Quantity of Gases absorbed by Water , See. 31 
bent at a right angle, and graduated, from a given point, into 
hundredth parts of a cubical inch. It is attached to the copper 
pipe, by a tube of Indian rubber D, over which is a covering of 
leather, forming a joint, which admits of the vessel A being 
briskly agitated. When the apparatus is used, it is first filled 
with quicksilver ; a transfer bottle of elastic gum, furnished with 
a cock, and containing water of a known temperature, is screwed 
on ; and a communication is opened, through the cocks, between 
the bottle and the glass vessel. The lower cock b is then opened, 
through which the mercury runs out, while its place is supplied 
by a quantity of water from above, measurable by the scale on 
A. This transfer is removed, and another containing gas being 
substituted, a measured quantity of gas is admitted in a similar 
manner. Strong agitation is now applied, by means of the joint 
D; and mercury is poured into the tube B, to supply the 
descent occasioned by the absorption in A; its level being 
exactly preserved in both legs of the syphon, both at the com- 
mencement and close of the experiment. The quantity of mer- 
cury required for this purpose, indicates precisely the amount 
of the gas absorbed. 
The only advantage of this apparatus over a cylindrical jar, 
inverted in the usual way over mercury, is, that by means of 
the tube B, very minute degrees of absorption may be measured, 
which would scarcely be perceived in a wide vessel. 
For the more absorbable gases, I found this instrument to 
answer perfectly well; but, for ascertaining the solubility of 
those which are taken up by water in only small proportion, I 
preferred one of different construction. It consisted simply of a 
glass vessel, of the capacity of 57J cubical inches, and shaped 
as in Fig. 2. At a was cemented a cock, provided with a screw : 
