210 PROCEEDINGS OF THE AMERICAN ACADEMY 



flask makes them respectively 10.90 and 33.64. Adding to these 

 tiie determinatious of K„0 and H^O, made by separate exjaeriments, 

 makes the complete analysis of the alum used as follows : 



100.12 100.00 



It is evident that the difference between the results of A and of 

 corresponding results of subsequent experiments in which precipitation 

 does occur, gives the amount of AUOg and SO3 in the precipitate. 

 The final results of each experiment are reckoned as parts of ALOg 

 in 100 parts of anhydrous precipitate, and as parts of AlgOg precipi- 

 tated out of 100 parts of AlgO, present in the total system. The fol- 

 lowing is an example of a calculation in full : — 



AloOj. SO3. 



A = .10766 grm. .33231 



D = .0689 .31134 



Precipitate = .03876 .02097 



» , -, • • .03876X100 nA onr^ Ai/^ 



Anhydrous precipitate = Q.jg^g _|_ qo cTq? = 64.89% Al^Og. 

 AlA of precipitate = '^^^H^^^^^^^ = 36-00% of total AlO.. 



The data of this series of experiments are recorded in Table VI. 

 The quantity of KOH varies, as will be seen, from 1.5 to 8 molecules. 



Before discussing the results, I describe another series made to 

 reverse, as it were, the conditions of the former. 



Fifty c.c. of alum solution were taken as before, and to this ten 

 molecules KOH added, and the whole made up to a little less than 

 125 c.c. Sulphuric acid solution, containing quantities of H.,SO^ pro- 

 portional to 1.5, 2, 2.5, and 4 molecules, was added, and the mixture 

 brought accurately to 250 c.c. (7^20°), thoroughly shaken, filtered, 

 and the filtrate analyzed as usual. The only variation was in 

 Exp. T, when 5 mol. HCl were used in place of 2.5 HjSO^, and this 

 will be more particularly discussed in a subsequent publication. The 

 data of these experiments are recorded in Table VII. 



