440 PROCEEDINGS OP THE AMERICAN ACADEMY 



cannot lead far away from the truth, since the difference between the 

 maximum aiul minimum corrections observed amounts for the quantity 

 of material employed in these experiments to but 0.0002 grm. De- 

 terminations of standard, original and corrected, are given in (28), 

 (29), (30), and (31). 



A portion of the solution, the standard of which is set by (28) and 

 (29), was diluted to ten times its volume, and 50 cm.^ of the diluted 

 solution were used in each of the experiments immediately following. 

 These were made to determine whether the precipitate was' definite 

 in composition, and might be weighed as such. Precipitation was 

 effected by a solution of microcosmic salt acidified with formic acid. 

 The precipitates of (32) and (33) were collected on paper, that of 

 (34) on asbestos, and all were ignited and weighed directly. 



The jihosphate obtained by Merz * by precipitating in presence of 

 hydrochloric acid containing two molecules of TiO^ to one of Pg^s' 

 requires 0.032G grm. of PgO^ to 0.0377 grm. of fiOa. It will be 

 noticed that in two determinations the PgO^ is in a considerable ex- 

 cess of the amount demanded by this proportion, and in one, in defi- 

 ciency. The idea of weighing the phosphoric anhydride and titanic 

 oxide togetiier was therefore abandoned. 



In experiments (35), (36), (37), the titanium was present alone; 

 in (38), (39), 2 grm. of alum were added to tlie solution before pre- 

 cipitating by means of the mixture of microcosmic salt and formic 

 acid, 5 grm. of the former being employed to 8 grm. of the latter. 

 The precipitates of all, excepting (35), were carefully washed, ignited, 

 fu.sed with sodic carbonate, and the product of fusion was dissolved 

 in water, the insoluble titanate separated from the soluble phosphate 



* Jour, prakt. Clicm., .\cix. 157. 



