154 - Onthe Affinities of Bodies for Light ; 
The barometer connected with the ‘interior of the prism, 
compared with that in the open air, always gave the exact 
aecount of the density of the contained fluid compared with: 
that of the ambient air. The prism was filled at pleasure 
with any given gas, by previously exhausting it of ar and 
then connecting it with a reservoir filled with the gas, which 
was ascertained to he the purest of its kind by several pre- 
cautions. This apparatus was constructed by M. Fortin, 
to whom the authors pay the compliment of having fulfilled 
their expectations in point of convenience and precision i 
his execution of that apparatus. 
The specific gravity of the gases empfoyed was arm esser- 
tial requisite im this kind of experiments. The authors as- 
certained it by the commron process of weighing a globe of 
glass previously emptied, and afterwards filled with the gas: 
all the corrections were made for the state of the barometer, 
the thermoineter, and even of the hygrometer; because, the 
vapour of the water being lighter than the air, as 16 to 14, 
at equal pressure, it was nceessary to have regard to this cir- 
eumstance, and it was done according to a formula which 
™M. de la Place deduced from the experiments of De Saussure 
and Dalton. Al! the weights were reduced to what they 
had been inthe vacuum at the freezing pomt, and under the 
coustant pressure of 0-76 of a metre. Regard was even paid 
to the dilatation of the glass, supposed equal to 00600262716 
of its volume for each degree of the centigrade thermometer 3 
a result discovered by Messrs. Lavoisier and De la Place in 
a work upon the dilatation of solids, which, unfortunately, 
has not been published. With all these precautions, the 
results obtained on different days, and in very different states 
of the air, scarcely deviated a few milligrammes from each 
other, when they were reduced to the same temperature and 
the same pressure. ° 
In order exactly to compare the weights of the gases with 
that of water, the authors gauged very exactly the globe with 
pure water at the temperature of 4° (39°2° Fahr.), the point 
at which the greatest condensation takes place, and at which 
the absolute weight of water was very preciscly determined 
by experiments made in I'rance for the determination of the 
gramme, 
