OF ARTS AND SCIENCES. 213 



is more complex than in Q, consisting in the precipitate acquiring 

 both Al,,03 and SO3. In other words, the equilibrium has moved the 

 crossed point Ji up to the point i? in VI., and the crossed point H 

 down to E in VIII. 



The results alread}' described indicate unmistakably the existence 

 of the equilibrium ; but, to vary the evidence still more, Exp. S', S", 

 and S'" were made, as follows : — 



Aluminum was dissolved in KOH solution and the KOH exactly 

 neutralized by HCl, independent experiments having proved that 

 the precipitate thus obtained is fi-ee of acid. The volume of this 

 mixture having the freshly precipitated hydrate in suspension was 

 brought up to 250 c.c. Fifty c.c. of alum solution (= 5 grm. alum) 

 were then added, the whole made up to 500 c.c. and thoroughly 

 mixed. Before filtering, portions of 50 c.c. each were withdrawn 

 with a pipette, care being taken to insure the uniform distribution of 

 the suspended precipitate by constant stirring. In these portions the 

 precipitate was redissolved by HCl and the total Al^Og of the system 

 determined. The remainder of the mixture was filtered as usual, and 

 portions of 50 c.c. of its filtrate taken for the usual determinations of 

 SO3 and Al^Og. The diiferences between the total AI.2O3 and SO3 

 and the AI2O3 and SO3 found in the filtrate give the composition of 

 the precipitate. In S' the amount of AloOn added as precipitate was 

 1.12 times that added as alum, making a proportion of 2.12 mol. ALO3 

 to 3 mol. SO3 in 500 c.c. of volume ; equivalent to 1 AI2O3 and 1.4 SO3 

 in approximately 250 c.c. This so far as relation between ALOg 

 and SO3 is concerned, should be equivalent to 1 alum -|- 3.2 KOH. 

 Using therefore the analytical results of S', Table VIII., and making 

 X = 3.2, we obtain the dotted points S', Plate II., approximating 

 closely to the values otherwise obtained ; whereas had the precipitate 

 remained unchanged it would have been lOO*}^ AI2O3 and contained 

 52.S2f/o of the total AI2O3. Tliese values are represented by the 

 crossed points >S''. The change consists mainly in acquiring SO3. 



Exp. S" is similar, the only variation being in the quantity of 

 precipitated AUO^ added. The results arc shown in the points S" 

 (x = 2.5) ; whereas tlie crossed points S" indicate what would have 

 been the results had the precipitate remained unchanged. Here, too, 

 the change is in acquiring 80.,. 



Still another experiment, S'", was made, varying again the quan- 

 tity of AI2O3 added. The actual results are shown in the points S'" 

 (a: = 2), and the initial precipitate in the crossed points, S'". Here it 

 is apparent that the change consists in gaining 8O3 and losing AI2O3. 



