88 On definite Proportions. 
les unes aux autres.” This is admirably describing them. Those 
in the water-lily are round as in fig. 2. in the Journal just men- 
tioned, and the square sort are found in the vail of the same plant, 
The water-lily is in the present print. (See fig. 0, GG.) 
To Messrs. Nicholson and Tilloch. 
——— 
XX. An Aitempt to determine the definite and simple Pro- 
ortions, in which the constituent Parts of wnorganic Sub- 
stances are.united with each other. By Jacoz BERZELIUS, 
Professor of Medicine and Pharmacy, and M.R.A. Stock- 
holm. 
[Continued from p. 54.] 
2. Alumina. 
Js order to be able to determine the quantity of oxygen con- 
tained in alumina, I dissolved some of this earth, which had 
been precipitated from alum by caustic ammonia, in sulphuric 
acid; and when the acid would take up no more of it, I filtered 
the solution, concentrated it by evaporation, and precipitated the 
neutral salt from it by the addition of aleohol. The precipitate 
was well washed with alcohol, in order to separate any uncom- 
bined sulphuric acid which might accidentally be present. The 
salt thus prepared had completely the taste of alum, but the taste 
was much stronger. In order to drive off the water, I heated it 
in a platina crucible over a spirit lamp, and weighed the crucible 
from time to time ; when it lost no more of its weight, I con- 
sidered the salt as free from water. It melted, swelled up, and 
exhibited the appearance of alum; at last I was obliged to force 
it down into the crucible, in order that the heat might be able 
to penetrate it uniformly. The dry salt appeared not to be so- 
luble in water, and in this respect resembled the dry sulphate of 
the protoxide of iron, the dry sulphate of magnesia, burnt alum, 
and other salts. But with the assistance of heat it was by de- 
grees completely dissolved. . 
I ignited 10 grammes of this dry sulphate of alumina in a 
platina crucible, as long as they lost any weight from the extri- 
cation of sulphurous acid ; they left behind 2°9934 gr. of a loose 
white andlight alumina. This salt, the dry sulphate of alumina, 
must therefore consist of : 
Sulphuric acid .. 70:066 100-000 
PAMIMIMNA ss 4 6 y0 255... 29 GOOF 42°722 
And if 42:722 parts of alumina contain 19°96 of oxygen, 100 
parts must contain 46-726. 
In order to prepare the hydrate of alumina, | first attempted 
4 to 
