262 
attention was first directed to ascertain the quantity 
and composition of the air that exists naturally in 
river water. For this purpose, they filled glass-bal- 
loons with given quantities of water, taken from the 
river Seine, and expelled the air from it by submitting 
it to ebullition. The air that came over was received 
in vessels filled with mercury, or with distilled water? 
recently boiled, that no foreign air might mix with 
that obtained from the water in the balloon. From 
the results of ten experiments, conducted in this man- 
ner, they found, that the water of the Seine contain- 
ed rather less than iV of its volume of air. This air 
they farther found to be composed of about -rVg- oxy- 
gen, with from six to eleven per cent, carbonic acid, 
and the remainder was nitrogen gas. The composi- 
tion of the air, contained in river- water, they state to 
be as constant in its proportions as that of the atmo- 
sphere ; for in experiments continued many months, 
in times of drought and during the melting of snow, 
the proportion of oxygen in the water of the Seine 
never varied more than from 0.309 to 0.314 *. 
This estimation of the quantity and composition of 
the air, contained in river-water, does not differ so 
much as might have been expected from that afford- 
ed by the water of springs ; for, according to Mi- 
Da! ton, air expelled from common spring water, after 
losing from five to ten per cent, carbonic acid, con- 
sists oiSSfier cent, of oxygen and 62 of nitrogen gasf. 
552. Having thus determined the quantity and 
kind of air contained in a given volume of river-w?.-. 
* Mem. de d'Arcueil, torn. ii. p. 3(j7- et seq. 
f Chem. Phil, part ii. p. 272. 
