459 
vegetable and animal substances. 
then to put it up in a phial, and by igniting one hundred 
grains of it in a proper glass tube, sealed at one end, and 
loosely closed with a glass plug at the other, to determine 
the proportion of moisture which it contains. This, then, in- 
dicates the constant quantity to be deducted from the loss of 
weight which the peroxide suffers in the course of the experi- 
ment. The mortar should be perfectly dry, but not warm. 
Experimenters have been at great pains to bring the 
various organic objects of research to a state of thorough 
desiccation before mixing them with the peroxide of copper ; 
but this practice introduces a similar fallacy to that above 
described. We ought, therefore, after having made them as 
dry as possible by the joint agencies of heat, and an absorbent 
surface of sulphuric acid in vacuo, to expose them to the air 
till they also come into hygrometric repose, noting the quan- 
tity of moisture which they imbibe, that it may be afterwards 
allowed for. The plan which I adopt for the purpose of 
desiccation seems to answer very well. Having put the pul- 
verulent animal or vegetable matter into short phials, fur- 
nished with ground glass stoppers, I place the open phials 
in a large quantity of sand heated to 21 2° F. in a porcelain 
capsule, and set this over a surface of sulphuric acid in an 
exhausted receiver. After an hour or more the receiver is 
removed, and the phials instantly stopped. The loss of 
weight shows the total moisture which each of them has 
parted with ; while the subsequent increase of their weight, 
after leaving them unstopped for some time in the open air, 
indicates the amount of hygrometric absorption. This is 
consequently the quantity to be deducted in calculating ex- 
perimental results. 
