IMAGE EVALUATION TEST TARGET (MT-S) V // /. // \/%^^£^_ V^ ^ >&, ^ m. 2a 1.0 I.I 3.2 2.5 S*? 1^ Ui m ■ 40 2.0 1.8 11.25 ill 1.4 i 1.6 Photographic Sciences Corporation 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716) 872-4503 *^--v--i,-.-'- y-e^^' }0p^:'s~a¥:immmm»'^?i^^sm^sm^^^m^S^^^S^M ^»\% ;% 1?/ 1^ i- CIHM/ICMH Microfiche Series. CIHM/ICMH Collection de microfiches. A- Canadian Institute for Historical Microreproductions / Institut Canadian de microreproductions historiques m^m^^ '^^rn'sm'' ^^n.'^^^""*^'^ a rf e**1 - --^f^ Technical and Bibliographic Notes/Notes techniques et bibliographiques The Institute has attempted to obtain the best original copy available for filming. Features of this copy which may be bibliographically unique, which may alter arsy of the images in the reproduction, or which may significantly change the usual method of filming, are checked below. D D D D D D Coloured covers/ Couverture de couleur Covers damaged/ Couverture endommag6e Covers restored and/cr laminated/ Couverture restaur^e et/ou pellicul6e Cover title missing/ Le titre de couverture manque I I Coloured maps/ Cartes g6ographiques en couleur □ Coloured ink (i.e. other than blue or black)/ Encre de couleur (i.e. autre que bleue ou noire) □ Coloured plates and/or illustrations/ Planches et/ou illustrations en couleur Bound with other material/ Relid avec d'autres documents Tight binding may cause shadows or distortion along interior marqin/ La reliure serr^e peut causer de I'ombre ou de la distortion le long de la marge intdrieure Blank leaves added during restoration may appear within the text. Whenever possible, these have been omitted from filming/ II se peut que certaines pages blanches ajout^es lors d'une restauration apparaissent dans le texte, mais, lorsque cela 6tait possible, ces pages n'ont pas 6t6 filmdes. rT/l Additional comments:/ V\\ Commenteiires suppl6mentaires>^ L'Institut a microfilm^ le meilleur exemplaire qu'il lui a 6t6 possible de se procurer. Les details de cet exemplaire qui sont peut-dtre uniques du point de vue bibliographique, qui peuvent modifier une image reproduite, ou qui ijeuvent exiger une modii^ication dans la m^thode normale de filmage sont indiqu6s ci-dessous. I I Coloured pages/ D D Pages de couleur Pages damaged/ Pages endommagdes Pages restored and/oi Pages restaur^es et/ou pellicul6es Pages discoloured, stained or foxe Pages ddcolor^es, tachetdes ou piqu6es Pages detached/ Pages d6tach6es I — I Pages damaged/ I — I Pages restored and/or laminated/ I — I Pages discoloured, stained or foxed/ I I Pages detached/ I I Showthrough/ Transparence Quality of prir Qualit^ indgale de I'impression Includes supplementary materis Comprend du materiel supplementaire I I Quality of print varies/ I I Includes supplementary material/ Only edition available/ Seule Edition disponible Pages wholly or partially obscured by errata slips, tissues, etc., have been refilmed to ensure the best possible image/ Les pages totalement ou partiellement obscurcies par un feuillet d'errata, une pelure, etc., ont §t6 film^es d nouveau de fapon d obtenir la meilleure image possible. This item is filmed at the reduction ratio checked below/ Ce document est film6 au taux de reduction indiqud ci-dessous. 10X 14X 18X 22X 12X 16X 20X 26X 30X H 24X 28X 32X re Idtails }s du nodifier )r une ilmage The copy filmed here has been reproduced thanks to the generosity of: Library of Congress * Photoduplication Service The images appearing here are the best quality possible considering the condition and legibility of the original copy and in keeping with the filming contract specifications. L'exemplaire film6 fut reproduit grdce d la g6n6ro8it6 de: Library of Congress Photoduplication Service Les imaytiS suivantes ont 6t§ reproduites avec le plus grand soin, compte tenu de la condition et de la nettatd de raxemplaire film6, et en conformity avec les conditions du contrat de filmage. Original copies in printed paper covers are filmed beginning with the front cover and ending on the last page with a printed or illustrated impres- sion, or the back cover when appropriate. All other original copies are filmed beginning on the first page with a printed or illustrated impres- sion, and ending on the last page with a printed or illustrated impression. les Les exemplaires originaux dont la couverture en papier est imprim6e sont film^s en commengant par le premier plat et en terminant soit par la dernidre page qui comporte une empreinte d'impression ou d'illustration, soit par le second plat, selon le cas. Tous les autres exemplaires originaux sont film6s en commenpant par la premidre page qui comporte une empreinte d'impression ou d'illustration et en terminant par la dernidre page qui comporte une telle empreinte. The last recorded frame on each microfiche shall contain the symbol -^ (meaning "CON- TINUED "), or the symbol V (meaning "END"), whichever applies. Un des symboles suivants apparaitra sur la dernidre image de cheque microfiche, selon le cas: le symbole — ^> signifie "A SUIVRE", le symbole V signifie "FIN". Maps, plates, charts, etc., may be filmed at different reduction ratios. Those too large to be entirely included in one exposure are filmed beginning in the upper left hand corner, left to right and top to bottom, as many frames as required. The following diagrams illustrate the method: Les cartes, planches, tableaux, etc., peuvent dtre film6s d des taux de reduction diffdrents. Lorsque le document est trop grand pour dtre reproduit en un seul clichd, il est film6 d partir de Tangle sup6rieur gauche, de gauche d droite, et de haut en bas. en prenant le nombre d'images n^cessaire. Les diagrammes suivants illustrent la m^thode. ' errata d to It e pelure. ;:on d n 32X 1 2 3 1 2 3 4 5 6 I in ■ -I III ■ iiaiii ^iiilii lUBIWwm ijiiiii iliiniii L mtmimiimmtimttW'""' "<*'' A CONTRIBUTION TO TIIK STITDY OK DOUBLE SALTS IN WATER SOLUTION. DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE ; JOHNS HOPKINS UNIVERSITY FOR THE iv^f •v-v»?, '? ^- DEGREE OF DOCTOR OF PHILOSOPHY, -BY- KBENEZER MACKAY 7V 1896 KASTON, PA.: CHEMICAL I'VBLISHINO COMPANY. 1S97. '^ - /" t stuB-jn^'A rjai-.^ :.,'>-ir-;,i'-*; ^ ^ V CONTENTS. Introduction Conductiviiy apparatus Solutions Conductivity Measurement Method of Titrating Aluminium Alums Conductivity Results Potassium Sulphate Aluminium Sulphate Chromium Sulphate Potassium Aluminium Alum Sodium Aluminium Alum Ammonium Aluminium Alum Potassium Chrome Alum Ammonium Chrome Alum Green Modification of Ammonium Chrome Alum Ammonium Iron Alum Stability of Alums in Dilute Solution Comparison of Conductivity Results Summary of Conductivity Results Freezing-point Measurements Comparison of Kreezing-Poiiit Results Double Chloride of Zinc and Potassium Preparation Conductivity Conclusion Biographical riiRc. • 5 . lo II II 12 • 13 '3 • '3 • 14 . i6 • 17 • '7 . i8 • '9 • '9 . 20 . 20 . 20 . 21 • 27 . . 28 ■ 30 • 32 • 32 • 33 • 34 ■ 35 "•i~;r~n: ACKNOWLEDGMENT. The author wishes to express his sense of gratitude to Professor Remsen, both for class-room instruction and for 'the helpful suggestion and encouragement received from him ; to Professor Morse and Professor Renouf for the benefit de- rived from their instruction ; and to Doctor H. C. Jones, at whose suggestion this work was undertaken, for valuable ad- vice and constant experimental assistance throughout its progress. He also wishes to express his appreciation of the instruction received from Professor Ames, Professor Frank- lin and Doctor Hulburt. A \ se of gratitude truction and for ceived from him ; r the benefit de- ll. C. Jones, at for valuable ad- throughout its preciation of the Professor Frank- Introduction. If a solution of a mixture of two salts having a common ion be subjected to crystallization, the two salts may crystallize out separately or crystals may form containing both. In the latter case two modes of separation are to be distinguished, according as the crystals formed are of variable or of constant composition. When of variable composition they are mixed crystals, or isomorphous mixtures. When the constituent salts are present in constant molecular proportions, the crystalline compound is a double salt. The constancy of composition of the crystals of a double salt raises the question whether the double salt is present as such in solution or is first formed only at the moment of crystallization. The question of the existence of double salts in solution has attracted the attention of many investigators, and a variety of methods have been employed in the investigations. The chief methods, classified according to the phenomena or properties investigated, may be grouped according as they relate (a) to diffusion; (b) to thermal changes; (c) to volume changes; (d) to solubility; (e) to electrical properties; or (f) to cryoscopic behavior. Graham," in his classic researches on diffusion, studied in- cidentally the diffusion of a few double salts, among them alum and the double sulphate of^magnesium and potash. He found that the constituents of the former were present in the diffused 1: in a proportion different from that in the alum itself, and inferred that the alum in solution was at 1 - st par- •tiall- .omposed. With regard to the latter salt, ht *ound that the amount of the double salt which diffused was equal to the sum of the amounts diffused of its constituents taken separately : and he concluded that this double salt was not decomposed in solution. As Graham had not determined the proportions in which the constituents of the double sulphate of magnesium and potash had diffused, but only the total amount of diffusion, his work was repeated by Marignac,' 1 Phil. Trans., 1850, i. » Aun. cbira. phys., [5], a, 546. (i374). who concludficl from an extensive series of observations that there is no difference between mixtures of 'alts capable of forming double salts and those in which no union can take place. He inferred that dou])le salts are only formed at the i.ioment of crystallization. The work of van der V.'al.' who diffused alums atu' other double sulphates, and of Ingenhoes,' who investigated barium acetonitrate and similar salts, lead to the same result. The latter found that the constituents of the salts examined diffused, not in the proportion in which they exist in the double salt, but nearly as if diffused sepa- rately, kiidorff,' following the same mode of investigation, divided double sn.lts into two classes, according as their con- stituents were found in the diffusate in the same proportion as in the double salt or in a different proportion. Riidorff also investigated the relation between dissociation and concentra- tion and found that, until near the point of saturation, the 'two were independent ; but that if crystals of the double salt were placed upon the diaphragm so as to maintain saturation, the proportion of the constituents diffused much more nearly approached their proportion in the double salt. He inferred that molecular compounds exist in fully saturated solutions. This inference was criticised by Ostwald,' and later by Tre- vor." on the ground that the effect observed was due to the more rapidly diffusing constituent becoming relatively less concentrated in the diffusion vessel ; and the latter undertook a series of experiments confirming this view. On the other hand, E. Fischer and Schmidner,' by allowing a saturated solution of ferrous ammonium sulphate to diffuse upwards through rolls of filter paper inside a glass tube, found the proportion of the constituents in the filter paper the same as in the double salt. In the field of thermochemistry, conclusions as to the ex- istence of double salts in solution are based upon a principle stated by Berthelot' as follows : " Everything indicates that double salts formed with a feeble disengagement of heat are to be regarded as separated in greater part into their constitu- 1 Iiiaujrnral Dissortatiovi. Leyden, (1869). 2 Ber. d. chem. Ges., la, 1678, (1879). » Uer d. chem. Ges., ai, 4 ; 3i. 1S82 ; ai, 3044. (1888) ; 33. 1846, (1890). 4 Ztsrhv. phys. Chem., 3, 60:, (18S9). & Ztschr. phys. Chem., 7, 468, (1891). « Ann. . 'i.-m. (Uebig), a?*. 156. (1892)- ' M^c. Chem., II, 324- jscrvatlotis that -alts capable of utiion can take y formed at the der Wal.' who d of Ingenhoes,' iiilar salts, lead e constituents of portion in which if diffused sepa- jf investigation, ng as their con- nie proportion as ti. Riidorff also 1 and concentra- f saturation, the f the double salt ntain saturation, uch more nearly lit. He inferred irated solutions, itid later by Tre- l was due to the ig relatively less ; latter undertook V. On the other wing a saturated diffuse upwards , tube, found the iper the same as ons as to the ex- upon a principle [ig indicates that ement of heat are nto their constitu- etn. Ges., la, 1678, (1879). 1846, (1890). lys. Chem., 7,468, (1891). a., 11,324- ents by water." On this principle Havre and Valson,' find- ing that the constituents of the alums evolved no heat oh mixing, concluded that the latter come into existence only throu^'h crystallization. Graham' had previou.sly made simi- lar experiments with other double sulphates, reaching the same results. Graham^ found, on the other hand, that ther- mal changes occurred when solutions of the chlorides of mer- cury and ammonium were mixed. vSimilarly, Berthelot* found that in general the thermal changes on mixing solu- tions of the halogen salts of mercury and potassium were not equal to the sum of the thermal changes for the component salts, and concluded that the existence of double salts in such solutions was thereby proved. Of importance in their bearing upon the question are also the volume changes which occur when salt .solutions are mixed. Kremers,' finding no change of volume on mixing solutions of salts capable of forming double .salts, inferred that no double salts existed in solution, on the ground that chemical changes are known to be accompanied by changes of volume frequently large. Favre and Valson* drew a simi- lar conclusion from the fact that the density of a solution of potassium cupric sulphate is the mean of the densities of its constituents. Gerlach,' however, noticed a slight contraction of volume in the case of the alums on mixing solutions of their constituents, which became more marked with increas- ing concentration ; and upon this and other grounds he in- clined to the belief that these double salts exist as such in their solutions. The conclusions of Groshans^ from observa- tions on double chlorides and sulphates confirmed those of Kremers. Researches upon the solubility of salt mixtures, especially upon the states of equilibrium in solution when two salts capable of forming a double salt are dissolved together, have contributed largely to a knowledge of the state of double salts in solution. Mulder' drew from the work of Kopp'" and 1 Compt. rend., 74, 1165. (1872). ■ Phil. Mag., 34, 401, (1844). 6 Pogg. Ann., 98, 58, (1856). •» Ztschr. anal. Chem., a8, 485. (1889). « Jahrsb. Chem., 1864, 92 ; 1866, 65. a Phil. Mag., jo, 539, (1842). 4 Ann. chim. phys., [5], 39, 198, (1883). 8 Compt. rend., 77, 907, (1873). 8 Des Dissolutions Aqueuses, 2-7. 10 Ann. chem. (Uebig), 34, 260, (1840). I others, cxtciulccl liy researches of hisown, the conclusion that well-defined double salts exist in saturated solutions of salt mixtures. This conclusion was founded upon the observa- tion that ill such solutions, saturated with respect to each component, the salts are often found in simple molecular pro- portions. Some researches of Riidorff' seemed to lead to a different conclusion. These, however, have been otherwise interpreted by Trevor.' The studies in equilibrium of Ditte' on the double iodide of lead and potassium (PbI,.K,I,.«H,/)). of Roozeboom* on astrakanite, of Meyerhoffer," Vriens," vSchreinemaker,' and van der Heide" have led to the {general result that there are certain limits of temperature outside of which double salts are capable of existing in their soh-tions. From van't Hoff's extension of the gas laws to .solutions Nernst" deduces the consequence that if to a saturated salt solution a solution of another salt having a coiumon ion be added, the solubility of the former salt is decreased and a ■part of it precipitated. On examining certain cases in which the solubility of the first salt is increased under the conditions stated, Le Blanc and Noyes'" succeeded in showing that the apparent exceptions were due to the formation of double salts in solution. To the same cause Rose" attributed those cases in which a salt is found to be more soluble in the solution of another salt than in water. The view that the solution of silver chloride in ammonia is accompanied by the formation of a double salt in the solution has been confirmed by Bod- lander." The application of electrolysis to a study of the state of double salts in solution was made early in the present cen- tury by Porret," who found that when potassium ferrocyanide was electrolyzed, alkali appeared at the negative pole, while oxide of iron and prussic acid appeared at the positive. The work of Daniell and Miller" on the same salt, extended by Hit- l Fogg. Ann., 148, 558, (i873). ' ^<^- "'• » Ann. chem. phys., [5], a4, 226 (18S1). 4 Ztschr. phys. Cliem., a, 513. (1888) ; la, 35Q (1893). 6 Ztschr. phys. Chem., 3. 336. (1889). " Ztschr. phys. Chem., 7. 194. (1891)- 7 Ztschr. phys. Chera., 9, 57. (1892) ; H. 289, (1893). 8 Ztschr. phys. Chem., la, 416, (189.O. » Zts.:hr, phys. Chem., 4, 372. ('889). 10 Ztschr. phys. Chem., 6. 38.5. (1890). '» Poge^ Ann.. 8a, 545, (1851). W Ztschr. phys. Chem., 9, 730, (1892)- " Phil. Trans., 1814. 527. 14 Pogg. Ann., 64, 18, (1845). ; conclusion that solutions of salt wn the obscrva- respect to each .' molecular pro- led to lead to a been otherwise ibriuni of Ditte' bI,.KJ,.«H,/)), loffer," Vriens," d to the general ature outside of their solftions. iws to solutions a saturated salt I coiunion ion be decreased and a n cases in which ler the conditions lowing that the )n of double salts buted those cases 1 the solution of t the solution of by the formation snfirmed by Bod- y of the state of the present cen- iium ferrocyanide ative pole, while le positive. The extended by Hit- I. Chera., 7. 194. (1891)- 1. Chem., 4, 372, (1889), 8a, 545, (1851). , 1814. 527. torf to many other double salts, lead to the view that double salts are of two classes : ( 1 ) those which tlo not exist as such in solution, and (2) those containing a metal coml)iued with a complex radical, which are capable of existing in water solution undecomposed. In the nomenclature of Ostwald* the latter are "complex salts," the term "double salt" being applied to members of the first class only. To the complex salts he refers potassium ferrocyanide and analogous salts; to the double salts he refers the alums. Retwecn these extreme types various degrees of dissociation may exist. Researches upon the electrical conductivity of double salts and salt mixtures are numerous. Of chief imi)ortance in the present connection are those of Bouchotte' and Paalzow* on sulphates of copper and zinc in common solution ; of vSvenson' on alums, of Grotrian' on the salt K,CdI,, of Freund,' Ben- der," Bouty,' Arrhenius,"and Chrouslchoff and I'ackhoff" on mixtures of electrolytes ; of Klein'" on mixtures and double salts: of Wershoven" on dilute solutions of cadmium salts; and of Kistiakowsky'* on dilute solutionsof "complex" salts. The results of these investigations, in so far as they refer to the dissociation of double salts, agree in showing that with increasing dilution, many double salts rapidly approach the condition of a simple mixture. For concentrated .solutions, Klein, who compared the conductivities of salt mixtures capable of forming double salts with that of those whose con- stituents were without mutual chemical action, showed that in the former case the difference between the conductivity of the mixture and the mean conductiviiy of its constituents was much greater than in the latter case ; and he inferred that in concentrated solutions there is only partial dissocia- tion. For complex salts Kistiakowsky found that e\ en in very dilute solutions their complex ions remained undecom- posed. The conductivity method has also been applied by 1 PogK. Aim., 106, 51.1, (i8,s9)- * Ztschr. phys. Chem., 3, 596. (1889)- 8 Compt. rend., 6a, 955. (1S66). * Pogg. Ann., 136, 489. (1H69). » Beibl, a, 46, {187S). » Wied. Ann , 18, 177. ("883). ^ Wied. Ann., 7, 44, (i879). ' Wied. Ann., aa, 179, (1884). 9 Ann. cliini. phys. [6], 3, 433. (1884) : 14. 74. (1888). ><» Biebl., 9, 437, (1885) : Wied. Ann., 30, 51, (i>^87). 11 Compt. rend., 108, 1162, (1889). W Wied. Ann., a?, 151, (1886). >8 Ztschr. phys. Chem., s, 481, (1890). " Ztschr. phys. Chem., 6, 97. (iS9o)- i^,m milinffia lO Le Blanc and Noyes,' by Bodliincler' and others to indicate the existence of double salts in the respective cases studied. The application of the freezing-point method has been especially studied by Raoult.' By comparing the lovering given by double salts with the sum of the lowerings of their constituents, he concluded that certain double sulphates, in- cluding the alums, are entirely dissociated, while other salts examined, as the chlorides of mercury, go into solution with but partial decomposition. From the foregoing summary it will be seen that the classi- fication of double salts into two groups according as they are wholly or but partially decomposed by water is well estab- lished so far as regards dilute solutions, but that in more con- centrated solutions the evidence is not sufficient to regard the classification as final. The present investigation has been undertaken with the object of obtaining such further data as may justify more definite conclusions. The work has been confined to .some alums as representatives of an extreme type of dissociation, and to one member of the dissociated class of double chlorides. The results reached are based upon a com- parison of the electrical conductivity and cryoscopic behavior of the double salts with that of their constituent salts, with a view to determining to what extent the solutions of the double salts correspond to mixtures. CotidticHvity Apparatus. The measurements of conductivity were made by the Kohlrausch method, using a Wheatstoue bridge, induction coil and telephone. The special form of the apparatus used was that described by Ostwald." The bridge wire was cali- brated by the method of Strouhal and Barus.' Conduc- tivity cells of the Arrhenius form were employed with elec- trodes at different distances, one for solutions less than o.ooi normal, the other for those more dilute. In this way satis- factory tone minima were obtained throughout a wide range of dilution. The temperature selected for the measurements wa3, in general, 25°. By means of a thermostat the tempera- 1 Loc. cit. * f'O':- "i- 8 compt. rend., 99, gM. OWH)- * Ztschr. phys. Chem., a. 561. (1888). B Wied. Ann., 10, ii6, (1880). others 1o indicate s cases studied, nethod has been ing the lovering )\verings of their )le sulphates, in- while other salts iito solution with en that the classi- rding as they are ter is well estab- that in more con- fficient to regard stigation has been h further data as le work has been [ an extreme type issociated class of based upon a com- yoscopic behavior uent salts, with a tious of the double ?re made by the bridge, induction le apparatus used Ige wire was cali- Barus.' Conduc- iployed with elec- )ns less than o.ooi In this way satis- hout a wide range the measurements lostat the tempera- s. Chem., a, 561. (1888). II ture was easily kept constant under ordinary conditions to within o. i degree. Solutions. The flasks and pipettes used in making up solutions were calibrated for a temperature of 20°, the former by the ap- paratus devised by Professor Morse of this laboratory and Blalock,' the latter by weighing in successive portions the water delivered at a known temperature. The solutions were prepared at 20° and standardized by analysis of a measured volume except in the few cases in which the concentration could be known from the weight of the salt dissolved. As the character of the water effects appreciably the con- ductivity of solutions more dilute than 0.02 normal, the water used was prepared by redistilling ordinary di.stilled water successively from solutions of alkaline and acid permanganate of potash, employing a block tin condenser. The water so prepared gave, in general, a conductivity at 25° varying from 1.5 to 2 X io"° mercury units. The conductivity of the water used was measured for each series of determinations, and tne correction applied. Conductivity Measurement, In making a series of conductivity measurements the general method adopted was to make up from the original solution, standardized by analysis, a .solution which was a simple multiple or fraction of normal ; 25 or 50 cc. of this solution were then measured with a pipette into the cell and successive solutions prepared from this in the cell itself by withdrawal of a measured portion and addition of an equal volume of water. As a check upon errors of dilution, a series of solutions were made up independently in measuring flasks at intervals corresponding to four or more dilutions in the cell. These solutions were used as standards, and when the conduc- tivity of the cell dilution differed a proportional correction was applied, in the manner indicated by Kohlrausch,'' to the members of the series immediately preceding, provided this correction did not exceed 0.5 or 0.75 per cent ; when it did, Am. Chem. J., l6, 479. (1894). * Wied Ann. , a6. 1S4, (1885). M>. utmmsmaK^^ "Hfc 13 the measurements were repeated. The cell was standardized at frequent intervals with 0.02 normal potassium chloride. Certain cases occurred in which the solutions were found to be unstable at great dilutions. In these cases the check solu- tions were made up immediately before measuring their con- ductivity. A Method of Titrating Alunmiiufn. As the necessity of standardizing solutions containing aluminium often occurred in the progress of the work, a reliable and convenient method of titration became desirable. None of the methods described in the literature seem to have proved satisfactory. That of Bayer' by titration of sodium aluminate with sulphuric acid, using litmus and methyl orange as indi- cators, is described as a "tedious, hot and uncertain pro- cess ;'" while, in the method of Lunge,' the ratio of the acid . consumed to the aluminium has been the subject of much dis- cussion. After some unsuccessful attempts with barium hydroxide, the following method of titration with ammonia, using litmus as an indicator, was finally adopted, and has been found to give fair results : A measured volume of the alum solution is heated nearly to boiling, and the sulphate precipitated with barium chloride avoiding a large excess. A 0.05 normal solution of ammonia is now added slowly and with constant shaking until the solution reacts weakly alkaline. The change of color is very gradual and could not be used in direct titration. The solution is now diluted to a known volume, the whole operation being conducted in a 100 cc. flask, and allowed to stand some hours, until most of the pre- cipitate has settled. Some of the solution is then decanted off and passed through a small filter, the first part of the filtrate being thrown away to avoid change in concentration. 50 cc. of the filtrate are next measured off, a few drops of litmus again added and the solution titrated to color with standard hydrochloric acid. A solution of potassium alum titrated in this way gave the following results : I Ztschr. anal. Chem., 34. 54J, (-SRs)- » K B., Ztschr. anal. Chem., as. '83- («»86). 8 Ztschr. angew. Chem., aa?, ags. ('890)- •SHN* 'as standardized uni chloride, tiswere found to 5 the check solu- uring their cou- ions containing e work, a reliable [esirable. None n to have proved odium aluniinate 1 orange as indi- d uncertain pro- ratio of the acid aject of muchdis- iriura hydroxide, nia, using litmus as been found to alum solution is precipitated with A 0.05 normal nd with constant alkaline. The i not be used in uted to a known :ted in a 100 cc. il most of the pre- is then decanted part of the filtrate entration. 50 cc. r drops of litmus lor with standard this way gave the al. Chem., asi 183. (i»86). 10). 13 I. I cc. solution contains 0.0948 gram alum. II. I cc. " " 0.0946 " III. I cc. " " 0.0945 " IV. I cc. " " 0.0944 " V. ICC. " " 0.0944 " Mean, 0.0945 " " The results of a gravimetric analysis by precipitation of the sulphate with barium chloride gave 0.09486 gram alum per cc. I. Alums. Double sulphates conform to one of two types : I. R', SO,. R"SO, + 6H,0. II. R', SO,. R'", (SO,),+ 24H,0. The alums, representing the latter type, form a group whose members are united by the closest analogies. Although taken by Rose' as the chief representives of a class of double salts not easily decomposed by water, since he found that repeated crystallizations gave rise to no decomposition, later investigators, as Ostwald and Raoult, regard them as types of extreme dissociation. Any conclusions reached, therefore, with regard to the alums from this point of view become im- mediately applicable to a large group of double salts. Conductivity Results. As no conductivity data at 25° for sufficiently concentrated solutions of the constituents of the alums were available, measurements were made for potassium sulphate, aluminium sulphate and chromium sulphate. Potassium Sulphate. The salt, purified by repeated crystallizations, was dried to constant weight in an air-bath, and solutions made up from a weighed amount of the dry salt. In the following tables conductivities are expressed in " molecular conductivity" units as defined by Ostwald, ^^ in- stead of in "equivalent molecular" units, in order to facilitate the comparison of the conductivities of solutions containing I Pogg. Anu., 8a, 545. (1851). 2 tehrbuch der allg. Chem., a, 621. .! ; f i 14 equal numbers of gram-molecules per litre but of different concentration expressed as normal. The symbols, /i„i5°. >".25^ denote respectively the molecular conductivities at 15° and 25°, the numbers in the columns being multiplied by 10'. The only direct measurements at 25° available tor compari- son are those of Walden,' which give by interpolation a very satisfactory agreement. Walden's results are as follows : Litres per gram-molecule. 64.0 128.0 256.0 512.0 1024.0 2048.0 /^i,25 . 232.2 246.0 256.8 265.4 272.8 278.6 K,SO. [I7434]- Gram-molecules per litre. 0.333 0.25 0.2175 0.10855 O.I 0.05427 0.05 0.0333 0.025 0.01666 0.005 0.0025 0.0005 0.00025 0.00005 Litres per gram-molecule. 0.3 4.0 4.6052 9.2104 10. 18.4208 20.0 30.0 40.0 60.0 200.0 400.0 2000.0 4000.0 20,000.0 /^«/i5 134-7 148.2 ■ • • • 162.9 167.2 171. 7 179.4 200.7 /^z>25 . 157-4 167.9 170.2 186.9 187.0 202.8 205.1 213.9 220.3 229.1 252.4 262.2 276.5 281.9 279.6 237-8 Aluminium Sulphate. The ordinary chemically pure aluminium sulphate was found to contain considerable quantities of sodium. The aluminium sulphate used was prepared from pure potassium alum. The aluminium was precipitated as hydroxide by am- monia, and the precipitate washed until it was found to be entirely free from potassium and ammonia. The washing was effected by boiling the precipitate with large quantities 1 Ztschr. phys. Chem., 3, 49. (i88S). ; but of different :ely the molecular ; in the columns lable for compari- terpolation a very ire as follows : ^25°. 132.2 !46.o J56.8 !65.4 272.8 278.6 /^f25 . 157-4 167.9 170.2 186.9 187.0 202.8 205.1 213.9 220.3 229.1 252.4 262.2 276.5 281.9 279.6 lium sulphate was of sodium. The )m pure potassium s hydroxide by am- it was found to be nia. The washing th large quantities 888). 15 of water, allowing it to settle and successively decanting and filtering. This treatment was repeated a dozen or fifteen times. The pure aluminium hydroxide was dissolved in sul- phuric acid in slight excess, as indicated by methyl orange, and the excess of acid removed by repeated precipitation of the salt with alcohol, the last traces of alcohol being re- moved by heating in an air-bath. The salt so obtained was in the form of a light powder and answered every test of purity. Analysis gave the following resuls : Calculated for A1,(S04),. Al 0.0479 SO, 0.2548 Analysis of the standard solution gave : I. 0.044706 gram A1,(S0J, per cc. II. 0.044778 gram A1,(S0,), per cc. Mean, 0.04474 gram A1,(S0,). per cc, equivalent to 0.1307 gram-molecules per litre. Found. 0.048 0.255 A1,(S0J. [342.34]- Oram- moleciiles per litre. O.21715 0.10855 0.06666 0.05427 0.03333 0.016666 0.003333 0.001666 0.0005 0.0003333 0.00025 0.0001666 0.00003333 0.000025 0.0000166 0.0000125 0.00000625 ivitres per gram- molecule. 4.6052 9.210 15.0 18.4208 30.0 60.0 300.0 600.0 1999.9 3000.0 3999-9 6000.0 30000.0 40000.0 60000.0 80000 o 160000.0 ;^i-i5 - • • • • 107.7 • • • • • • • • 139.1 • • • • 243.6 • • • • 382.4 • • • • 447-4 « « • • • • • • 678.4 • • • • 708.7 715.2 1. 105.7 130.3 148.2 153-6 172.5 202.3 300.4 359-5 II. O /^x/25 ■ Mean. 148.5 148-3 172.7 202.1 300.8 358.6 172.6 202.2 300.6 359.0 530.0 531-7 530.8 611.6 806.1 511.4 809.1 611. 5 807.6 869.4 869.4 869.4 i6 These results for aluminium sulphate show a good agree- ment with Walden's measurements of dilute solutions at the same temperature. Chromium Sulphate. The chromium sulphate used was a specimen prepared under the direction of Professor Renouf of this laboratory. The process consisted in the reduction of chromic acid with ether under a bell-jar, filtration of the solidified crystalline mass with the aid of a pump, and careful drying of the prod- uct on plates of unglazed porcelain.' A concentrated solution of the salt was prepared. To diminish errors of analysis, the solution standardized was pre- pared from the original by diluting one volume to twenty. Analysis gave : I. 0.007426 gram Cr,(SO,), per cc. II. 0.007410 gram Cr,(S0,)5 per cc. Mean, 0.007418 gram Cr,j(vSOJ, per cc. Hence the original solution contained 0.3779 gram-mole- cules per litre. The conductivity values obtained by Walden" differ widely from those in the following table for the conductivity of chro- mium sulphate. A comparison of a few approximately cor- responding dilutions will show the extent of the difference. Column I gives Walden's values reduced to molecular con- ductivity units ; column 2, those from the following table for chromium sulphate at the same temperature : Column 2. ;Uj, 25'' = 288.2 (1"= 200) 346.5 {v=^ 400) 558.2 (j/= 2000) Column I. A*.'25°= 378.0 {v= 192) 439.2 {v= 384) 589.2 {v= 1536) 669.6 {v •= 3072) If the mean be taken of the conductivity values obtained by Walden for chromium sulphate and those corresponding for potassium sulphate, it leads to values for the conductivity of potassium chrome alum greatly in excess of those obtained in the present investigation. The conductivity measurements for chromium sulphate are shown in the lollowing table : I Renouf: Inorganic Prep., p. 78. 2 Ztschr. phys. Chem., 1, 541, (1887). £ a good agree- slutions at the imen prepared his laboratory, amic acid with fied crystalline ig of the prod- prepared. To irdized was pre- mie to twenty. '79 gram-tnole- :n° differ widely ictivity of chro- roximately cor- the difference. molecular con- owing table for iluran 2. 88.2 (V= 200) 46.5 iv= 400) 58.2 (V = 2CXK)) iralues obtained )rresponding for conductivity of hose obtained in iim sulphate are lem., I, S4i> (1887). 17 Cr,(SO,). [392.58]. Gram-moltcules Litres per _ . per litre. Kraiii-molecule. A'fi5 • A'j'25 0.333 30 • • • • 94.2 0.25 4.0 86.2 107.3 O.I 10.0 • • • • 139.0 0.05 20.0 132.3 162.7 0.025 40.0 • • • « 190.7 0.005 200.0 233-2 28S.2 0.0025 400.0 • • • • 346.5 0.0005 2000.0 • • • • 558.2 0.00025 4000.0 • • • • 671.2 Potassium Aluminium Alum. The salt was purified by repeated crystallizations, and was found to be free from iron, sodium and ammonia. A solution was prepared approximately saturated at 20°. The solution was standardized by precipitation of the sulphuric acid with barium chloride. The results of analysis were : I. 0.0952 gram KA1(S0,),-|- 12 H,0 per cc. II. 0.0948 gram KA1(S0,),+ i2H,0 per cc. III. 0.09458 gram KAKSOJ, + i2H,0 per cc. Mean, 0.09486 gram KA1(S0,),-|- i2H,0 per cc. of solu- tion, or equivalent to 0.19995 gram-molecules per litre. The conductivity obtained is shown in the following table : KAl(SO.),-f i2H,0 [4744I]. Grnm- Litres per L u. molecules :'ram- n per litre. molecule. /^z'25 . ;'r.25 . Mean 0.19995 5.0012 133-9 133.9 133-9 0.125 8.0 149.2 149.2 149.2 0.05 20.0 178.3 178.3 178.3 0.025 40.0 202.5 202.5 202.5 0.005 200.0 268.2 269.9 269.0 0.0025 400.0 304.2 306.3 305.2 0.0005 2000.0 407.0 406.8 406.9 0.00025 4000.0 468.2 466.8 467-5 Sodium Aluminium Alum. The alum was made to crystallize from a water solution of its constituents by pouring a layer of alcohol upon the surface of the solution. With the gradual diffusion of the alcohol large, well-formed crystals of soda-alum, together with a few f. ! u i8 crystals of sodium sulphate appeared. When the mixture was washed quickly with cold water the small crystals of so- dium sulphate dissolved, leaving the alum perfectly homo- geneous throughout. Analysis of a standard solution gave : [. 0.02402 gram NaAl(SO,),-h i2H,0 per cc. II. 0.02407 gram NaAl(SO.),+ i2H,0 per cc. Mean, 0.024047 gram alum per cc. equivalent to 0.05247 gram-molecules per litre. Conductivity results are shown in the following table : NaAl(SO.),-hi2H,0 [458.33]- Oram- molfcules per litre. 0.05 0.025 0.005 0.0025 P.OOO5 0.00025 Ulres per Kr.-\m- molecule. 20.0 40.0 200.0 400.0 2000.0 4000.0 I. 161.5 184.4 250.0 286.7 380.0 435-5 II. ;^^25°. 161.6 184.7 251.2 284.2 378.8 435-2 Mean. 161.6 184.6 250.6 285.4 379-5 435-5 Ammonium Aluminium Alum. After eleven crystallizations the salt was found to be free from sodium, but still retained a minute trace of potassium. The standard solution was analyzed by titration of the alumin- ium. The analysis gave the following results : I. 0.04550 gram (NH,)A1(S0J,+ i2H,0 per cc. II. 0.04585 gram (NH,)AUSO,),+ I2H,0 per cc. Mean, 0.04567 gram alum per cc, equivalent to 0.10076 gram-molecules per litre. (NHJAl(SOJ, + i2H,0 [453-19]- I. n. 152.8 174-5 198. 1 261.4 297.1 390.0 452.4 568.4 608.4 Gram- molecules per litre. O.I 0.05 0.025 0.005 0.0025 0.0005 o 00025 0.00005 0.000025 Utres per gram- molecule. 10. 20.0 40.0 200.0 400.0 2000.0 4000.0 20000.0 40000.0 152-9 175-0 198.4 260.6 294-9 389.2 452.4 560.0 602.8 M ean. 152.8 174.8 198.2 261.0 296.0 389-6 452.4 564.2 605.6 the mixture rystals of so- rfectly homo- alution gave : ;c. nt to 0.05247 tig table : Mean. 161.6 184.6 250.6 285-4 379-5 435-5 and to be free of potassium. . of the ahirain- >er cc. per cc. ent to 0.10076 9]. - M ean. 9 4 6 152.8 174.8 198.2 261.0 9 2 296.0 389-6 4 8 452-4 564.2 605.6 19 Potassium Chrome Alum. The salt, purified by repeated crystallization from water at 35°, was dissolved and the solution standardized by precipi- tation of the sulphuric acid as barium sulphate. The analy- sis gave : I. 0.11742 gram KCr(SO,), + i2H,0 per cc. II. o.ii746gram KCr(SO,),-f- i2H,0 per cc. Mean, 0.1 1 744 gram chrome alum per cc, equivalent to 0.2351 gram-molecules per litre. Following are the results of conductivity measurements : KCr(S0J,-j-i2H,0 [499-53]. Gram- molecules per litre. Litres per gram- molecule. O.I TO.O 0.05 20.0 0.025 40.0 0.005 200.0 0.0025 400.0 0.0005 2000.0 0.00025 4000.0 I. II. /^.25°. A'„25°. Mean. 147.8 170-5 195-4 266.7 148. 1 170.3 195.0 266.5 147-9 170.4 195-2 266.6 306.1 418.6 472.4 304-9 418.8 471.6 305-5 418.7 472.0 Ammonium Chrome Alum. The salt was purified as in the case of the potassium chrome alum, and the solution standardized by precipitation of the sulphate. The analysis gave : I. 0.09266 gram (NH.)Cr(SO,), -|- i2H,0 per cc. II. 0.09282 gram (NHJCr(SO,), + i2H,0 per cc. Mean, 0.09274 gram alum per cc, equivalent to 0.194 gram- molecules per litre. The conductivity observed is shown in the following talile : (NHJCr(SO,),+ i2H,0 [478.44]- Gram- molecules per litre. Litres per grom- molecule. /'.25°- O.I 10. 145-4 0.05 20.0 167.2 0.025 40.0 I9I-5 0.005 200.0 263.8 0.0025 400.0 304.0 0.0005 2000.0 .... 0.00025 4000.0 484.6 II. /Wri25°. 144.7 167. I 191. 1 263.6 303-6 4J5-3 477-5 Mean 145.0 167. 1 I9I-3 263.7 303-8 415-3 481.0 I T" 20 Green Modification of Ammonium Chrome Alum. As is well known, solutions of chrome salts at a tempera- ture between 70° atul 80° lose their original color, passing to f. green, and cease to have the power of depositing crystals. Recoura' has shown that the green modification is a well- defined basic salt mixed with free acid ; and Monti' has found that the change i accomplished by a rise in conductivity. With a view to fiiuling whether the passage to the green modi- ficatior is discontinuous at any point, a series of qualitative experiments were made by placing the conductivity cell con- taining a solutii n of ammonium chrome alum in a bath, and regulating the heat so that the temperature rose uni- formly with t)io time, noting at the same time, at regular interval.*, the rise in conductivity. The latter showed no rapid rate o^ change, and it may be inferred that the pas- sage to the gieon modification is a continuous change. Ammonium Iron Alum. A solution of the pure alum was standardized by titration of the reduced ferric salt with potassium permanganate, the reduction having been effected by a platinum-zinc couple. The results of analysis were : I. 0.24772 gram (NHJFe(SO,), + i2H,p per cc. II. 0.24722 gram (NH,)Fe(SO,), + I2H,0 per cc. Mean,o.2t747 (NH,)Fe(vSOj, + i2H,0percc., equivalent to 0.5080 gram-molecules per litre. Conductivity results are shown in the following table : UNH.)Fe(SO.),-f i2H,0 [482.24]. Gram-m()!ecu!es per litre. o,::5 o 0.5 o 025 o.cx)5 Litres per gram-molecule. 4.0 20.0 40.0 200.0 118. 9 177.4 211. 5 320.2 StibiUty of Alums in Dilute Solutions. In the course of the preceding measurements, it was noticed that the conductivities of dilute solutions of some of the alums » Ann. chim. phys., [7]. 4. 494. (i«9.s)- a Zlschr. auorg. CUem., la, 75. Refer., (April, 1896.) |Ml)WtM<>Mpil|fcMagjMiSR ^J I« 'a W«».*^to--«-«'*":.Aiifa»-'a ^w * - -« ^ - Alum. It a tempera- ^r, passing to iting crystals, on is a well- )nti'' has found conductivity, le green niodi- of qualitative ivity cell con- ni in a bath, ure rose uni- ne, at regular ter showed no that the pas- hange. d by titration anganate, the n-zinc couple. er cc. per cc. cc, equivalent ng table : /'7'25°. 118.9 177.4 211.5 320.2 , it was noticed lie of the alums 2t incrca.scd at a more or less rapid rate on standing for a time. As such a cliange of conductivity must depend upon changed conditions in the solution, this observation indicates a tliffer- ent degree of stability in the presence of a large amount of water. Ammonium iron alum furnished the most marked iti- stance of instability. At a dilution of 200 litres the conduc- tivity of a solution of iron alum rose 3 per cent, in two hours. At 400 litres it showed the same increase in twenty minutes, and at a dilution of 2000 litres the rate of increase was about 1.2 per cent, per minute. This rate soon diminished and, at the end of twenty-four hours, an apparent state of equilibrium was reached after a total increase in conductivity of 30 per cent. At o" a dilution of 1000 litres is reached before decom- position begins. The precipitate which accompanies the de- composition does not become visible until after the change in conductivity has become apparent. A rise in conductivity was also noticed in the case of soda alum and ammonium chrome alum, but not until a dilution of 20,000 liters was reached. The former at this dilution showed an increase of 4 per cent, in twenty-four hours. Similar examples of a rise in conductivity in dilute solutions have been noticed by Kohl- rausch' in the case of barium nitrate, copper sulphate and other salts. In the former case it was clearly accompanied by the formation of a basic salt, and the liberation of free acid. In the case of chromium salts the rise in conductivity accom- panying the formation of a basic salt has already been men- tioned. The well-known tendency of ammonium iron alum to form basic salts is doubtless the cause of the change of con- ductivity noticed in its case. It seems probable, therefore, that a similar change occurs in very dilute solutions of soda and ammonium chrome alums brought about by the action of a large mass of water. Such a change would be analogous to that studied by Rose'' in the case of acid potassium sulphate, which is decomposed by an excess of water forming the neu- tral sulphate and free acid. Comparison of Conductivity Results. If the foregoing conductivities of the alums be compared, 1 Wied. Ann., a6, 175, (1885). » Pogg. Ann., 8a, S45. (•■851). it will be seen that throu^jhout all dilutions tlicy can be ar- ranKcd in order of magnitude with respect to their conductivi- ties from a knowled^;e of the conductivities of their constitu- ents. The conductivities of potassium, sodium and ammo- nium suli)hates are in the same order of magnitude as the conductivities of the cnrrespondinj; alums. A detailed comparison of the conductivities of the alums with the arithmetic means of the conductivities of their con- stituents is shown in the follov.inR tables. The data for the values of conductivity of aluminium sulphate have beeti ile- rived from the table already aWt-n, supplemented by WaU den's' results. The values given have been interpolated by the graphic method. The data for sodium sulphate are inter- polated from Kohlrausch's' tables reduced to 25" and changed to molecular conductivity units. The values for potassium sulphate and chromium sulphate are taken directly without interpolation from the tables already given. No data for am- 'nionium sulphate and ferric sulphate are available. As the latter only exists in acid solution, direct observations of its conductivity would have no value for the present purpose. /. Potassium Aluminium Alum. Litres molecule. KjSO,. AI,{SO,),. /'r25°. Alum, observed Aritliinetic ,, „_o mtun. f*v2^ , Difference. Per cent. 5.0012 172.7 108.0 140.3 133.9 —4.5 8.0 183.3 124.2 153.7 149.2 —3.0 20.0 205.1 158. 1 181.6 178.3 — 1.7 40.0 220.3 185.7 203.0 202.5 — 0.2 200.0 252.4 290.4 271.4 269.0 —0.8 400.0 262.2 342.6 302.4 305.2 +0.9 1 Ztschr. phys. Chem., i, 54i. (i***?). a Wicd. Ann., a6, iy6, (18S5). i«M»ii'rii, ey can be ar- icir coiuluctivi- their coiistitu im and anuno- piitiulc as the ; of the alums s of their con- le data for the have beeti ile- Mited by Wal- jiterpoUitcd by phate are inter- '.^" and changed for potassium irectly without Mo data for am- ilable. As the ervations of its 2nt purpose. observed. If25°. Difference. Per cent. 133-9 —4-5 149-2 —30 178.3 — 1.7 202.5 — 0.2 269.0 —0.8 J05.2 +0.9 //. Sodium ,Vnminiitlii Alum, I.ilrFit per writni- mol.cule 20.0 40.0 200.0 400.0 No,HO«. A'r25°. 171-5 183-1 211. 6 219. 1 i5«-i 18.5-7 290.4 342-6 mean. 164,8 184.4 251.0 280.8 Alum. AP^irrvcd. Mr25". I6I.6 184.6 750.6 285.4 Ulfferrnce, I'er cent, — 1-9 -1-0. 1 —0.2 + 1.6 ///, Potassium Chrome Alum, Mtren per Ktnitj- molecnle. K,SO,. /'r25°. Cr,(SO.),. ^'.25°. Arithmetic mean. Alum, observed ;<^25^ Difference. I'er cent. 10. 187.0 139.0 163.0 147.9 9.2 30.0 205.1 162.7 183.9 170.4 —7-3 40.0 220.3 190.7 205.5 195-2 —5.0 200.0 252.4 288.2 270.3 266.6 —1-3 400.0 262.2 346.5 304.3 305-5 +0.4 aooo.o 276.5 558.2 417-3 418.7 +0.3 4000.0 281.9 671.2 476.5 472.0 —0.9 A study of the preceding tables justifies the following state- ments : ( I ) In concentrated solutions the conductivity of the alums is notably less than the mean of the conductivities of their constituents. (2) The defect of the conductivity from the mean is the same in amount for the aluminium alums, while it becomes much greater for chrome alum, (3) The differ- ence between the conductivity of the alums and the mean con- ductivity of their constituents grows rapidly less with increas- ing dilution, so that at a dilution of 40 litres for the aluminium alums and of 400 for chrome alum it has disappeared. Deductions from these facts must depend upon the relations which exist between the conductivity of a mixture of two electrolytes having a common ion and the conductivities of the constituents taken separately — assuming that the state i ! 24 of the alum in solution is the same as if it had been formed by mixing equal volumes of solutions of its constituents having the same molecular concentration as itself. It has been shown, especially by the work of Bender' and Klein", that in such mixtures the conductivity is in general less than the mean of the conductitivities of the constituent .^alts. But Klein showed that a more marked difference existed when the salts were capable of forming a double salt. Thus while potassium sulphate and sodium sulphate gave a difference of I per cent., potassium sulphate with ferrous sulphate gave for the same concentration 6 per cent. The differences found for the alums as given in the above table are of the same order (2U0 II percent.) as those found by Klein for the double salts examined by him. The condition that two electrolytes having a common ion may mix without undergoing a change in ionization, has . been deduced from the dissociation theory of electrolytic con- duction by Arrhenius.' If no change of volume occurs on mixing, that condition is that the concentration of the ions, i. e., the quotient of the coefficient of ionization divided by the volume containing one gram-molecule, shall be the same for the both solutions, or where «,, «„ are the coefficients of ionization of the respective solutions ; «,, «,.the number of gram-molecules contained re- spectively in a unit of volume of each; and z*,, z\, the re- spective volumes of the solutions mixed. The following table shows the extent to which this condi- tion is fulfilled in some of the solutions of alums being studied. On the assumption already made, that we may regard a solu- tion of the alum as having been formed by mixing equal volumes of solutions of its constituents of the same molecular concentration, «, and «, in the preceding equation become equal as also z\ and v,, and it is only necessary to compare the values of a, and ct„ for the constituent solutions. The values of a have been calculated from freezing-point data in- stead of from conductivity, since the experimental error in 1 Loc. at. "^ ^-'"■- "■'■ « Ztschr. phys. Chera., a, 284, (1888). < 25 [lad been formed nstituents having f. It has been id Klein", that in ;ral less than the tuent jalts. But ce existed when salt. Thus while Lve a difference of sulphate gave for ferences found for if the same order n for the double ng a common ion n ionization, has )f electrolytic con- rolume occurs on ation of the ions, zation divided by shall be the same n of the respective :ules contained re- and z',, ^',, the re- which this condi- ums being studied, may regard a solu- . by mixing equal he same molecular ; equation become iary to compare the t solutions. The zing-point data in- jerimental error in 388). measuring n^ for aluminium sulphate and chromium sulphate may become very large. For potassium sulphate the two methods give concordant results, as is shown in the table given by Arrhenius.' In calculating the valuesof i, thatis, the ratio between the observed and the normal osmotic pressure, the value 1.88 is taken for normal molecular-lowering. The corresponding values of a are calculated from the equation ?■= /+ {k — /) a. where k denotes the number of ions given by each active molecule, namely, 3 for potassium sulphate and 5 for alumin- ium sulphate. AljCSOjj) Gram-molecules K,SO, per litre. i. a. 0.117 2.22 o.6i o.ro86 2.28 0.64 O.I 2,30 0.65 0.098 .... • ■ • • 0.088 2-33 C.66 0.0653 • • • • .... 0.059 2.41 0.70 0.054 2-45 0.72 0.05 2.44 0.72 0.0333 2-57 0.78 0.0294 2.57 0.78 0.026 .... • • • • Cr,(S04), a. 2.13 0.28 2.21 2-34 0.30 2.44 0.38 t. 2.20 • • ■ • 2.21 ■ • • • 2.28 2.40 2.44 2.55 n. 0.30 0.30 • • • * 0.32 • • • • 0-35 • • • • 0.38 0.387 2.60 o 40 .... It is seen that the values of a and, therefore, of or the concentration of the ions for potassium sulphate, differ widely from those for either aluminium or chromium sulphate: while, for the latter sulphates, the values of n show an almost com- plete agreement throughout the range of dilution given. The change of dissociation or ionization produced by mixing with potassium sulphate should, therefore, be very nearly the same in both cases, if we neglect the effect of changes in volume which in this case are extremely small. Reference to the pre- ceding tables I. and III., however, shows that this is by no means the case. The divergence from the mean is four or five times greater in the case of the chrome alum than for the aluminium alum at the same dilution , or, is also much greater if we compare them at equal distances from the point of satura- 1 Ztschr. phys Chem., a, 491, (1887). 26 tion. Even if we conclude, therefore, that the potassium aluminium alum is entirely dissociated we must infer that the molecules of the double salt in a concentrated solution of chrome alum are only partially broken down. This con?lu- sion with respect to chrome alum agrees with that reached on other grounds by Carey Lea,' who finds by the use of a reagent for the detection of free sulphuric acid in the presence of sulphates that chrome alum is the only one present as such in .solution. In order to find whether the solution of an alum is simply equivalent to a mixture of solutions of its equivalents, a num- of mixtures were prepared of equal volumes of solutions of potassium sulphate and aluminium sulphate of the same molecular concentration and also of potassium sulphate and chromium sulphate. The conductivities of the constituents salts were observed, then that of the mixtures. The results appear in the following tables : Potassium Sulphate and Aluminium Sulphate. Gram- molecules per litre. KjSO,. Alj(SO,),. /'t-25°. Mean. Differ- Mixture observed, eiice. Per /'„25 . cent. 0.2175 170.2 105.7 137-9 129. 1 -6.4 0.10855 186.9 130.3 158.6 I5I-8 —4-3 0.05427 202.8 153.6 17S.2 174.2 — 2 .,2 0.03333 213.9 172.5 193-1 190. 1 —1.6 0.01666 229.1 202.2 215.6 214.3 —0.6 Potassium Sulphate and Chromium Sulphate. Gram- olecules ;)er litre. KjSO,. Cr.j(SO,),. f^v25°. Mixture observed. Mean. /'r25 . Difference Per cent. 0.333 157-4 94-2 125.8 II3-0 lO.I 0.25 167.9 107-3 137-6 125. 1 —9.0 O.I 187.0 139.0 163.0 155-1 —4.9 0.05 205.1 162.7 183-9 177-9 —3-3 0,025 220.3 190.7 205.5 202.8 — 1.3 1 Ztschr. anorg. Chem., 4, 445, (i893)- lie potassium infer that the ;(1 solution of This coii?lu- lat reached on the use of a II the presence resent as such ilum is simply alents, a num- )f solutions of of the same sulphate and e constituents . The results Iphatc. Differ- jre observed, eiice. Per /'„25 . cent. 129. 1 -6.4 151-8 —4-3 174.2 — 2.2 1 90. 1 —1.6 214-3 —0.6 'phate. rved. Difference. Per cent. - -lO.I I —9.0 I —4.9 9 —3-3 8 —1-3 27 The results for mixtures of potassium and aluminum sul- phates show a close agreement with those obtained for the alum ; those for potassium and chromium sulphates show a somewhat remarkable difference — the alum having a consid- erably lower conductivity at corresponding dilutions. It may be remarked that volume-changes play an unimportant part in this instance since, according to Gerlach's ob-servations,' the contraction at 15° for a mixture of potassium and alumin- ium .sulphates containing 0.21 gram-molecules per litre is 0.06 per cent., and for a mixture of potassium and chromium sulphates containing 0.27 gram-molecules per litre is at the same temperature 0.12 per cent. If the difference noticed is well founded, it might be accounted for by supposing that the molecules of the double salt remain in part undissolved when its crystals are dissolved, but that the molecules of the con- stituent salts do not necessarily unite when their solutions are mixed. This is the explanation given by Graham' of the fact observed by him that the diffusion of the double sulphate of potassium and magnesium differed from that of a mixture of its constituent salts. Summary of Conductivity Results. The preceding results in so far as they bear upon the ex- istence of alums as such in solution may be summarized as follows : 1. The alums in dilute solution are entirely dissociated into their constituent salts. 2. In concentrated solution they have a lower conductivity than the mean conductivity of their constituents. The differ- ence is more marked as the concentration increases ; and is of the same order as that observed for other double sulphates, and greater than that observed in the case of mixtures of sul- phates incapable of yielding a double salt. There is, there- fore, some evidence that the alums are partially undissociated in concentrated solution. 3. The amount of the divergence from the mean conduc- tivity of its constituents for chrome alum as compared with the aUtminium alums affords strong evidence that this alum, at least, exists as such in solution. 1 Ztschr. anal. Cbem., a8, 505, (1889). ••J Phtl. Trans., 1850, i. JF 28 Freezing- Point Measurements. The apparatus used in tlie freezing-point observations was of the ordinary Beckmann form, except that the inner tube containing the sohitions was slightly larger in diameter to prevent the danger of the stirrer coming into contact with the thermometer, and had no side branch. The ice-bath was in- cased with felt. The solutions used were made up from the same solutions as those employed in conductivity work. In, every freezing-point determination the same water was em- ployed as that used in making up the solution in question. The following tables show the freezing-point lowenngs ob- tained for the alums and their constituent salts. The follow- ing svmbols are used : G denotes the number of grams of salt dissolved in a litre of solution. N denotes the number of gram-molecules per litre of solu- trion. L denotes the corrected lowering. A denotes the gram-molecular lowering. In the case of the alums the double formulas are used for convenience in comparison with their constituents, the sum of the lowerings for the constituents being then directly com- parable with the lowering of the alum. I. K,SO, [I74-34]- G. N. No G. I 37-857 N. I.. 0.217 0.890° 4.10 ][8.9285 0.1086 0.466 4.29 9.4643 0.0543 0.250 4-6i 5.8110 0.0333 0.161 4.83 2.9055 0.0167 0.084 5.03 No. 1 58.1134 0.333 2 43-5850 0.250 3 17-434 o.ioo 4 8.717 0.050 5 4-3585 0.025 II. A1,(S0,), [342-34]- No. G. I 44-740 33-555 22.370 15-659 8.948 6. 711 4-474 2 3 4 5 6 7 N. r.- No. 0.I3I 0.516° 3.94 I 0.098 0.410 4.18 2 0.0653 0.288 4.41 3 0.0457 0.212 4.64 4 0.0261 0.127 4.87 5 0.0196 o.ioi 5-15 0.0131 0.073 5-57 N. 0.217 L. 1.311° 1.004 0.432 0.230 0.122 0.832° A. 74-338 37.169 0.1086 0.43b 18.5845 0.0543 0.245 II. 411 0.0333 0.189 5-7055 0.0166 0.088 3-93 4.02 4-32 4.60 4.88 A. 3-83 4-03 4-51 4-77 5-30 29 srvations was he inner tube 1 diameter to ntact with the e-bath was in- ; up from the 'ity work. In, vater was em- iti in question, lowerings ob- . The follow- )lved in a litre ;r litre of solu- is are used for Luents, the sum n directly com- f. L. 133 1.3"" 3-93 !50 1.004 4.02 GO 0.432 4-32 ),SO 0.230 4.60 >25 0.122 4.88 L. J. 17 0.832° 3.83 o86 0.438 4-03 543 0.245 4-51 333 0.189 4.77 1 66 0.088 5-30 III. Cr,(SO.), [392.58]. No. G. N. I,. ^■J. I 130.860 0.333 • • • • .... 2 98.145 0.250 1.029" 4.12 3 39.258 O.IOO 0.417 4.17 4 19.629 0.05 0.230 4.60 5 9.8145 0.025 O.I2I 4-84 IV. (NH,),SO, [132.16]. No. (.. N, I- J. I 26.432 0.2 0.829° 4-145 2 13.216 O.I 0.437 4370 3 9.2512 0.07 0.316 4-514 4 6.608 0.05 0.237 4-740 5 3.9648 0.03 0.148 4-933 6 1.9824 0.015 0.075 5.00 V. K,A1,(S0J,-|-24H,0 [948.8 2]- No. G. N. I-. J. I 47-441 0.05 0.409° 8.18 2 35.5807 0.0375 0.320 8-53 3 23.720 0.025 0.222 8.88 4 "•8575 0.0125 0.120 9.60 5 4- 7430 0.005 0.057 11.40 VI. Na,Al,(SO,). + 24H,0 [916 66] No. G. N. h. J. I 22.9165 0.025 0.229° 9.16 2 11.4582 0.0125 0.125 10.00 3 5.7291 0.00625 0.009 11.04 VII. (NH,), Al,(SO.), + 24H,0 [906.6]. No. G. N. I-. J. I 45-330 0.05 0.408° 8.16 2 33-997 0.0375 0.318 8.48 3 22.665 0.025 0.224 8.96 4 "•332 0.0125 O.I2I 9.68 5 5.666 0.00625 0.066 10.56 VIII. K,Cr,(SOJ, + 24H,0 [999]- No. G. N. h. ^. I 117.44 O.II7 0.888° 7-59 2 88.08 0.088 0.686 7.80 3 58.72 0.059 0.480 8.14 4 29.36 0.0294 0.267 9.08 5 17.618 0.0176 0.170 9.66 6 11.744 O.OII7 O.II9 10.17 7 5872 0.0059 0.065 11.02 30 IX. (NH,),Cr,(SOJ, + 24H,0 [956.8I. No. I 2 3 4 5 6 No. I 2 3 4 5 6 7 8 G. 92.740 46., -,70 27.822 16.6932 I 1. 1 288 5-5644 X. N. 0.097 0.0484 0.0291 0.0174 o.oi 16 0.0058 0.768° 0.400 0.266 0.168 0.II7 0,064 J. 7.92 8.26 9.14 9-65 10.09 11.04 (NH,),Fe,(SO,). + 24H,0 r964.42]-^ 7.08 G. 247.72 185-79 123-86 92.895 61.930 37-158 24.772 12.386 6.193 N. 0.257 0.192 0.128 0.096 0.064 0.0385 0.0257 0.0128 0.0014 U. 1.820° 1.400 0.927 0.713 0.505 0.322 0.227 O.I2I 0.066 7.29 7-24 7-43 7-89 8.36 8.83 9-45 10.31 Comparison of freeziyig- Point Resxdts. Comparing the results for the different alums as shown in the foregoing tables, it will be seen that, as regards theahunm- um alums the soda alum (Table VI) gives a greater de- ;rssrd: fo; corresponding dilutions tljan either potass.urn (Table V) or ammonium alum (Table VII), whi e the depressions given by potassium alum correspond closel>^ XI are slightly larger in concentrated solution than ule for tl^ Limonium alum. For dilute solutions "reverse is the case; but the experimenta error is here greater and less value is to be attached to the resul^^ The ammoniun chrome and potassium chron.e alums alo show closely corresponding values for corresponding dilutions (Tablec VIII. IX), while these values are somewhat g eater or corresponding dilutions than thoseof the aluminium alums In the following tables is shown a comparison of the lower- ings given by the alums with the sum of the lowenngs for 1^? r'c onstituents at the same dilution. The values for the constituent salts are interpolated from the preceding tables I to IV. 31 6.8]. Potassium Aluminium Alum. A. 7.9a 8.26 9.14 965 10.09 11.04 4.42]. J. 7.08 7.29 7.24 7-43 7-89 8.36 8.83 9-45 10.31 mlts. US as shown in ;ardsthealumin- es a greater de- iither potassium ^11), while the rrespond closely, [ solution than dilute solutions imental error is d to the results, rome alums also ponding dilutions somewhat greater iluminium alums, rison of the lower- the lowerings for bie values for the )receding tables I No. I 2 3 4 No. I 2 3 4 No. No. I 2 3 4 N. 0.05 0.0375 0.025 0.0125 4.61 4.76 4.89 5.02 ,(SO,),. J. Sum. PotHSsium alum. Z/ obaerveU Differ eiice. 4.60 9.21 8.18 1.03 4.72 9.48 8.53 0.95 4.91 9.80 8.88 0.92 5-61 10,63 9.60 1.03 Ammonium Aluminium Alum. N. 0.05 0.0375 0.025 0.0125 (N11,),S04. J. 4-74 4.86 4-95 5.02 Al,(SO,), J. 4-55 4.72 4. ox 561 Sum. 9.29 9-58 9.86 10.63 Potassium Chrome Alum. N. O.II7 0.088 0.059 0.0294 K,SO,. J. CrjCSO,),. J. 4.28 4.16 4.38 4.27 4-54 4-51 4.83 4.80 Sura. 8.44 8.65 905 9-63 Amm.-A'.. alum. A observed. 8.16 8.48 8.96 9.68 K-Cr alum. A observed. 7-59 7.80 8.14 9.07 Ammonium Chrome Alum. (NH4),SO,. J. Cr,(SO,),. J. Sum. Amm.-Cr alum. A observed. 7.92 8.26 Differ- ence. 1.I3 1. 10 0.90 0.95 Differ- ence. 0.85 0.85 0.91 0.55 Differ- ence. 0.65 I. II 0.00 1 0.097 4-38 4-19 8.57 2 0.0484 4.76 4.61 9.37 3 0.0291 4.93 4-8i 9-74 9H These tables show that for the most concentrated solutions the lowerings given by the alums, with the exception of ammo- nium chrome alum, is uniformly 10 or 11 per cent, less than he sum of the lowerings for its components The abnormal ■"f -y j* -);^.-- 32 value for ammonium chrome alum— giving 7.3 per cent.— may be attributed to experimental error, since the other values ob- tained for it are in agreement with those of the other alums. Raoult' found for potassium aluminium alum a lowering of 1.2 per cent, less than the sum of those obtained for itscomponents, and he inferred that its dissociatfon was con.plete. He does not, however, state the concentration observed ; and the ex- tent to which the difference obtained depends upon concen- tration is apparent from the above table. These results must be regarded as comfirming the indications given by conduc- tivity of the existence of alum molecules in concentrated solution. II. Double Chloride of Zinc and Potassium. Riidorff' distinguished, as has already been mentioned, two classes of double chlorides. ( 1 ) those which suffer decompo- sition in water solution, and (2) those which remain undecom- posed. The first class are to be regarded as double salts of the same type as the alums. The second class he regards as true binary compounds. The following examples of the first class are given : 2KCI+CUCI, -f 2H,0 2NH,Cl + CuCl,+ 2H,0 2KCH-ZnCl,-+-H,0 KCH- MgCl, 4- 6H,0 2NaCl+CdCl,+ 3H,0 BaCl, + CdCl, + 4H,0. It seemed of interest in connection with the foregoing in- vestigation to study an example of this class by the con- ductivity method. For this purpose, the double chloride of potassium and zinc, 2KCl.ZnCl,+ H,0 was selected. Preparation, In preparing the double chloride of potassium and zinc, potassium chloride was used which had been twice crystal- lized, and was shown by the flame test to be free from sodium. The zinc chloride used was tested and found free I Compt. reud., vy, 9>4. (>884). » Ber. d. chem. Ges., ai. 3048. (1888). er cent. — may her values ob- other alums, oweringof 1.2 iscomponents, te. He does ; and the ex- upon concen- e results must :n by conduc- i concentrated nentiLdied, two uffer decompo- nain undecom- iouble salts of he regards as pies of the first ; foregoing in- ss by the con- ble chloride of elected. isium and zinc, I twice crystal- 3 be free from and found free 11,3048, (1888). 33 from impurities. The two chlorides were brought together in solution in the proportion necessary to form the double salt, and the solution evaporated to crystallization. The salt was separated as rapidly as possible from its mother-liquor on the filter-pump, and dried under a press. A specimen, previ- ously dried at 130", was immediately weighed off, and an- alyzed for zinc. The analysis gave : Percentage of zinc found 22.91 " calculated for 2KCl.ZnCl, . 22.94 The solution was standardized by precipitating zinc by so- dium carbonate and weighing as zinc oxide. The conductivity results are tabulated below : 2KCl.ZnCl, [285.58]. Gram- molecules per litre. Litres per Kram- molecule. 0.2 0.5 O.I I.O 0.5 3.0 0.25 4.0 0.05 20.0 0.025 40.0 0.005 200.0 0.0025 400.0 0.0005 2000.0 0.00025 4000.0 0.00005 20000.0 I. 11. «:.25°. f^v2f. Mean. 117. 2 • • • • 117. 2 176.0 • • • • 176.0 244.2 244.2 303.7 .... 303-7 397-4 397-2 397-3 426.2 423.8 425-3 468.4 469.0 468.7 483.2 484.2 483-7 504.0 504.6 504-3 510.8 512.3 5"-5 516.8 509.0 512.9 518.8 516.0 517-4 0.000025 40000.0 A comparison of the conductivity of the double salt with the conductivities of its constituents is given below. The values for zinc chloride and potassium chloride are interpo- lated from the tables of Kohlrausch' and reduced to 25°. The molecular conductivity of the zinc chloride is taken at the same volume as the double salt. That of potassium chloride is taken as twice the molecular conductivity of potassium chloride at one-half the volume of the double salt. 1 Wied. Ann. J6, i6i, (1885). ■ffr ii i L ii ff i i e ii i ,uwi ii r 1.*^ -' -^.- rj#j" Double Ch'iride of Zinc and Potassium. Utresper Uii-sper gram-mol*- K :i KHo'« cult v> ^ cvile orKCl. 2X/-25 .o(ZuCl,. 0.5 I.O 2.0 lO.O 20.0 1CX3.0 200.0 lOOO.O 2000.0 lOOOO.O 196. '• 211. a ?2t ,6 241.4 3; J.C 2''H.8 269 o 2754 277 6 27 JO O, I 2.0 4.0 20.0 40.0 20O.C' 4UO.O iooo.o 4000 o 2C00C).0 /.nCl,. )"t-25°. 78.4 120.0 140.3 179.2 186.7 213.6 220.1 232.0 234.8 240.2 240.7 The foregoing comparison leaves no room for doubt as to the existence of thi-i double salt as such in concentrated solu- tion, while at a dilution of between 1000 and 2000 liters it is entirely dibsociated into its constituents. Conclusion. Regarding the doable chloride of zinc a^id potassium as a type of the class of dissociated double haloids, we may draw the conclusion that the.se salts are present as such in their concentrated solutions As compared with the alums, they show a greater degree of stability towards water in concen- trated solutions, bet, i.'. dilute solutions, resemble them in becoming wholly dissociated into their component salts. 2C1K0OO.O 280.0 40000.0 Sum. 274.4 33«-8 361.9 420.6 436.7 478.4 489.1 507-4 512.4 519.2 520.7 iKCI.ZnCI, observed, 176.0 244.2 303.7 397-3 425.3 468.7 483.7 504.3 5" -5 512.9 517-4 Dtfler- enc"?. 98.4 87.6 58.2 23-3 II.4 9-7 5.4 3.1 0.9 6.3 3-3 I.ZnCI, •rved, 25°. ^6.0 33-7 37-3 25-3 68.7 83.7 04.3 "•5 12.9 17.4 r doubt as to intrated solu- 00 liters it is tassium as a we may draw such in their ; alums, they er in concen- nble them in nt salts. Difller- 98.4 87.6 58.2 23-3 11.4 9-7 54 31 0.9 6.3 3-3 I BIOGRAPHICAL. The author is a native of Plainfield, Nova Scotia, where he was born January 24, 1864. He graduated as Bachelor of Arts from Dalhousie College, Halifax, N. S., in 1886, and thereafter occupied a position as ateachor in the High School, New Glasgow, N. S., until 1892. In October, 1892, he en- tered the Johns Hopkins University, where he has since been pursuing a post-graduate course in Chemistry, Physics, and Mathematics. He was awarded a University scholarship in Chemistry in January, 1895, and was appointed a Fellow in June of the same year. A CONTRIBUTION ro TJKB HtiroY Of DOUBLE 8ftl,T8 IN DISSERTATION ■' v^' v.- „ ;, '. ■■ V- y.- 1 ■; - ■ ■■■.-■■ > -.■ . ■ ^—»X- E:BEN,-K2:ER' MJ>kGlKjkX 11196 4 ■'■. !■• ■ ■ ■ ■ .1 ' ■ tit-.'S'.y ft '■-' si"'.' '»"