152 
Proceedings of the Royal Society 
mother liquor till the next day, and then filtered and dried at 100° 
C. We called these respectively A, B, and C. 
On titration they gave the following results : — 
B was selected for further analysis. On ignition it lost 41*35 
percent. The percentage of water found was 36 '05, and that of 
C0 2 , 3'5. The percentage of alumina 58*65, The percentage 
composition was therefore 
On deducting the water and C0 2 the percentage of oxygen 
equals 2*69 in the anhydrous oxide. 
The presence of carbonic acid in this preparation is probably due 
to absorption from the atmosphere during the preparation. The 
water and carbonic acid were estimated directly; the oxygen by 
oxidising a standard solution of ferrous sulphate, and titrating the 
excess of ferrous salt by means of a standard permanganate of 
potash solution ; the oxide by igniting a portion of the substance, 
and estimating the residue. 
Although the above analyses show that the substances we have 
prepared are not pure, yet they hold out the hope that, by modi- 
fying our experiments, we may yet succeed in preparing the pure 
peroxides in a state fit for analysis and for the determination of their 
formulae. 
5. On Papers, by MM. Haas, Cleve, and Lecoq de Bois- 
baudran, on the Production of Peroxides by means of 
Peroxide of Hydrogen. By J. Gibson, Ph.D., and 
E. M. Morrison, D.Sc. 
The foregoing paper was read on 5th July 1880. We postponed, 
however, its publication, feeling that however interesting the forma- 
tion of the bodies therein described might be, much careful quanti- 
A, 1*04 per cent, of oxygen. 
B, 1*6 
C, 1 less than on percentage of O. 
A1 2 0 3 . . , 
H 2 0 . 
co 2 . . , 
O (by titration) 
58*65 
36*05 
3*5 
1*6 
99*70 
