . . Stas' 1865 values International table, 191 8 Chlorine. 3 5.457 35'46 Bromine. 70.955 79.92 Iodine. 126.850 126.92 Silver. 107-83 107.88 J. S. Stas also repeated the work with lithium, sodium, and potassium ; incidentally, the value for nitrogen was determined. The values deduced in this way were : . .Pot,assiuzu. 3 9.14 Stas' 1865 values International table, 191 8 ' 3 9.10 Sodium. 23-04 23.00 Lithium. 7.03 6.94 Nitrogen. 14.04 14.01 T he chief weakness in J. 8. Stas' work arose (1) from the difficulties in manipulating the unusually large amounts of material employed in each determination, and ( 2) precipitating his compounds from too conc. soln. in order t o keep down the bulk of the liquid. This led to the adsorption of relatively large amounts of the precipitant by his precipitates. Indeed, says T. W. Richards (1911),6 the presence of residua1 water and the loss of traces of " insoluble " p recipitates by dissolution cluring filtration, have perhaps ruined more at. wt. determinations than any other two causes-unless indeed the adsorption of foreign substances by precipitates may be ranked as a.n equally vitiating effect, Much of Stas' work has been revisedchiefly by T. W. Richards and his co-workerewith the idea of providing for the small errors whic.h affected Stas' work. It may here be emphasized 7 t hat different results may be obtained by varying the method of calculating at. wt. The ratios actually measured are affected by unavoidable errors of experiment, and during the calculai~ion these errors may be ibst'ributed over the several ratios concerned, or they may all accumulate upon that value last determined. Thus, in calculating the at, wt. of fluorine from the obscrved ratio, CaF2 : CaS04, the errors of experiment are superposed upon the errors involved in obtaining the at. wt. of calcium and sulphur when oxygen 16 is the standard of reference. I n the ideal case,,the observed data should be so treated that the errors are properly distributed among the different weighings, and their influence red~rcedto a minimum. The mat,hematical operations are well understood, and the laboratory should furnish the necessary data. II. T he quantitative synthesis of hydrogen chloride.-H. B. Dixon and E. C. Edgar8 prepared chlorine by the electrolysis of fused silver c.hloride and weiphed it in t'hc liquid form ; hydrogen was prepared by the electrolysis of a soln. of barium hydroxide and weighed occluded in palladium. The hydrogen was burnt in a globe filled with chlorine, and the excess of chlorine determined by absorption in potassium iodide and titration of the liberated iodine with sodium thiosulphate. The mean of nine determinations was 35'463(H=1*00762). W . A . Noyes and H. C. P. Weber passed hydrogen weighed in palladium over heated potassium chloro-platinate. The loss in weight of the fatter salt gave the weight of chlorine employed. The hydrogen chloride produced was also collected either by absorption in water, or condensed to a solid by cooling with liquid air. The mean of twelve determinations gave 35%2(H=l). E. C. Edgar then synthesized hydrogen chloride by weighing the hydrogen, chlorine, and hydrogen chloride. The latter was freed from the excess of chlorine condensed along with the hydrogen chloride by allowing the liquid t o evaporate ; passing the vapour through a quartz tube filled with mercury vapours ; INORGANIC AND THEORETICAL CHEMISTRY and condensing the purified hydrogen chloride, back to a liquid, or dissolving it i n water. Result, 35.461. R . W. Gray and F. P. B urt first determined the weight of a normal litre of hydrogen chloride to be 1.63885 grms. ; t hey then passed the gas over heated alumininm, and measured the volume of the liberated hydrogen. Two volumes of hydrogen chloride gave 1.00790 vols. of hydrogen. With Morlev's value for the density of chlorine, it was found that 36.4672 grms. of hydrogen chldride give 35'4594 grms. of chlorine, if the unit of hydrogen be 1.0078 grm. P. A. Guye and C . Ter-Gazarian's determination of the relative density of hydrogen chloride also gave 35.461 for the at. wt. of chlorine. 3. Deutsch obtained a rather lower value. 1 1 The a nalysis of nitrosy1 chloride.--P. A . Guye and G. Pluss 9 first distilled a 1. known weight of purified nitrosyl chloride, NOCI, over heated silver-this retained the chlorine ; i t was then passed over heated copper-this retained the oxygen ; a nd finally it was passed over metallic calcium--this retained the nitrogen. The sum of the weights of chlorine, oxygen, and nitrogen so determined was usually a little less-up to 0.0012 grm.-than the weight of the nitrosyl chloride employed. The mean val'ue for chlorine gave 35.468. IV. P hysical methods.-Several physical methods have been employed for determining the mol, w t. of hydrogen chloride. The density of hydrogen chloride determined by R . W. Gray and It'. P. B urt 10 was 1'62698 grms. for a normal litre ; a nd 1.42762 for oxygen, Hence, with the mol. wt. of oxygen, 32, as the standard of reference, the mol. wt. x of hydrogen chloride is given by the proportion 1.42762 : la62698=32 : x, where x=36.469 ; with the at. wt. of hydrogen 1'0076, the ah. wt of chlorine is 36.469-19076=35.461. The compressibility coeff.-that is, the mean deviation of hydrogen chloride from Boyle's law-required for an application of the D. Berthelot's method of limiting densities (1. 6, 8), has been determined by A. Leduc to be Aol=O-00758 ; E. Briner, 0.00750 ; R. W. Gray and F. P. B urt, 0'00743 ; a nd ,4. J aq~zerod and 0. Scheuer found' for oxygen Ao1=0'00097 ; R. W. Gray and F. P. B urt, 0.000964. Hence, A. Leduc c,alculated the mol. wt. of hydrogen chloride f to be 36-460 ; a nd E. Briner, 36.462. A. Leduc's n?uetAod o molecular v o l u m , where it is assumed that all gases a t correspouding temp. and press. have the same mold vol., gave the value 36.450 for- the mol. wt. of hydrogen chloride; and E. Briner, 36.453. The method by the reduction o tins critical constants, based on f the equation : Mol. wt, referred to oxygen 32 is equal to 22'412W/(l+a)(l-b), where W denotes the weight of a litre of gas a t 0' and 760 mm., reduced to sea-level, and latitude 45". P. A. Guye and G. Ter-Gazarian 11 found for a normal litre of hydrogen chloride, W=1%398 ; a nd (l+ao)(l -bo)=l~00773, so that the mol. wt. 01 hydrogen chloride is 36.4393, and if the at. wt. of hydrogen is 1'0078, that of chlorine will be 35.4615. Starting with oxygen 16 as the standard of reference, F. W. Clarke (1910) regards 35'4584 0.0002 as the best representative value for the at. wt. of chlorine if the at. wt. of the silver be 107'880-0.00029 ; a nd B. Brauner (1913) considers the best representative value to be 35.457, if silver be 107.880 ; 35'456, if silver be 107.876 ; a nd 35.454, if silver be 107-871. B. Brauner also inquires : W hat reliance can be placed on the third decimal ? a nd answers that the uncertaintv is smaller if silver be regarded as a fixed basis with respect to oxygen 16, and g e i t e r when the at. wt. is referred to oxygen 16 alone regarded as a fixed constant. I referred to silver f fixed a t 107'876, the at. wt. of chlorine lies between 35.454 and 35.458, or 35'456 f0.0019 ; b ut if the value for chlorine is made to depend upon the cycle of re1a tIons ' between it and oxygen=16, " t he at. wt. of chlorine lies between 35:453 and 35.459, or possibly 35.452 and 35.460 ; " or the at. wt. of chlorine is 35.456 0.003, or possibly 35'456 0.004. B. Brauner illustrates the idea geometrically by means of Pig. 22, which he says makes it much clearer than is possible by mathematical symbols, how the greater the distance right or left from the middIe line corresponding with Ag=107?376, C1=35'456, the smaller the probability that the corresponding value for the at. wt. will be correct. The international Galue for the a t , wt. of chlorine is 35'46. - + + + T HE HALOGENS It is astonishing what a vast amount of labour has been expended in the struggle for increased accuracy in the determination of the at. wt. ratios. Some of the later determinations are masterpieces of precision. It is probable that the majority of these researches has been directed towards the chlorine : oxygen or the chlorine : hydrogen ratio on account of its fundamental importance. Even now J. F. W. Herschel's words 12 a re not inapplicable : I t is doubtful whether such accuracy in chemical analysis has yet been attained as to enable us to answer positively for a fraction not exceeding the three or four-hundredth part of t he whole quantity to be determined ; a t least, the results of experiments obtained with the greatest care'often differ by a greater amount. Many have had misgivings as to the utilitarian valuc of the enormous labour which has been expended in this direction, and particularly when the best representative values of all the best results are usually rounded off to the nearest tenth when the at. wt. are employed in chemical calculations. Lord Kelvin's words are often quoted as a stimulus t o greater and still greater precision : Accurate and minute measurement seems to the non-scientific imagination a less lofty and dignified work than looking for something new ; discoveries o science have been the f reward of accurate measurement, and patient, long-continued labour in the minute sifting of numerical results. The eighty or more individual numbers we call at. wt., a dds T. W. Richards, " a re perhaps the most striking of the physical records which Nature has given us concerning the earliest stages of the evolution of the universe. They are mute witnesses P I G . 2 2 . B . Brauner's Table of the Relation of the Atomic Weight of Chlorine to that of Silver (Oxygen 16). of the first beginnings of the cosmos out of chaos, and their significance is one of the first concerns of the chemical philosopher." We are now promised atoms of one element of different weight ; so that the observed at. wt. is a kind of average. If t he atoms of different weight could be separated, the fractions would occupy the same position on the periodic table, and be chemically identical-they are called isotopes, and the s u b j e ~ t discussed in is Vol. 1 1 According to the positive ray spectrograph, F. W. Aston l3 obtained 1. results which show that chlorine contains two isotopes of at. wt. 35 and 37 ; a nd W . D. H arkins has claimed that in the atomysis of hydrogen chloride the density of the fraction which remains in the diffusion tube increases a t a rate corresponding with the assumption that the chlorine isotopes have at. wts. 35 a nd 37-with ~ossibly a third of at. wt. 39. F. W. Aston added : At first sight it may seem incredible that chIorine, whose chemical combining weight hat3 been determined more often and with greater accuracy than almost any other element, should not have given evidence of its isotopic nature in the past ; but it must be remembered that, in d l probability, every one of these determinations has been performed with chlorine originally derived from the sea in which the isotopes, if ever separate, must have been perfect'ly mixed from the most remote ages. Chlorine from some other source, if such can be found, may well give a different result, as did radio-lead when examined. T he atomic w eight of bromine.-The at. wt. of bromine has been determined by methods which follow in principle those employed for chlorine. A. 3 . Balard (1826),l*t he discoverer of bromine, transformed a knownweight of potassium bromide into the sulphate, and also reduced silver bromide to metallic silver by means of zinc ; t he numbers 74.7 and 75.3 were respectively obtained. J. von Liebig (1826) 106 INORGANIC AND THEORETICAL CHEMISTRY also transformed potassium bromide into silver bromide, and obtained 75'2 for the at. wt. of the element ; C. Lowig (1829) obtained 75'76. These numbers are veryOW ; t his is, no doubt, due to the impurities present in the salts. J. J. Berzelius (1828) converted silver bromide into the chloride by the action of chlorine gas, and obtained a value tr& approche'e du but, namely, 79.36. J. B. A. D umas (l859),by the same method, obtained 79.95. W. Wallace precipitated silver bromide from arsenic tribromide, AsBr,, by the addition of silver nitrate, and obtained an at. wt. of 79'738, when the at. wt. of silver is 107.97, and arsenic 75. The preceding determinations are usually disregarded in modern estimates of the at. wt. of bromine because of the then imperfect state of the art of chemical analysis as involved in the work. J. C. G. de Marignac 15 decomposed potassium bromate by careful calcination, and precipitated the bromine from the resulting potassium bromide by treatment with silver nitrate. J. S. Stas reduced silver bromate by treatment with sulphurous acid ; th'e ratio of silver to bromine was also determined by J. C. G. de Marignac and by J. S. Stas either by synthesizing silver bromide from its elements ; or by converting potassium bromide into silver bromide. J. S. Stas' value, 79'9628 f0.0032, has been recalculated by F. W. Clarke and by J. D. van der Plaats, who obtained respectively 79-951 and 79.955. A. Scott obtained values varying from 79.899 to 79'911 from his analysis of ammonium bromide. J. 8. Goldbaum (1911) by the electrolytic method used for the ratio Na : Cl, obtained for the at. wt. of bromine 79.927, when the at. wt. of sodium is taken as 23.00. G . P. B axter (1906) dissolved the purest silver in nitric acid, precipitated silver bromide by th? addition of ammonium bromide, and finally fused the washed and dried product in bromine vapour. He obtained values ranging from 79.914 to 79-918-average 79.915-as the best representative value of his determinations. By collecting together all the direct and incidental determinations of the silver-bromine ratios since about 1843, P. W. Clarke obtained 79.9197 as the best representative value for the at. wt. of bromine (silver=107.880, oxygen=16) ; B. Brauner obtained 79-916, if silver be 107'880 ; 79'913, if silver be 107,876 ; a nd 79.909, if silver be 107.871 ; t he error in the third decimal place is estimated to be about 0.004. P. A. Guye, E. ]r\l!toles,C. K. Reiman, and W. J. Murray have computed the at. wt. of bromine from determinations of the density and compressibility of hydrogen. bromide, The results lie between 79.924 and 79.926, when H=1.0076. The International table gives 79'92 as the best representative value when silver is 107.88. The atomic w eight of iodine.--In his historic M6rnoire sur l'iode (1814), J . L . Gay Lussac 16 a ttempted to determine the at. wt. of iodine by the synthesis of zinc iodide, and in this manner he obtained the value 125 for the constant. H. Davy obtained 132 by converting sodium hydroxide into iodide. By a n analogous process to that eniployed by J. L. Gay Lnssac, W. P rout obhined 126. I n 1825, T. Thomson obtained 124 as a result of decomposing potasgium iodide. I n 1828, J. J. Berzelius converted silver iodide into the chloride by the ac,tionof chlorine on the heated salt, and obtained values ranging between 126.26and 126.39; a nd by the same method, J'. B. A. D umas obtained the value 126.59. I n 1843, N. A. E. Millon determined the percentage of oxygen in potassium iodate and silver iodate and obtained respectively 126.697 and 125.33. The preceding dete.rnfinations are usually disregarded in .modern estimates of the at. wt. of iodine because " t he art of qualitative analysis was then in its infancy." J. C. G. de Marignac showed that the ignition of the iodate is not suitedfor the determination since some iodine is l o ~during thecnlcination, and he preferred thesynthesis t of silver iodide by dissolving a known weight of silver in nitric acid and precipitating thc ccmtained silver a.g iodide, by the addition of potassium iodide. The result furuished 126.537 and 126.550. J . S. Stas 17 analyzed silver iodate and determined both the water and the oxygen given off i n eac,h calcination, as well as the amount of silver iodide. He also reduced the iodate to iodide by sulphuro~isacid, and synthesized silver iodide by precipitation from silver nitrate by means of hydriodic acid, and by t reating silver sulphate with ammonium iodide. The results varied THE HALOGENS from 126.85 t o 126.864-average 126.855. Both J. C. G. de Marignac's and J. S. Stas' results were affected by constant errors, the chief one being due to the occlusion of silver nitrate by the precipitated silver iodide. I n 1902, A. Ladenburg claimed that the value 126.85 was about one-tenth too low ; a nd soon afterwards, A. Scott obtained the value 126.912 as a result of two syntheses of silver iodide. P. Kijthner and E. Aeuer synthesized silver iodide by heating the metal in iodine vapour, and by precipitation from silver nitrate by the addition of hydriodic acid, taking precautions t o eliminate the occlusion of mother-liquor. As a result, they obtained 126.936 (oxygen 16). A very careful series of determinations were made by G. P. B axter about the same time. He precipitated silver iodide from silver nitrate by treatment of the soln. with ammonium iodide in the presence of an excess of ammonia ; a nd also by converting iodine into ammonium iodide and precipitating the silver iodide by the addition of silver nitrate, taking precautions to avoid an excess of the latter reagent. As a result of the whole work, 126-929 was obtained for the at. wt. of iodine. G. P. B axter and G. S. Tilley also measured the ratio of silver to iodine pentoxide, and found 0.646230, which makes the at. wt. of iodine 126.891. M. Guichard decomposed iodine pentoxide into oxygen and iodine by heat ; t he former was absorbed ,by red-hot copper, and the iodine condensed. He found that if oxygen has an at. wt. of 16, iodine is 126.915. G. Gallo electrolyzed a soln. of silver salt so that silver was deposited on the cathode, and the iodine liberated a t the anode by titration with sodium thiosulphate. His determinations varied from 126.82 t o 126.9&average 126.89. Summing up the various determinations, F. W. Clarke obtained 126.920 fO.OOO33 a s the best representative value for the at. wt. of iodine ; a nd B. Brauner 126.932, if silver be 107.880 ; 126.927, if silver be 107.876 ; a nd 126.921, if silver be 107.871. It is very probable that the at. wt. is greater than 107'870 (oxygen 16) and smaller than 107'880, consequently, says B. Brauner, " t he uncertainty in these values for the at. wt. of iodine does not extend to many units in the third decimal place." This makes the at. wt. of iodine 126.93. The International table gives 126.92 as t he best representative value. The anomaly in the at. wt. of iodine and tellurium with respect t o their position in the periodic table has greatly stimulated researches on the at. wt. of these elements. It h as been asked : Does iodine contain an undiscovered halogen element of higher atomic weight than iodine ? I n answer, G. P. B axter converted iodine into hydriodic acid by hydrogen suliphide, and the hydriodic acid was converted back t o iodine by distillation with potassium permanganate in small quantities at a time so s s to obtain four fractions. No difference could be detected in the different fractions. If a halogen element were present in iodine with properties to be expected from the analogies with other members of the halogen family, it should have accumulated in the first fraction. Hence it is unlikely that iodine contains a halogen element of higher at. wt. t han iodine. E . Kohlweiler, however, claims to have evidence of the existence of iodine isotopes. The molecular weights of iodine, bromine, a nd chlorine.-When the vapour density determinations of all known volatile chlorides are collected together, a table illustrated by the excerpt Table X I is obtained. The smallest combining weight of chlorine in any one of these compounds corresponds with the combining 35.46oxygen=16-and accordingly this number is taken to represent the at. wt. of chlorine. The at. and eq. wt. of chlorine have the same numerical value. Similar results are obtained with the volatile fluorides, bromides, and iodides. The results show that the best representative values for the at. wt. of the halogens are F1, 19.0 ; C1, 35'46 ; B r, 79'92 ; a nd I, 126.92. The vapour densities of the halogens correspond with diatomic molecules; a t elevated temp., as already shown, there are signs of dissociation into monatonlic molecules, and this inore particularly with iodine and bromine than with chlorine and fluorine. E . P aternb and R. Nasini 18 have determined t he mol. wt. of bromine in aq. and acetic acid soln. by the f.p. INORGANIC AND THEORETICAL CHEMISTRY method, and the results correspond with the formula Br2. Iodine in dil. benzene soln. corresponded with the molecule I%, in more conc. s o h , the molecule seemed but to be more complex. I n acetic acid soln, the results were intermediate between those corresponding with mono- and di-atomic molecules. There has, however, been much discussion on the molecular condition of iodine in different solns. Volatile chloride. Vapour density. Formula o comf pound : Mol. wt. = vapour density. Amount o chlorine f in the molecule. -- . . . . Hydrogen chloride Chlorine Mercuric chloride Arsenic trichloride Tin tetrachloride Phosphorous pentachloride :! -1 36-5 70.9 273% 182.1 260.2 208.3 HCl GI2 HgCL AsCI, SnClp PCl, 35.46 70.92 70.92 106.38 141.84 177.30 T he valency of chlorine, bromine, and iodine.-Compounds are known in which the three halogens act as uni-, ter-, p i q u e - , or septa-valent elements. Usually, however, these elements are univalent. In chlorine dioxide, C102, the chlorine & bi- or quadri-valent,19 I n M. Berthelot's hydrogen perchloride, HCl,, the chlorine is probably tervalent, and R. Meldola (1888) showed that the oxygen in the hydrochloride of methyl oxide is best regarded as quadrivalent, the chlorine tervalent ; thus, (CH3)?: 0 : C1.H. Iodine also appears to be tervalent in the so-called iodonium compounds. When chlorine is passed into a chloroform soln. of iodobenzene, C6H51, an addit.ion compound phmyliododichloride, C6H,I.C1,, is in which the halogen atoms are probably tervalent. When this compound is treated with alkali hydroxide, iodo80benzene7 C6H510, is formed; and when this compound is boiled with water, i t produces iodoxybenzene, C,H,IO,, and iodobenzene, C6H,I. By the action of silver hydroxide upon a mixture of eq. amounts of iodosobenzene, C6H510,and iodoxybenzene, C6H510,, followed by treatment of the clear filtrate with potassium iodide. V. Meyer prepared the so-called : diphenyl iodonium hydroxide, (C6H6I2 I .OH. The precipitated hydroxide forms an iodide, (C6H,), : 1.1, which crystallizes from alcohol in yellow needle-like crystals melting between 175" and 176'. The iodonium bases and salts resemble those of lead and silver but particularly those of thallium. The colour and solubility of the halides resemble the corresponding salts of these met&. The chloride is white, the bromide pale. yellow, and the iodide yellow. The hydroxide and carbonate are soluble in water, and the soln. give an alkaline reaction, as is the case with the corresponding thallium salts. Diphenyl iodonium hydroxidegives a precipitate with ammonium sulphide, which looks like freshly precipitated antimony sulphide, and i t consists mainly of a trisulphide-possibly I -S - S - S --I =(C6H6),. The nitrate, (C,H,),I.NO,, acid sulphate, chromate, periodide, and other salts have beeh prepared. The base in question seems to be a derivative of an hypothetical iodonium hydroxide, HO -I - H2, analogous with hydroxylmnine, H O - N = Hz. Monoiododiphenyl. iodonium derivatives have been prepared by treating iodosobenzene, C,H,IO, with sulphuric acid a t a low temp., and afterwards treating the dil. soln. with potassium iodide. The iodide, (C6H5)I=I-C6H:,-I, and the other salts, as we1l.a.s the base, (C6H5)OH=I-CgHb-I, resemble the correspondmg salts of diphenyliodionium. Compounds of the type KBr3 are usually supposed to contain a tervalent halogen ; t he chlorates, bromates, and iodates, and compounds of the type &I5, t o .contain quinquevalent halogens ; and the perchlorates and periodates and compounds of the type CsI,, to contain septivalent halogens. F. W. Clarke, A Recalculation of the Atomic Weight&, Washington, 1910 ; B. Braunor, Abegg's finndbltch der anorgandschen Chemie, Leipzig, 4. ii, 50, 1913 ; 5. D. van der Plaats, Ann. Chim. P hys., (6), 7. 499, 1886. 5. 5. Berzoliug Pogg. Ann., 8. 1, 1826 ; T. Thomuon, Ann. Phil., 15. 89, 1820. THE H ALOGENS a F Penny, Phil. Tram., 129. 30, 1839 ; T. J. Pelouze, Compt. Rend., 15. 959, 1842 ; C. Gerhardt, ib., 21. 1280, 1845 ; P. A. Guye and G. Ter-Gazarian, ib., 143. 41 1, 1907 ; J. C. (2. de Marignac, Liebig's Ann., M.11, 1842 ; L. MaumenC, Ann, Chinz. Phis., (3),18. 71, 1846 ; V. Faget, db., (3), 18. 20, 1846 ; J. 8. Stas, Mdm. Acad. Belgique, 35. 3, 1865 ; A. Stikhler and F. Meyer, Zeit. a?iorg. Chem., 71. 378, 1911 ; T. W. Richards and H. H. Willard, Journ. Amer. Chem. Sot., 32. 4, 1910. E. Turner, Phil. Trune., 119. 291, 1829 ; 123. 529, 1833 ; F, Penny, ib., 129. 28, 1839 ; J. C. G. de Marignac, Liebig'a Ann., 44. 11, 1842 ; J. B. A. Dumas, Ann. Chim. Phy,?., (3), 55. 134, 1859 ; J. S . Stas, Mdm. Acad. BelgQue, 35. 3, 1865 ; T. W. Richards and R. C. Wells, A Revision 0f the Atomic Weight o Sodium and Chlorine, Washington, 1005, . f J. C. G. de Marignac, Liebig's Ann,, 44. 11, 1842 ; T. J. Pelouze, Compt. Relzd., 20. 1047, 1845 ; J. B. A. Dumas, Ann. Chkm. Phys., (3), 55. 134, 1859 ; J. S. Stas, Mkm. A d . Belgigue, 35. 3, 1865 ; L. MaumenO, Ann, Chim. Phys., (3), 18. 41, 1846 ; A. Thiel, Zeit. anorg. C.bm., 4 . 0 313, 1904 ; A. S cott, Jozcrn. Chem. Soc., 79. 147, 1901 ; T. W. Richards and R. C. Wells, A Revision o the Atomic Weight o Sodium and CAZorine, Washington, 1905 ; T . W. Richards and f f E. H. Archibald, Proc. Amer. Acad., 38. 456, 1903 ; E. H. Archibald, Trans. Roy. Sot. Canada, 3. 47, 1904: ; T. W. Richards and A, Staehlcr, Furlher Researches concerning Atomic Weights, Washington, 1907; T. W. Richard*, P Kiithner, and E. 'Siede, Journ. Amer. Chem. Soc., 31. 6, . 19@9 J. 8. Goldbaum, ib., 33. 35, 1911. ; T. W, Richards, Journ. Chem, Soc., 99. 1201, 1911 ; N ~thods used in P,recise Chemical Investigation, Washington, 1910. F. W. Clarke, Amer. Chem. Journ., 27. 321, 1902. H. B. Dixon and E. C. Edgar, Phil. Trans., 205. A, 169,1905 ; E . C. Edgar, ib., 209. A. 1,1908 ; W. A. Noyes and H. C. P. Weber, Journ. Amer. Chem. Soc., 30. 13, 1908; R. W. Gray and P. P. B urt, Journ. Chem. Soc., 95. 1633, 1909; Proc. Chem. Soc., 24. 215, 1908; R. W. Gray, ib., 23. 119, 1907 ; P. A, Guye and G. Ter-Gazarian, Compt. Rend., 14.3. 1233, 1906 ; 11. Dentsch, Essai sur une nouvelle mdthode p o u ~ d&termimtiondu poids atomique du chlore, Genhe, 1905. la P. A. Guye and G. Fluss, Journ, Chim. Phys., 6. 732, 1908. l o R. W. Gray and F. P. B urt, Jwurn. Chem. Soc., 95. 1633, 1909 ; A. Leduc, Recherches sur les gaz, Paria, 1898 ; Compt. Rend., 140. 642, 1905 ; P. A. Guye, Journ. Chim. Phy?., 3. 321, 1905 ; 0. Scheuer, Zeit. phys. Chem., $8. 575, 1909; E. Briner, Joz1.m. Chim. Phys., 4. 476, 1006, A. J aquerod and 0. Scheoer, Mkm. Soc. Uenkve, 35. 659, 1908 ; K . Olschewsky, M o ~ l s h . ,5. 127, 1884. l1 P. A. &ye, Cmpt. R e d . , 138. 1213, 1904; Jozrrn. Chim. Phys., 3. 321, 1905; G. TcrGazarian, i b., 7. 337, 1910 ; 9. 101, 1911. 1 2 J. F W. Hergchel, A Preliminary Discourse on the Study of Natural Philosophy, London, 307, . 1851 ; T, W. Richards, Journ. Chem. Soc., 99. 1201, 1911. 1 l 3 F. W. Aston, Phil. Mag., (6), 39. 620, 1920 ; Science P r ~ r e s s , 5, 212, 1920 ; N atwe, 104. 393, 1919; D. L Chapman, ib., 105. 487, 611, 1920; 106. 9, 1920; F. Soddy, ib., 105. . ,516, M3, 1920 ; T. R. Merton and H. Hartley, ib., 105. 104, 1920 ; A. F. Core, ib., 105. 592, 677, 1920; W. D. Harkins, ib., 105. 231, 1920; Jcience, 51. 289, 1920 ; E. Kohlweiler, Zeit. phys. Chem., 95. 95, 1920. l4 A. J. B alard, Ann. Chim. P h p ~ . , (2), 32. 337, 1826 ; J. B. A. Dumas, ib., (3), 55. 162, 1859; J. von Liebig, Schweigger's Journ., 4 . 108, 1826 ; J. J. Berzelius, Pogg. Ann., 14. 566, 1828 ; 8 C. Lowig, Das Brom und seine chemischn Verhaltnisse, Heidelberg, 1829 ; W. Wallace, Phil. Mag., (4), 18. 279, 1859. 1. " C. G. de Marignac, Bibl. Univ. QenJve, 48. 357, 1843 ; J. 8. Stas, MCm. A d . Belgique, 35. 3, 1865 ; G. P. B axter, Journ. Amer. Chem. Soc., 28. 1322, 1906 ; J. 8. Goldbaum, ib., 33. 35, 1911 ; A. Scott, Journ. Chem. Soc., 79. 147, 1901 ; F. W. Clarke, Chm. News,, 49. 68, 1883 ; Phil. Mag., (5), 12, 101, 1881 ; J. D. v an der Plaats, Ann. Chim. Phys., (6), 7. 499, 1886; P. A. Guye, Journ. Chim. Phys., 14. 361, 1916 ; 17. 171, 1919; E. Moles, i b., 14. 389, 1916; Compz. Rend., 162. 086, 1916 ; 163. 94, 1916 ; C. I<. Reiman, ib., 164.44, 180, 1917; W. J. Murray, ib., 104. 182, 1917. 18 J. L . Gay Lossac, Ann. Chim Pliyfi., (1), 91. 5 , 1814; N. A. E. Millon, ib., (3), 9.407, 1843 ; 5 . B. A. Dumas, ib., (3), 55. 163, 1859 ;W. ProuC,, Ann. Phil., 6. 323, 1815 ; H. Davy, Phil. Trans., 104. 87, 1814 ; J. J. Berzelius, Pogg. Ann., 14. 558, 1828 ; 1.Thomson, An Attempt to establish ' the First P r i ~ i p l e s f Chemistry by Experiments, London, 1. 189, 1825 ; J. C. G . de Marignac, o &%l. Univ. G'en&ae,46. 363, 1843. 17 J. S. Sfas, MCm. Acad. Belgique, 35. 3, 1866 ;A. Ladenburg, Ber., 35. 2275, 1902 ; A. S cott, Pwc. Chem. Soc., 18. 112, 1902 ; P. K othner and E. Aeuer, Liebig's Aqen., 337. 123, 1904 ; Ber,, 37. 2536, 1904 ; G. P. B axter, Journ. Amer. Chem. Soc., 26. 1577, 1904 ; 27. 876, 1905 ; G . P. Baxter and G. 8. Tilley, i b., 31.. 201, 1909 ; 32. 1591, 1910 ; G. Gallo, Atti A m d . Lincei, ( 5), 15. 24, 1906; Qazz. Chim. ltal., 36. 116, 1906; M. Guichard, Compt. Rend., 159. 185, 1914; Ann. Chim. Pkys., (9), 6. 1916 ; ( 9), 7. 5, 1917 ; E . Kohlweiler, Zeit. phys. Chem., 95. 95, 1920. Is E. P atem6 and R. Nasini, Ber., 21. 2 153, 1588. ' 9 V . Meyer and C. Hartmann, Ber., 26. 1727, 1893 ; 27. 426, 502, 1592, 1894 ; P A sbnasy . and V. Meyer, i b., 26. 1354, 1893 ;A. P, Mathews, Journ. Phys. Chem., 17. 252, 1913; R. Meldola, P hil. Mag.,! 5 ) , 26. 403, 1888. r 10 INORGANIC AND THEORETICAL CHEMISTRY 5 15. The Colour of Solutions of Ioding The colours of iodine soh. may be roughly classed in two groups-violet and brown-but there are intermediate tints ranging from violet-red to reddish-brown. Thus, iodine gives a violet soln. with chloroform, and by the progressive addition of akohol, the colour changes : Violet. Reddish-violet. Red. Brown. Alcohol . e . 0.1 0.2 0.4 0-6 per cent. According to A.Lachmann,lsoln. with pure solvents are always either violet or brown, the intermediate tints are produced by impurities in the solvents. He adds that violet soh. are furnished by hydrocarbons, halogen compounds (not iodides), nitrocompounds, and carbon disulphide, while brown soln. are furnished by iodides, alcohols, ethers, ketones, acids, esters, nitriles, nitrilo-bases, and various sulphur compounds. There is a general tendency for the brown soln, to become violet when heated, and the violet soln. brown when sufficientlv cooled.2 For example, the violet soh. in oil becomes brown a t -90' ; and'the brown soln. in the fatty esters becomes violet a t 80°, while the soln. in ethvl ether remains brown up to the critical temp. of the solvent. H. Rigollot tried to establish a connection between the mol. w t . of the solvent and the colour of the soln. He found that for homologous compounds, and for derivatives of the same radicle, the absorption band is very slightly displaced towards the violet end of the spectrum. It is curious that what P. Walden called the multivalent solvents are those which produce brown soln., and, according to W. Vaubel, a similar remark applies to the solvents containing oxygen, sulphur, and nitrogen--e,g. ether, organic acids, esters, alcohols, aldehydes, and ketones. Dimethylpyrone produces brown soln., and it, according t o A . W. Stewart and R. Wright, is an olronium compound with quadrivalent oxygen. Similarly, the members of A. Lachmann's list of solvents which produce brown soln. have the characteristics of unsaturated compounds in t hat they form associated molecules, etc. The absorption spectrum of the violet soln, is not very different from that of iodine vapour, although one is a band and the other a line spectrum. In his H andbuch der Spectroscopic (Leipzig, 1905), H. Kayser thus summarizes the observations on the absorption spectra of red and brown soh. of iodine : T he red and brown soln. are quite transparent to the longer waves of light, and opaque to the shorter waves ; t he transition occurs i n t he visible spectrum for wave-lengths which are not very different. The violet soln. have a broad absorption band al; about 600pp, in consequence of which they have a stronger absorption in the red portion of the spectrum bhan red soln. ; t hey begin in the blue, and a ain become quite transparent i n the ultraviolet, where there is another absorption b a n f whose position is not known very exactly, but which seems to be about 34OpP, and is probably extended to the limits of the spectrum. The absorption spectrum of violet soln. is bnt little influenced by the nature of the d v e n t , by the temp. or by the concentration of the soh. With brown soln. of the same concentration, the absorption in the violet end of the visible spectrum and in the ultraviolet is much more marked. H. Gautier and G. Charpy, E . Wiedemann, and H. E bert explain the peculiarities in the optical properties of iodine soh. by assuming a polymerization of the solute iodine which in the violet soln. contain Iz-molecules, and in the brown soln. I,+,-molecules. Under any particular set of conditions, there is a state of equilibrium 1,+z+d2, which determines the tint of the soln. From measurements of the raising of the vap. press. of iodine in solvents which produce brown and violet soln., M. Loeb assumed that the iodine in the brown soln. is present as I4molecules and in the violet soh. as I z molecules. He explained the change from brown to violet with a rise of temp. by assuming that the equilibrium 1,+21, is displaced in favour of the I 2 molecules, and conversely with a lowering of the temp. E. Beckmann, however, has shown that the lowering of the f.p. of the violet and THE HALOGENS 111 brown s o h do not support M. Loeb's hypothesis.3 For example, the mol. wt. of iodine a t the f.p. of the following s o h . is near t o that theoretically required for I?, namely 253.84. . . Acetic acid Urethane E thyl acetate Methylal . Brown s o h . . . . . Violet or reddish-violet soin. 2 56 2 56 2 45 253 Carbon tetrachloride Chloroform . Benzene Ethylene dibromide . . . . . 252 267 252 2 44 Similarly with the raising of the b.p. in violet or reddish-violet s o h of iodine in benzophenone, carbon disulphide, ethyl chloride, chloroform, carbon tetrachloride, ethylene chloride or benzene ; or in brown s o h . of ethyl alcohol, methyl alcohol, th~rnol, ethyl ether, m e t h ~ l a l or acetone. The values for the last three solvents , were rather low, presumably because of the chemical action of solute on solvent. High values with benzene are attributed to the formation of a solid s o h . of solvent and solid. Confirmatory results were found by J. H ertz with naphthalene, and by E , Beckmann and P. Wantig with pyridine. The results by I. von Ostromisslensky (0-nitrotoluene), by G. Kriiss and E . Thiele (glacial acetic acid), and by 8. Gautier and G. Charpy indicate polymerization, but they are not considered to be reliable. I n view of the fact that the mol. wt. of. iodine in both brown and violet soln. is the same, E. Beckmann assumes that there is a partial combination of iodine with the solvent, S, forming a compound, say SIz. U nder any particular set of conditions there is a balanced reaction : S12+Sf12. With violet s o h the amount' of combination is very much less than with brown s o h and the spectra approximate to t.hat of iodine vapour. When the iodine s o h . are heated the absorption band is displaced towards the red end of the spectrum approximating to that of iodine vapour ; t he converse obtains when the soh. are cooled. The displacement of the abeorption bands with temp. is explained by assuming that the additive compound, S12,has a maximum absorption in the ultravioIet, and this is more or less affected by the absorption band due to free iodine. The spectra of some brown soh. become permanently altered on heating, showing that there is probably a considerable association between solvent and solute. Since the equilibrium is displaced in favour of an increase i n t he dissociatio~iof the complex, S12, with a rise of temp., the heat of formation of the complex is probably positive. This agrees with the observed effect of temp. on the heat of s o h . and solubility of iodine in different solvents. The negative heat of soh. for all solvents except pyridine is less for brown than for violet soln. P. Wantig has isolated a compound of pyridine, Py, with iodine of the formula, Py12. D . Macintosh (1910) found that by cooling alcohol or acetone s o h of iodine to -80" or -90°, iodine itself and not an addition product separates out, and he argues that the negative heat of soln. of iodine in most solvents rather lends itself to the assumption that the addition compound should dissociate with a failing temp. .The observed negative heat of s o h , however, is a difference between the exothermal heat of formation of the complex, S12, and the endothermal heat of soln. I n support of the theory that in brown soln. a complex of solute and solvent is formed, F. Dolezalek ,k having shown that the partial press. of each form of a substance in a soh. is proportional to the molecular proportion of i t present in the mixture, P. MTantigfound that boiling poln. of iodine in ether, carbon disulphide, carbon tetrachloride, chloroform, and benzene agree with the assumption that even at the b.p. there is a considerable amount of association bekween iodine and the solvents which form brown soln. with this hypothesis also before t,hem, J . H. Hildebrand and B. L. Glascock measured thc depression of the f.p. of certa8inneutral solvents-bromoform and ethylene dibrornide-produce by iodine and certain liquids separately and together. With mixtures which produce violet soln. the total depression of the mixture in the constituents are considered separately or together ; with mixtures which produce brown s o h . the total depression with the mixture is less than the sum of the separate depressions. This is taken as a proof INORGANIC AND THEORETICAL CHEMISTRY that in brown soln. the solvent and solute are more or less combined. H. M. Dawson compared the extrapolated value for the mol. vol. of iodine liquid a t 18" w ith the mol. vol. in nitrobenzene (violet) s o h . with its value in ethyl alcohol (brown) soln. Other solvents were used. The mol. vol. increases in the violet soln. and decreases in brown s o h . This points to the formation of an addition compound of solvent and solute. U. Pomilio found that for the same concentration of iodine with red (benzene) and brown (alcohol, or ether) s o h . the increase in the viscosity is greater than with violet soln., pointing to the formation of larger molecules in the brown soln. A. Hantzsch and A. Vagt, and M. L andau found that the partition coeff. of iodine between two solvents which form violet or brown s o h . remains constant, but with two solven's which form violet and brown s o h . respectively, the partition coeff. of iodine between the violet s o h . is raised with decreasing concentration. This cannot be caused by a polymerization of the iodine in the violet s o h , and it is assumed that it is due to an impoverishment of the solvent in the brown s o h . ; a rise of temp. diminishes the effect owing to the assumed dissociation of the addition compound. H . Gautier and G. Charpy found that the chemical behaviour of brown soln. towards lead amalgam is different from violet ones in that brown s o h . form yellow lead iodide before any green mercurous iodide is formed, whiIe violet s o h . form green mercurous iodide. It is said that brown soln. first form mercuric iodide which passes into s o h . ; t his is reduced by the lead forming lead and mercurous iodides, the latter is reconverted into mercuric iodide, and so on until the lead is a11 converted. into the yellow iodide. If no lead is present the mercuric iodide is reduced to the green mercurous salt by the excess of mercury. With violet soln. green mercurous iodide is supposed to be formed a t once. Hence. " violet s o h . of iodine contain the element in a more simple molecular condition with a tendency to form mercurous iodide a t once." According to L. Carcano, violet soln. do not attack the skin so quickly as the brown ones. According t o A. Beer's law 6 t he absorption of light i n passing through a l ayer of fluid of unit thickness increases i n geometrical series when the concentration increases i n arithmetical series ; a n d according t o K . V ierordt's rule, t h e coeff. of absorption, a,i s proportional t b the concentration C of the s o h . These rules are applicabIe provided t h e molecular state of t h e s o h . is not altered b y a c hange i n the concentration-say b y ionization, or degree of association of soln. or solvent. P. W iintig found t h a t neither rule is applicable t o soln. of iodine, a n d i t is assumed t h a t the solute is partly ionized. According to J. Amann,' brown soln. of iodine contain ultra-microscopic particles, while the violet s o h . contain none. The violet-red soln. in benzene, toluene, and xylene are photosensitive in that clouds of ultramicroscopic particles are formed in white light, and the soln. rapidly turn brown. The s o h . return to their original state in darkness. I n some cases, iodine soln, are true soln. either of free iodine or of a n addition compound ; a nd in other cases they contain free colloidal iodine or of a polymerized addition compound. I n some cases ultra-microscopic suspended particles of iodine are present. REBERENCIES. A. L achmann, Journ. Amer. Chem. Soc., 25. 50, 1903 ; H. Gautier and G. Charpy, Compt.Rend., 110. 189, 1890 ; W. Vaubel, Journ. prakt. C h m . , (2), 6 . 381, 1901. 3 E. Wiedemann, Wied. Ann., 4 . 299, 1890 ; A. Lachmann, Journ. Amer. Chem. Soc., 25.50, 1 1903 ; P. W lntig, Zeit. phys. Chem., 68. 51 3, 1909 ; R. Rigollot, Compt. Rend., 112. 38, 1891 ; W. Vaubel, Journ. prakt. Chem., (2), 631. 381,1901 ; A. Hantzsch and 0. Benstorff, Liebig's Ann., a 9 . 1, 190G ; A. W. Stewart and R. Wright, Ber., 44. 2819, 1911 ; J. Amann, Kolloidchem. Beihefte, 3. 337, 1912 ; H. E bert, Siizber. Phys. N e d . Soz. Erlangen, 8. 3, 1899. M. Loeb, Zeit. phys. Chem., 2. GOA, 1 888; E. Beckmann and A. Stock, ib., 17. 107, 1895 ; E. Beckmann, ib., 58. 543, 1907 ; J. Hertz, ib., 6 358, 1890 ; I. von Oatromidensky, ib., 57. . 341, 1906 ; E. Beckmann and P. W lntig, Zeit. anmg. Chem., 67. 1'7, 1910 ; G. Kriiss and E. Thiele, ib., 7. 52, 1894 ; G. Oddo and E. Serra, Qnzz. Chim. Ital., 29. ii, 343, 1899 ; J. Sakurai, Journ. Chem. Soc., 6 989, 1892 ; H. Gantier and G. Charpy, Compt. Rend., 110. 189, I 890. 1. * F. Dolezalek, Zeit. phys. Chem., 64. 7.2'7, 1908; A . Hantzsch and A. Vagt, ib., 38. 705, THE HALOGENS 1 13 1901 ; D. Macintosh, J ourn. A mer. Chem. S oc., 32. 1330, 1910 ; J. H. Hilde brand a nd B. L. Glascock, i b., 31. 26, 1909 ; H . B f. Dawson, Journ. Chem. Soc., 97. 1041, 1896, 1 910; U. Pomilio, Chem. Zlg., 36. 437, 1912. H. Gautier a nd G. C harpy, Cornpt, Rend., I l 645, 1800 ; E. Beckmann and A. Stock, l. Zeit. phys. Chem., 17. 107, 1895 ; L. Carcano, Bull. Chim. Farm., 47. 5, 1908. A. Beer, Pogg. Ann., 86, 78, 1852 ; K. Vierordt, D e A nwendung des Spektralappclrates z u ~ i Photometric der A bsorptionsspektren, Tiibingen, 1873. ' J. Amann, Zeit. Koll., 6. 235, 1910 ; 7. 67. 1010 ; KoU. Beihefte, 3. 337, 1912, § 16. Binary Compounds of the Halogens with One Another Fluorine and chlorine form neither compounds nor mixed crystals, a1though there may be a eutectic. H. Moissan could detect no signs of a reaction when fluorine gas was led into an atm. of chlorine gas. The two liquids are miscible. Several fluochlorides arc known. Fluorine, however, does unite vigorously with bromine and iodine, forming BrFs and IPS respectively. Chlorine and bromine form mixed crystals but no compound ; bromine and iodine form mixed crystals and a comand IC13-in pound IBr ; while chlorine and iodine form two compounds-ICl which the iodine appears t o function as a metal. Iodine thus. combines with all the halogens ; bromine does not combine with chlorine, but i t combines with both iodine and fluorine ; chlorine combines with iodine alone ; a nd fluorine with both bromine and iodine, but not with chlorine. When the four halogens are taken in pairs in the order of their at. wt., the chemical affinity between fluorine and chlorine is least and that between bromine and iodine is greatest. The known binary inter-halogen compounds are : Fluorine. Chlorine. Bromine. N o action Nixed crystals Chlorine BrP, Mixed crystals Bromine IBr Iodine IF5 I CI ; I Cl, . . . ... Fluorine and bromine.-While no sensible reaction between fluorine and chlorine has been observed, H. Moissan 1 f ound that fluorine unites violently with cold bromine vapour, and the reaction is attended by unecflamm Bclairante, but with the evolution of little heat. P. Lebeau found that no flame is produced if dry liquid bromine is employed, and he showed that the product of the reaction is bromine trifiuoride, BPS, a result almost simultaneously established by E . 33. R. P rideaux. No reaction --solvent or chemical-occurs between liquid fluorine and solid bromine, and the fluorine can be distilled from the latter without any sign of interaction. There is no indication of the formation of a lower bromide, say, BrF ; a nd attempts to prepare a higher fluoride, say BrF4, 'by passing a large excess of fluorine over the trifluoride, were fruitless. Bromine trifluoride was also made by the action of fluorine on potassium bromide : K Br+2F2=KF+BrF,. Bromine trifluoride is a colourless liquid with a smaller ~ p gr. than bromine ; . i t freezes to a crystalline mass melting a t -2" (E. B. R , P rideaux), or a t 4" t o 5" (P. Lebeau), and boils between 130" and 140"-the exact temp. was not determined because the vapour attacked the thermometer. The liquid fumes strongly in air, and acquires an orange-yellow colour. The vapour is very irritating and corrosive. The reactivity of the trifluoride recalls that of fluorine itself. It r eacts violently with water, giving off oxygen, and forming a mixture of hypobromous, hydrofluoric, and bromic acids ; a nd an analogous reaction occurs with soh. of the alkali carbonates. These reactions are taken as evidence that bromine is the electropositive and fluorine the electronegative component in the compound. Even a t -10" the solid trifluoride reacts incandescently with iodine to form iodine pentafluoride, IF5, similarly with bromine ; with sulphur, to form sulphur bromide ; a nd with red phosphorus, arsenic, antimony, boron, and silicon. It r eacts with warm VOL. 11. I INORGANIC ANT) THEORETICAL CHEMISTRY ca,rbon ; a nd attacks most metals and their compounds. It behaves like ffuorinv towards organic compounds. I n 1844, H. B. Leesen claimed to have made a bromine fluoride by leading fluorine into bromine, but fluorine was n ot unequivocally isolated until 1886, a nd t,herefore this claim does not hold good ; some other products must have been formed. The same remark applies t o M. A ubde, M. Millet, and M. Leborgne's claim to the use of bromine chloride in photographic work. Fluorine a nd iodine.--According to H. Moissan,3 when a current of fluorine is directed on to a fragment of dry iodine, the iodine burns with a pale flame ; t here is a large evolution of heat. Iodine liquefies almost immediately the fluorine begins its attack, forming a homogeneous dark liquid which on further action forms an upper colourless and a lower dark layer. Under suitable conditions, the product of the action can be condensed t o a dense colourless, fuming liquid which resembles the iodine pentdluoride, first prepared by G. Gore in 1875 by the action of iodine on silver fluoride in a platinum vessel : 5AgF+312-+IF5+5AgI. The same compound is formed by the action of fluorine on hydrogen iodide : H I-\-3F2+IF5+HF, Liquid ff uorine exerts no chemical or solvent action on iodine, and the liquid can be distilled f from the solid without any sensible change. I a t ube of liquid fluorine and iodine be sealed off, and removed from the cooling agent, the layer of fluorine near the iodine acquires a dark colour, the iodine liquefies, an energetic action sets in, and white fumes are projected some way up the tube, and finally a pale green flame appears for a few seconds. An analysis of the liquid by E. B. R. P rideaux corresponded with the formula IF5. Fruitless attempts were made to prepare a higher fluoride. The sp. gr. of the liquid pentafluoride approximates to 3.5 ; i t boils a t 97" without decomposition, but it dissociates between 400" and 500" ; t he liquid freezes a t -8' t o a white solid with a smell resembling that of camphor ; t he fumes from the liquid affect the respiratory organs. The liquid dissolves iodine and bromine forming brown-coloured soh. Iodine pentafluoride is very reactive although it is the most stable of the halogen fluorides. It reacts with water with b bruissement of red-hot iron forming iodic and hydrofluoric acids: 2 1 P 5 + 5 H 2 0 = I 2 O 5 + l O ~ . An analogous reaction occurs with alkaline s o h ! when the alkali iodate and fluoride: are formed. The liquid can be distilled in hydrogen without chemical change ; oxygen has no action a t 100" ; chlorine reacts in the cold, forming iodine chloride ; bromine produces bromine trifluoride, and iodine bromide, IBr ; phosphorus reacts vigorously, forming phosphorus pentafluoride, PF5, and iodine ; arsenic and antimony act like phosphorus ; with sulphur it forms sulphur hexafluoride, SFe,'and free iodine ; carbon reacts energetically forming carbon tetrafluoride, CF4, and iodine ; crystalline silicon is not attacked in the cold, but on warming, the reaction resembles t,hat with carbon ; boron acts similarly. The alkali metals a t f i s t react energetically with the liquid, but the reaction soon comes to a standstill owing to the formation of a p rotective layer of iodide and fluoride ; t he molten metals react with explosive violence. Silver, mercury, iron, and magnesium are not attacked a t 100", platinum is a ttacked a t a red heat. Mica is attacked slowly in the cold, rapidly when heated ; dry glass is attacked with but little vigour ; a nd the silicon alloys-e.g. ferrosilicon -are rapidly attacked when warmed. Potassium hydride, KH, reacts energetically producing a violet vapour of iodine and hydrogen iodide, and solid potassium iodide and ff uoride. Calcium carbide does not react a t ordinary temp., but it does so when warmed. Calcium carbonate and phosphate do not react with the pentafluoride. The liquid sinks in conc. sulphuric acid and is slowly decomposed ; i t mixes in all proportions with nitric acid and is likewise slowly decomposed-bydrofluoric acid is formed in both cases. It reacts vigorously witb hydrochloric acid. Carbon disulphide dissolves the liquid forming a dark violet soh. ; t urpentine is energeti'callp decomposed ; benzene a t firsf; dissolves the liquid, but a vigorous reaction soon starts. Chl~rine mdbromine.-In 1826 A. J . Balard 4 believed that he had forped a T HE HALOGENS St5 compound of these two elements as a strongly refracting yellowish-red liquid, by passing chlorine into bromine, and condensing the vapour in a freezing mixture. C. Lowig and C. P . Schonbein (1.863) confinled these observations, and the former also claimed to have made a chlorobromine hydrate, BrCl.5H20. W. Bornemann (1877) further investigated the subject, and said that the chlorobromine is fo.rmed only a t temp. below 0°, and that the alleged hydrate is probably a mixture of t he hydrates of chlorine and bromine. According to J. Krutwig, chlorine acts on silver bromate, AgBr03, a t 50" forming silver chloride, oxygen, and chlorine monobromide. P . Donny and J. Mareska (1845) believed that the two elements united chemically a t -90" ; a nd V. Thomas and P. D ~ ~ p o (1906) not only claimed that bromim is monochloride, BrCl, is formed by the action of liquid chlorine on bromine, but they state that bromine trichbride, BrC18, is formed as a red solid by the action of Iiqnid chlorine on bromine monochloride. L. W . Andrews and H. A. CarIton (1907) determined the sp. gr. of mixtures of liquid chlorine and bromine, and concluded that because there is a small contraction "chemical combination probably occurs to a limited extent between these two elements." I n 1882, M. Berthelot expressed doubts about the alleged formation of compounds of these two halogens because of the small heat developed during their supposed union : Brli,.+Cla,,=BrCliiq.+0.7 cal. The work of P. Lebeau (1906) and B. J. K arsten (1907) has shown that M. Berthelot's suspicion was well-founded. Atomic p e r ce& FIG. 23.-J3oiling Point Curves of Mixtures of Chlorine and Bromine. A t ~ m i c e r cent' p F I ~ 24.-Roiling . Point Curves of M i s tures of Chlorine and Iodine. The reports of the formation of compounds of chlorine and bromine, and the description of the properties 0.f thede compounds are to be blue-pencilled ; in all cases the products under investigation were no doubt mixed crystals of the two elements. The f.p. and the b.p. curves show no signs of the formation of a chemical compound, but the evidence points to the formation of a continuous series of mixed crystals of the two elements. The liquidus and solidus curves, as in H. W. B. Roozeboom's Type1 (Fig. 6-1. 10,2), fall regularly from the m.p. of bromine to that of chlorine, and from about 20 to 90 per cent. of chlorine the two curves are from 10" to 14"a part, being respectively convex and concave to the horizontal axis ; b ut there is no sign of a maximum. The results show that the composition of the crystals is dependent on the temp. The fact that analyses of the supposed compound furnished numbers in agreement with BrCl was a consequence of the solubility of chlorine in bromine. Fig. 23 shows the composition of the vapour and that of the liquid with which it is in equilibrium a t the b.p. with mixtures containing varying proportions of the two elements. There is no sign of an approximation of the two curves a t intermediate points such as would occur if chemical combination occurred, and such as actually occurs with the corresponding curves for chlorine and iodine, Pig. 24, where the curves approach, and almost touch one another a t a point where the two components are present i n equi-atomic proportions, IC1. I n the latter case, therefore, it is inferred that the coinpound I Cl exists in t he vapour phase, and is only slightly 1 16 I NORGANIC AND THEORETICAL CHEBfISTRY dissociated ab 100°, t he b.p. of the particular nlixture in question. While the evidence frolri physical chemistry thus gives no support t o the existence of a bromine chloride, hI. D el6phe and L. Ville believe that, such a compound does exist because a s o h . of chlorine in bromine reacts with ethylene forming chlorobromoethane. Chlorine and iodine.-In t h e course of his historic research on iodine, J. L . G ay Lussac (1814) prepared a c ompound of iodine and chlorine by the action of chlorine gas on iodine-the gas was absorbed by the solid forming a reddish-bromn liquid which is s o r emarkably like bromine, that before that element had been recognized as a distinct chemical individual by A. J. B alard, J. v on Liebig mistook bromine for iodine chloride. If t he chlorine be in excess, citron-yellow needle-like crystals are formed. The liquid product is iodine monochloride ; t he crystalline solid is iodine trichloride. H. D avy called the product formed by the action of iodine on ode chlorine, chlorionic acid, a nd he regarded it a s a c ompound consisting of proportion of iodine and one of chlorine " 4 . e . iodine monochloride. W. S tortenbeker's 7 i nvestigation on the f.p. curve, Pig. 25, left no doubt as t o thc existence of these two iodine chlorides. The f.p. curve shows two maxima, one at 27.2" corresponding with the so-called iodine a-monochloride, and the other a t 101" corresponding with iodine trichloride ; t he corresponding eutectics are at 7 -9" and 22.7". I n 1854, j ~ . r showed ~t h a t ~there are a two modifications of the monochloride which W. S tortenbeker respectively distinguished as the a- a nd 13-forms. The a-modification is f ormed if t he fused mass is rapidly cooled when it f orms ruby-red, transparent, needle-like crystals belonging t o the cubic system. The sp. gr. is 3.18223 (0") and 3-12988 (18") according t o T. E . T horpe. The iodine 13-monochloride is f ormed when the undercooled liquid is slowly cooled by a freezing mixture t o -10". The crystals are brownish-red, six-sided rhombic plates resembling solid iodine ; t hey melt a t 13'9" or more exactly 13.92". The a-form is t he stable variety. Both forms when melted furnish the same liquid ; a nd the m.p. of e ach variety is FIG. 25.-Freezing Point Curve lowered if e ither iodine or chlorine be in excess as of I odine and Chlorine. indicated in the diagram. The 13-crystals are more stable in the presence of an excess of iodine, and less stable in the presence of an excess of chlorine. The existence of iodine monochloride in freezing s o h . has also been established by the experimental results graphed in Pig. 25. (6 I n 1861, H. Kiimmerer described what h e considered to be iodine tetrachloride, ICl,, but J . B. H annay could not confirm the report. I odine pentachloride, ICI,, analogous w ith iodine pentduoride, IF,, h as not been isolated. This compound may be formed by the action of liquid chlorine on iocline; i f s o,it is so unstable at o rdinary press. that i t has n ot been recognized. T h e preparation of i odine monochloride.-Iodine monochloride can be ma& by pasking dry chlorine over dry iodine confined in a retort until the solid has completely liquefied ; or, as recommended by W. Bornemann,Q until crystals of iodine- trichloride appear. The liquid is then distilled; a reddish-brown fluid is should be rectified by distillation obtained which, according t o P. S ~hiitzenber~er, a couple of times between 100" and 102". The a n a l p i s approximates t o iodine inonochloride, ICI. A. B unsen (1852) obtained the same compound by boiling iodine with an excess of aqua regia, diluting the liquid with water, extracting witb ether, and evaporating off the ether. T o prepare iodine monochloride, mix 100 g rms. of finely powdered iodine w ith 300 c.c. of hydrochloric acid (sp. g r. 1-15)i n a porcelain basin ; a nd add 28 C.C. of nitric acid (sp. gr. THE HALOGENS 1.41) t o provide sufFicient chlorine t o convert d l the iodine into its chloride. The mixture is continuously stirred on a water bath a t about 40°, t he colour changes from brown to paleyellow when the reaction begins ; t he iodine then dissolves and the s o h . becomes orangecoloured-no chlorine escapes under these conditions. The water in the bath is then boiled to expel the nitrosy1 chloride. With the excesB of hydrochloric acid used, the soln. of iodine chloride is stable and does not decompose on boiling. G. 8. Skrullas recommended passing chlorine through tincture of iodine, but there is, then a complication due t o t he action of the gas on the alcohol. Iodine monochloride is formed by the action of h ypochlorousacid or of sulphuryl chloride, SOzClz, on iodine (0. Ruff), or, according to A. Tohl, on iodobenzene, p-iodotoluene, and analogous aromatic iodides-a trace of water may be required t o s tart the reaction. These reactions are not recommended as processes for preparing the compound. According to 0. R uff, the sulphuryl chloride does not act on iodine alone, but in the presence of aluminium chloride or iodide, iodine monochloride is formed i t he f ; sulphuryl chloride be not in excess : S 02Clz+I~==21C1+S0z if the sulphuryl chloride be in excess, some iodine trichloride is also formed, According to R. Bunsen, nitrogen iodide dissolves in hydrochloric acid without developing a gas, and the liquid contains iodine monochloride. I n general, A. Wkrabel and P. B uchta 10 h ave s h o r n that iodine monochloride is alone produced when iodine or an iodide is treated with a strong oxidizing agentiodic acid, chloric acid, chlorine, hypochlorites, permanpnates, etc.-in the presence of conc. hydrochloric acid. It is supposed that hypoiodous acid, I.OH, is first formed, and subsequently transformed to iodine monochloride by the acid. Thus, C. Roberts recommended reducing iodic acid by $otassium iodide in the presence of conc. hydrochloric acid. Hydrogen chloride also reduces iodine ,pentoxide with the formation of iodine monochloride. J. J. Berzelius' process is one of the most convenient modes of preparing iodine monochloride, namely, by heating an intimate mixture of iodine and potassium chlorate; P. S ~hiit~zenberger recommended four parts by weight of chlorate to one of iodine. The reaction is symbolized : 2KC103+Iz =2KI03+C12. The chlorine so formed unites with the excess of iodine to form the monochloride. According to T. E. Thorpe and G. H . Perry, potassium chloride, potassium perchlorate, and iodine pentoxide are formed only when the mixture is heated to too high a temp. T . E. Thorpe recommends purifying t!he product by distilling it from powdered pot,assium chlorate. The properties 01 iodine monochloride.-As just indicated the hyacinth-red or ruby-red solid e x i ~ t in two forms ; a s indicated above, W . Stortenbeker 11 f ound that s the a-variety has a m.p. of 27.2" or 27.165" (G. Oddo), and the 13-variety, 13.9" or 13.92" (0. Oddo). The older determinations by J. T rapp, J. B. Hannay, and P. Schutzenberger are inaccurate on account of the use of impure specimens. The b.p. of the molten iodine m.onochloride has not been yet determined very precisely. Determinations by T. E. Thorpe, G. Oddo and E . Serra, and J. B. H annay varied from 100" to 102" ; W. Stortenbeker interpolates from his vap. press. measurements 94.7", and B. J . K arsten finds 97.4". Possibly the higher boiling product contains some iodine in soln. owing t o t he decomposition : 31CI+12+1C13, observed by R. Kane and W. Bornemann, and the sublimation of the trichloride,which leaves the molten liquid richer in iodine. The dissociation of the molten monot;hloride can be but small since W. Wtortenbeker found a t 30°, 51.1, and a t SO0,52.1 atomic per cent. of chlorine; even a t the b.p., the dissociation is but small, as is evident from B. J. Karsten's diagram, Pig. 25. As the coniposition of the liquid phase deviates from the C1: 1=1 : 1, so does chlorine tend to accumulate in the vapour phase, and only when thc composition IC1 is approached do both phases have the same composition. The lower end of the boiling curve corresponds with the separation of IC13 from the. molten liquid. The vap. press. of iodine monochloride at its m.p. 27.2" is 39 mm. The brownish-red colour of the vapour indicates that iodine monochloride is stable ; t he absorption syectrnm of a 30 cam.layer of vapour a t 40°, and 30 mm. press., was found by D . Gernez to be different from tbat of ' INORGANIC AND THEORETICAL CHEMISTRY chlorine and that of bromine, chiefly in virtue of its possessing fine lines in the green. H E. Roscoe and T. E. Thorpe drew attention to the resemblance of the ppectrum o f iodine monochloride to that of bromine. W. Stortenbeker found the heat of the transformation to be IClp+IC1,+203 cals., or, according to S. Tanatar, 273 cals. The heat of fusion of the a-form is -16.42 cals. per gram or -2658 cals. (-2320 cals. according to S. Tanatar) per mol. ; a nd for the /3-form, -14 cals. per grm., or -2267 cals. per nlol. W . S tortenbeker also found the sp. ht. of the a-form to be 0.083 between -13-5O and 15" ; for the /3 form, 0.085 between -10" and 0" ; a nd 0.158 for the liquid between 153 and 77". K. Strecker found the ratio of the two sp. ht. to be y=la317 ; and C,=0.0512 ; Cv=0.0389. The sp. ht. of both the a- and /3-forms are, nearly alike, and hence the heat of transformation is nearly independent of temp. The h~3.t:of formation, according t o J. Thomsen, Isolid+Clgas'~Clliquid+5.8 Cals., and according to R. Berthelot, for I mlia+C!gas=ICI~oII,lf6.8 cals. No diff erence can be detected in the nature of t'he brownish-red oily Iiquid, iodine monochloride, derived from the a- and /3-solids. S. Tanatar found the sp. gr., and K. Beck the viscosities, to be the same, and the heat of resolidification to the a-form are the same. Iodine monochloride is miscible with iodine or bromine in all proportions. The sp. gr. of the a-solid is 3.18223 a t 0°, 3'1288 a t 17.93'-J. B. Hannay's value3 are a little higher; the sp. gr. of the liquid is : T. E . Thorpe's formula for the specific volume v a t the temp. I3 is v=1+0~0009158960 + 0-00000083296~2+0~0000000027~I33. The vapour density is that theoretically required for ICl ; with oxygen 32, the value for IC1 is 162.38 ; a t 120°, the vaponr density is 160-6, and a t 512": 156.4. Conclusions as to the degree of dissociation a t different temp. cannot be derived from the vapour density determinations since it proceeds without changing the number of moIecule~-2ICl+I~+C1~. Iodine trichloride vapour is almost completely dissociated into the monochloride and chlorine. K . Beck's value for the viscosity is 7.029 a t 16", 5'069 a t 28.4"-benzene a t 5" unity. Unlike iodine, the monochloride does not blue starch, but it bleaches indigo and litmus soh. The aq. soln. can dissolve much iodine. Iodine monochloride dissolves in carbon tetrachloride and in liquid sulphur dioxide, forming red soln. ; arsenic trichloride forms a reddish-brown s o h The mol. wt." of the a-form of iodine monochloride in phosphoryl chloride determined by the cryoscopic method, corresponds with the formula IC1. The b.p. of carbon tetrachloride is lowered by iodine trichloride owing to the fact that the solute sublimes between 70" and 75". The b.p. of carbon tetrachloride is also lowered by iodine monochloride ; C. Oddo and E. Serra attribute this to dissociation in accord with the equation91C1+12+14+31C13, a nd the effect of the trichloride in lowering the b.p, masks the rise due to the iodine. The depression of the f.p. of glacial acetic acid by iodine monochloride is normal, corresponding with the formula ICl. Iodine monochloride is stable in hydrochloric acid in which it forms a yellow soln., but in water 13 i t is presumably hydrolyzed forming hypoiodous and hydrochloric acids : IC1+H20=HOI+HCI, corresponding with the fact that the iodine is more electropositive than chlorine. The hypoiodous acid immediately breaks up The joint effect of into iodic acid, HIOs, and iodine : 5 HIO=2H20+21,+HI03. these two consecutive reactions is symbolized : 51C1+3H20=212+5HCl+HIOy, a n equation which is virtually that employed by J. L . Gay Lussac in 1814 to represent the formation of iodine and iodic acid when iodine monochloricle is decomposed by water. According to W. Bornemann, the iodine which separates is brought into s o h . by treatment with hydrochloric acid. The iodine monochloride forms a complex, ICl.HC1, in the presence of hydrochloric acid. P. Schiitzenberger found that this compound may be isolated by extracting the soh. with ether. As the acid is diluted, the compound is progressiveIy hydrolyzed. Aq. soln. of t he alkali hydroxides behave in a somewhat similar manner to water furnishing the alkali chloride and iodate, and free iodine.; the latter may react with the excess of alkali lye forming a mixture of iodide and iodate. I t he alkali be not in excess, H. Griinef berg found some alkalichlorate is formed. According to K. J. P. Orton and W. L. Blackman, an alkali hypoiodite can be detected for a short time in the soln. obtained by treating iodine monochloride with a soln. of an alkali carbonate, or lime-or baryta-water. W ith ammonia, A. Mitscherlich and R. Bunsen found that ammonium chloride and nitrogen iodide are formed. The salts of t he oxychlorine acids are decomposed by iodine monochloride, forming iodates and free chlorine ; t hus a t ordinary temp. the hypochlorites are rapidly decomposed : 3KOCI+ICI+KI03+2KC1+C12, a nd the chlorates slowly : X C I O ~ + I C I + K I O ~ + C ~ Consequently, both iodine, and iodine as the mono~. chloride, displace the chlorine from the oxychIorine acids. The reverse is the case with the halide salts. Iodine monochloride behaves towards mercury, aluminium, phosphorus, arsenic, antimony, bismuth, tin, etc., very much like free chlorine forming the metal chloride, some iodide, and free iodine. Aluminium foil is not attacked very much a t first, but it, afterwards burns with a bluish-white flame. Copper foil acts very slowly, but the powdered metal reacts quickly. Potassium explodes in contact with iodine monochloride, but the action with sodium is very slow. Sulphur reacts slowly forming free iodine and sulphur chloride, selenium acts more quickly, and tellurium more quickly still. Hydrogen sulphide and sulphur dioxide are decomposed with the separation of iodine, and iodine reacts with the excess of sulphur dioxide.14 A sat. soln. in carbon disulphide gives off white fumes, and on the addition of water, an oily mixture of carbon disulphide, carbon tetrachloride, sulphur chloride, and c,arbon thiochloride'is precipitated. Metal oxidese .g. lead dioxide, cupric oxide, mercuric oxide, etc.-form metal chlorides and iodides, free iodine, and oxygen. With galena, iodine monochloride forms lead iodide and sulphur chloride. Mercuric chloride precipitates mercuric iodide from conc. soln. of iodine monochloride and iodine trichloride remains in soh. A little stannous chloride, SnC12, precipitates iodine and forms stannic chloride, SnCI4, but with more stannous chloride, stannous iodide, SnT,, is formed. M. P araday 15 found that liquid iodine chloride can be electrolyzed, the iodine collecting a t the cathode, the chlorine a t the anode. According t o t he ionic hypothesis, this means that io&ne chloride is partially ionized, IC1=Ia+CI', that is, iodine monochloride is a compound of the cation, I., with the anion, Cl'. This also corresponds with the chemical behaviour of this compound, and with 0 . Walden's experiments on the electrical conductivity of soln. of iodine monochloride in inorganic solvents-liquid sulphur dioxide, arsenic trichloride, and sidphuryl chloride. L. W. Andrews 16 assumed that iodine monochloride is the chlorine derivative of hypoiodous acid, HOI ; A. Skrabel and P. Buchta consider it best to regard the hydroxyl derivative HO.1 as a base. Iodine monocIdoride forms an addition compound with hydrogen chloride, ICI.HCI, as found by P. Schiitzenberger,l7 in which the stability appears to be greatly increased. P. Schiitzenberger's compound may be regarded as a complex acid, HIC12, with the ions H' a nd IC1,' ; a nd the polyhaloids of H. L . Wells. and S. L. Penfield, and A. Geuther as salts of this acide .g. t he compound KCI.IC1 becomes KIC12 ; N(CH3)4Cl.ICIbecomes N(CH3)4.1CI, ; etc. Phosphorus pentachloride with iodine monochloride forms phosphorus trichloride and the compound PC121. Iodine trichloride.-This compound was discovered by J. L . Gay Lussac as the result of treating warm iodine or iodine monochloride with an excess of chlorine. The trichloride collects as a citron-yellow crystalline sublimate on the cooler parts of the vessel. It is also formed by the action of liquid chlorine on iodine, or an iodide-say lead iodide.18 The iodine trichloride is almost insoluble in liquid chlorine, and hence, say V . Thomas and P. Dupuis, this method of preparation is very convenient. It is also formed by the action of dry chlorine on hydrogen . iodide ( A. Christomanos) ; silver iodate (J. K rutwig) ; or methyl iodide ( L. von ' 120 INORGANIC AND THEORETICAL CHEMISTRY Ilosvay) ; b y the action of hydrogen chloride on iodic acid ((3. S . S6ruIlas) ; of phosphorus pentachloride on iodine pentoxide ( 0 . Brenkn) ; of iodine on sulphuryl chloride in excess (0. Ruff) : 3SOzCI2+I2=2ICl3+3SO2 ; a nd by heating iodine monochloride. T o prepare iodine trichloride, heat 20 grms. of iodine i n a retort, A , P ig. 2 6, which delivers into a glass balloon fi filled with chlorine, and connected with n Kipp's apparatus C delivering chlorine. The chloririe is rapidly absorbed as soon as i t comee in contact with the vapour of iodine, and reddish-yellow crystals of iodine trichloride are formed on the wnlls of t he balloon. The excess of chlorii~eis filially cspelled by a stream of carbon f dioxide. . I t he c otals nre desired, the ,balloon must be broken ; if a soln. of iodine triohloride is desirez the crystills car1 be dissolved in about ten times their weight of water. Iodine trichloride forms long citron-yellow needles, and also large reddish-brown rhombic plates. The sp. gr. a t 15" is 3.117 (A. Christomanos). This compound melts in a sealed tube under the press. of its own vapour at 101' an.d 16 atm. press. forming a reddish-brown liquid which freezes to crystals of the same colour. The citron-yellow crystals are obtained by sublimation. The differences in colour and appearance led W. Stortenbeker to say that the trichloride is dimorp hous. The crystals readily decompose in air, but they can be preserved unchanged in an atm. of chlorine, and A . Christomanos states that the compound is so volatile that it sublimes a t -12" in an atm. of carbon monoxide or dioxide, and a t 0" in an atm. of oxygen. According to P . G. Melikoff, the vapour of iodine trichloride is almost completely dissociated into the monochloride and free chlorine a t about 77" a nd atm. press. The f.p. curve is indicated in Fig. 25, and the b.p. curve in Fig. 24. The heat of formation Isolin+3CIgas=IC13s01i,I+21.49 Cals. ( J, Thornsen), or 3-16-3 Cals. (3%. Bert.helot) ; or IC~solid+2C~g~s=IC~ Bo1id+15.66 Cals. (J.Thomsen), or +9.5 Cals. (M. Berthelot).lQ At ordinary temp. hydrogen has no action on iodine trichloride, but when warmed, it forms the Fro. 26.-Tho Preparation 01 monochloride and hydrogen chloride : H2+IC13 I odine Trichloride. =IC1+2HC1; and a t a still higher temp., hydrogen iodide and chloride are formed. The trichloride behaves towards potassium, phosphorus, and the oxychIorine acids in an analogous manner to the monochloride. According to H Basset and E . Fielding, A chlorine monoxide-gaseous or dissolved in carbon tetrachloride-furnishes iodine pentoxide and free chlorine : 21C13+5C120=1205+8C12 ; i t does not act on iodic acid or on iodine pentoxide. Iodine trichloride dissolves in benzene, carbon tetrachloride, nitrobenzene, liquid sulphur dioxide, etc., and the two last-named soln. conduct electricity.20 Iodine trichloride dissolves in alcohol and ether, but slowly decomposes these liquids ; a nd with carbon disulphide, said J . B. Hannay, it reacts : 4CSz+61C13=2CC~+2CSC12+3S2C12+31,. T he trichIoride is Iess soluble in water than the monochloride, and it forms a dark yellow, strongly acid liquid which fumes slightly in air. Water furnishes similar products to those obtained with the monochloride. According to A. Skrabel and F. Buchta, the first action of water is to hydrolyze the compound : IC13+H20 +HCI+OClZOH. I t is more generally supposed that the first action of water is to hydrolyze the trichloride : IC13+3H20+H3103+3HC1 ; t he iodic acid, HsI03, imnlcdiately breaks down into iodic and hypoiodons acids : 2 H3103=HOI+HI03 -/-2H20, and the hypoiodous acid reacts with the hydrochloric acid formed in the first stage of the reaction : HOI+HCl==ICl+H,O. According to P. Schiitzenberger, the reaction progresses : 21C1,+3H20--~HCI+-ICI-I-HIQ~,nd the hydrolysis is a 121 T HE HALOGENS complete if 10 to 20 mols. of water are used for each mols. of the trichloride ; if less water is used, some trichloride remains in soln. undecomposed. I n consequence of the hydrolysis, the aq. soh. is stable only in the presence of much hydrochloric acid -possibly a complex IC13.HC1 is likewise formed. The action of soh. of alkali hydroxides follows similar lines---chlorides, chlorates, iodides, iodates, and free iodine are formed as with the monochloride. 1 t he alkali be not in excess, the f reaction is represented : 51C13+18KOH+3K103+12+15KCI+9H20. There are some side reactions furnishing hypochlorites and chlorates, and, according to J . Philipp, perchIorates, corresponding with the tendency of the trichloride to lose free chlorine : IC13=ICl+C12. I n alcoholic solution, potassium hydroxide forms with the trichloride, potassium iodide, iodate, chloride, and iodoform. According to A. Christomanos, ammonia in excess forms nitrogen iodide, and ammonium iodide and chloride. Sulphuric acid precipitates from aq. or hydrochloric acid soln. of the trichloride a white precipitate which soon changes to a yellow colour, and which has ;i some analogies with P. Chrhtien's iodosulphate,21 1203.SOs.~H20 t dissolves when the mixture is heated but separates out on cooling ; nitrlc acid precipitat,ea iodine with the evolution of chlorine. Silver foil is transformed by an aq. soln. of the trichloride into silver chloride and iodide ; silver oxide with an excess of the trichloride is transformed into the chloride and iodic acid ; with more silver oxide, silver iodate is formed ; a nd with an excess of the oxide and a boiling s o h . some silver periodate is formed. Mercuric oxide is slowly transformed into mercuric chloride and oxide ; chlorine, oxygen, and possibly chlorine monoxide are evolved. Aq. s o h . of the trichloride give a precipitate of iodine with a little stannous chloride ; with more stannous chloride, some stannous iodide is formed. Consequently, although chloroform extracts no iodine from the aq. s o h , it will do so after the addition of stannous chloride. Sulphur dioxide and ferrous sulphate are oxidized. Iodine trichloride resembles the monochloride in forming a series of addition products; these can be regarded as derivatives of the acid IC13.HCI, that is HIC14. E. Filhol 22 made a series of these salts of the type ICI3R*C1-where R sta'nds for Cs, Rb, K , Na, Li, NH.4, N(CH3),-by leading chlorine into an aq. soh. of the iodide ; or by adding iodine to an aq. soln. of the chloride. The salts crystallize readily, without water of crystallization, excepting that the sodium salt has two and the lithium salt four molecules of water of crystallization. A. Werner likens them to the salts of chloroauric acid : [I,. Pentahalides. K;Cl]lt' Chloroaurates. R . F. Weinland and P. Schlegelmilch also obtained a series of salts of the general formula 21C13.RC12.8H20,mostly in needle-like crystals-when R represents cobalt the crystals are dark orange-red ; nickel, green ; manganese, orange-red ; zinc, golden yellow ; beryllium, yellow ; magnesium, yellow ; a nd calcium or strontium, yellow. Salts with R=Ba, Cd, Cu, Hg, P b were not obtained. The salts are not very stable ; t hey decompose in a desiccator in a few days owing to the volatilization of iodine trichloride and water. They decompose rapidly when heated. I n another series of salt's of t,he type IC13.RCl.aH20, where R r epresenh univalent K, NH4, Rb, Cs, Na, Li, and N(CH3)4,A. Werner says that " since iodine can be extracted from the salts by carbon tetrachloride, these salts must be addition products of RC1 and IC13." The argument is not sound if it be intended to prove that these products are molecular compounds as distinct from compounds of the formula R(IC14), because it couId be argued that they are decomposed by this solvent : R(ICh)+IC13+RC1. Xeveral other polyhalides of this kind have been report,ed.23 P . Jaillard described a compound SCI2.2ICl3; a nd R. Weber. SC14.1C13. 0. R uff 122 INORGANIC THEORETICAL CHEMISTRY AND and G. Fischer obtained golden yellow crystals of SC&.21C13 by leading chlorine into a cold soh. of t he trichloride in sulphur chloride. The compound easily decomposes. E. Beckmann and F. Junker 24 obtained the value 233 for the mol. w t. of iodine trichloride when dissolved in phosgene-the theoretical value for IC13 is 233.4. G. Oddo found the depression of the f.p. of soh. of t he trichloride in phosphoryl oxychloride corresponded with a mol. wt:134*21 to 173'42 ; t he calculated value for the trichloride is ICI3, and hence the trichloride must be dissociated or ionized -G. Oddo thinks ionized IC13+IC12-+CI' ; in water the mol. wt. 42.0 t o 46.2 also represents decomposition. P. Walden found the electrical conductivity measurements of soh. of the trichloride in liquid sulphur dioxide, arsenic trichloride, and sdphuryl chloride corresponded with an ionization which increased with increased dilution. The temp. coeff. of the conductivity is negative ; with sulphuryl chloride soh. a che~nicalchange obscured the results. G. Oddo and E . Serra found the b.p. of carbon tet,rachIoride to be lowered, not raised, with iodine trichloride as solute. This was attributed to the volatilization of the trkhloride at the b.p. of the solvent. The mol. wt, calcqlated from the depression of the f.p. of soh. in glacial acetic acid approximated with increasing dilution to 120, that is half the value required for ICls ; this is attributed to the dissociation IC13+ICI+C12, a nd possibly also ionization. P. Schiitzenberger 25 regarded the trichloride as a molecular compound of the mono- and penta-chldrides ICl.IC15 ; J. Philipp supposed it t o be a molecular compound of t he monochloride and chlorine, IC1.C12 ; G. Oddo, as a salt similar in constitution to (C6H&=I-Cl, and C6115-I=C12 ; and R . Stanley as a compound of univalent iodine wlth tervalent chlorine : Iodine and bromine.-In his historic memoir on brpmine, A. J . BaIard (1826) 26 said : Iodine appears to be capable of forming with bromine two different compounds. By taking iodine with a certain proportion of bromine, a solid is produced which, when heated, furnishes reddish-brown vapours, which condense into small fern-like crystals of the same colour. A f urther addition of bromine transforms these crystals into a liquid compound which has the appearance of hydriodic acid. C. Lowig confirmed the observations of A. J. BaIard, and assumed that the liquid rich in bromine corresponded with IBr3, and the one poor in bromine, with IBr. H. Lagermarck made a crystalline solid by mixing eq. amount of the two elements ; a nd W. Bornemann also made the same compound by adding a slight excess of bromine, and volatilized the excess of bromine by heating the product at 50" in a stream of inert gas. The product when cold formed a hard crystalline mass, the colour of iodine, and whose composition corresponds with iodine monobromide, IBr. J. B. Hannay also made the monobromide by the action of iodine monochloride on sulphur dibromide : 2ICI +S2Br2=2IBr+SzCI2 ; and L. von Ilosvay made it by the action of bromine vapour on alkyl iodides. Iodine monobromide forms dark grey crystals resembling iodine and smelling of bromine. The physiological action of the vapour is similar to that of bromine. By slow sublimation a t 50" i t may be obtained in fern-like crystals 2 cm. long. According to P. C. E . M. Terwogt,27 the sp. gr. of t he solid is 4'4157 (0°), 4.4135 ( lo0), 3'7616 (42"), and 3.7343 (50'). When the sp. vol. or sp. gr. of mixtures o f iodine and bromine a t different temp, are plotted against the composition, the curves are continuous except where there is a change of state from solid to liquid ; a nd the curvature is concave upwards from the concentration axis, showing that there is a contraction which may be due either to the formation of a compound or to somephysical cause. The heat of formation, according t o M . Berthelot, is IaolidfBrliSuld YHE HALOGENS 123 =IBrs91idf 2.47 Cals., and with all the constituents solid, 4-2 -34 Cals. According L K . Strecker, between 99.7" and 214-5", the sp. ht. of the vapour is C,=0.039, o C,=3'029 ; a nd the ratio of the two sp. ht. varies from 1.325 to 1.411-mean 1-33, It is, however, doubtful if these results refer to anything but the dissociated vapour, i.e. to a mixture of the vapours of iodine and bromine. The m.p. of the monobromide is near 42", the lower values which have been reported are due to the presence of ilnpurities; the b.p. is near 116", although it volatihzes a t a Iower temp. than this. The liquid rich in bromine formed by t he action of bromine on the monobromide has been styled iodine tribromide, IBr3, and ioliine pent,abromide, IBr5. C. Lowig nixed the liquid wit,h a little water, and allowed it to stand in the cold; the yellowish- ,wo brown crystals which separated were said to be hydrated iodine pcntabromide. 1 . Bornemann showed that in all probability these crystals are 6$ a mixture of ice and iodine bromide ; he also tried t o prepare a higher bromide than IBr, but failed, the reddish-brown liquid obtained by the zoo action of bromine on the monobromide when cooled furnished crystals of the monobromide aud an excess of bromine. A 20 40 60 80 IOOJ I n 1907, P. C . E. M . Terwogt 28 investigat,ed Br/oo 80Atom~c erUcent29 0 60 + p the system iodine-bromine, and showed that the liquidus and solidus f.p. curves, Pig. 27, of mixtures FIG. Mixture8 Of P oint Curves of 37.-Freezing alld of iodine and bromine nearly coincide when the B romine4 ratio I : B r=1 : 1, while with a11 other mixtures there is an unbroken series of nixed crystals, in the one case, AB, between iodine monobromide and an excess of bromine, and in the other case, BC, between iodine monobromide and an excess of iodine. The coincidence of the liquidus and solidus curves a t the point B shows a high degree of probability that a single individual is present in the system. P. C , E . M. Terwogt atso investigated the b.p. curves and obtained two curves, one representing the percentage (atomic) composition of the liquid, and the other that of the vapour a t the b.p. of the liquid. Some results are shown in Table XII. The general shaps of the f . ~curves is that indicated in F ig. , -90 T ABLE IL-BDXLXNB X POINTS F MIXTURESO F I ODINE O Atomic per cent. of iodine. I A ND BROMINE. I3.p. and barometric press. - . -- _______I__-- 5 8-7" (771.2 m m.) 72'7" (759.3 m m.) 104.3" (756-3 mm. j 126.0' (750'0 m m.) 145.4' (708.8 m m.) 159.4" (757.8 mm.) 187.5" (771.7 mm.) . 27, but the curves do not approach one another so closely as in the case with iodine and chlorine corresponding with the fact that a t the b.p., bromine monochloride is probably dissociated more than the corresponding iodine inonochloride. This is in harmony with W. Bornemann's observation that iodine monobromide cannot be distilled between 90" and 100" without some decomposition-the distillate indeed contains so much bromine in excess tha.t it remains liquid. The accumulation of bromine in the vapour ~ h a s e correspond with the greater volatility of that element. The vap. press. a t 50.2" and 92.8" are as follows : 124 INORGANIC AND THEORETICAL CHEMISTRY TABLE X1II.-VAPOUR PRESSURES IODINE-BROMINE Or MIXTURES. --- Temp. 50'2". Atomic per cent. of iodine. Atomic per cent. of iodine. - Liquid. 25 50 100 Yapour, - Temp. 92'8". Yap. press. mm. Liquid. I I Yap. press. Vapour. 0 8 -23 1 00.00 J . J . van Laar has shown how the form of the vap. press. curve$ of a liquid mixture can furnish an indication, not a precise computation, of the degree of dissociationof any compound which may be formed, on the assumption that the different kind of molecules in the liquid-Iz, Br2, and IBr-possess partial press. each of which is equal to the product of the vap. press. of a given component in the unmixed state and its fractional molecular concentration in the liquid. It is assumed that in the liquid, there is a balanced reactioll 21Br+Iz+Br2, to which the law of mass action applies, where K is the equilibrium constant, and C1, C,, and C respectively denote the concentration of the free iodine, free bromine, and iodine bromide. Prom this, P. C. E. M. Terwogt infers that a t 50-2", K for the liquid is ,$; a nd that for iodine monobromide about 20 per cent. of the liquid and about 80 per cent. of the vapour is dissociated. That the vapour of iodine monobromide is not quite dissociated into its elements is evident from its absorption spectrum, which shows some fine red orange and yellow lines in addition to those which characterize iodine and bromine. I n thin layers, the colour of the vapour is copper red. 0. Ruff 29 could uot prove the formation of a compound by the measurements of the light absorption of soln. of iodine and bromine in carbon tetrachloride. Iodine monobromide has a specific electrical coxductivity of 3'078 x 10-4 reciprocal ohms when liquid a t 40.6", and 6'51 x.10-4 when solid a t 16.3". The iodine collects a t the cathode, and the bromine a t the anode when melted bromine monoiodide is electrolyzed between silver electrodes.30 Iodine mouobromide dissolves in chloroform, ether, and alcohol, forming reddish-brown soh. According to E. SoUy, the ethereal soln. id electrically conducting. but not the soh. in sulphur chloride, S2Clz, or carbon disuIphide. Iodine monobromidc is hydrolyzed b y watcr, the liberated iodine colours the s o h . brown, and some iodine may bc precipitated. Indigo soh. are decolorized ; a nd starch paste is colourcd reddish-brown. According to P. WaIden, the molecular conductivity of s o h . of the nlonobromide in liquid sulphur dioxide, arsenic trichIoride, and sulphuryl cbloridc iilcrcases with dilution ; t hc simultaneous s o h . of broilhe and iodine in these so1vent.s raises the conductivity more than the additive sum of the constituents i n the same solvent ; a nd the couductivity of a mixture I+4Br is greater than that of P+2Br, possibly becausc of the formation of iodine tribronlidc. The conductivity decreases with a rise of temp. and conversely. Iodine monobromide, like the other biiary halogen compounds, has the faculty of forming poIyhalides additively. They are obtained by the s o h . of the halogen halide in a conc. soh. of the alkali halide salts. Thus, H. L . Wells and 8. L . Penfield 3' prepared KBr.IBr, CsCLIBr, CsBr.IBr, RbBr.IBr, etc. Since CsBr21 is more stable than CsBr12. it follows that it is the mutual affinity of the halogens themselves, rather than the ~ola~tility the contained halogen, which determines of the stability. ...
View Full Document

Ask a homework question - tutors are online