18 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



pointed out that this occlusion is probably due to the distribution of a 

 definite molecular species between the solvent and the precipitate at the 

 moment of formation of the latter. Various data concerning different 

 precipitates seem to support this conclusion so strongly that it may "be 

 used as a working hypothesis.* 



Thus the composition of the occluded matter may afford a clue as to 

 the nature of some of the molecular species present in the solution. 

 The only difficulty in the interpretation of the results is due to the possi- 

 bility that more than one molecular species may thus distribute itself 

 between the solution and the precipitate. 



In the cases cited above, which furnished the basis for these conclu- 

 sions, it was shown f that as a rule electrically neutral substances rather 

 than ionized ones seemed to be thus distributed. This is indeed a rea- 

 sonable conclusion ; for the ionizing power of cool solids such as baric 

 sulphate is probably much less than that of water. 



In view of these inferences, it seemed possible that the careful analysis 

 of baric sulphate precipitated in the presence of much green chromic 

 salt might furnish valuable evidence as to the nature of the complex 

 substances present. Accordingly precipitates were made under varying 

 circumstances, and analyzed with great care. The first experiments 

 showed that green chromic salt is occluded by baric sulphate to an 

 extent fully as great as ferric iron, while the violet salt is scarcely oc- 

 cluded at all. Since the baric sulphate, after all the sulphuric acid had 

 been precipitated from the solution, weighed less than the amount corre- 

 sponding to the sulphuric acid known to be present, it was clear that the 

 chromium must be present in the precipitate as sulphate. The case 

 seemed, in short, to be precisely similar to that of iron, carefully studied 

 by Jannasch and Richards in the paper already cited. 



Our present object was to carry further the method of study followed 

 in that paper, and to determine more precisely the nature of the occluded 

 compound. A first step in this direction was to discover whether or not 

 green chromic chloride might be occluded as well as sulphate ; for the 

 occlusion of chromic chloride would seriously interfere with the possibility * 

 of satisfactory interpretation. 



For this purpose it was not enough to determine the chlorine in the 

 precipitated baric sulphate, since baric chloride is known to be occluded 



* See for example Richards, McCaffrey, and Bisbee, These Proceedings, 36, 

 377 (1901). 



1 Richards, loc cit. 



