Hare’s Eudiometers, &c. 315 
be kept open, and under water, so as to permit the latter to 
take place of that portion of the gas which disappears.— 
_ Whatever remains unabsorbed, is expelled from the glass 
spheroid, as in the case of the tube when used alone; and 
the divisions on the rod remaining without, will shew how 
much the fluid has taken up. 
When atmospheric air, or oxygen gas is to be analyzed 
by nitrous gas, the glass spheroid is filled with water, and in- 
verted with its orifice closed over the well of the pneumatic 
cistern. It should be supported by a wire stand, so as to 
leave the neck unobstructed. Any number of measures of 
nitrous gas, and of oxygen gas, or atmospheric air, may 
then be drawn into the measurer, and expelled into the 
spheroid successively, and the absorption estimated as al- 
ready explained. When the residuum is too great to be ex- 
pelled by returning the whole of the rod into the tube, by 
depressing the orifice of the spheroid just under the surface 
of water, the wire may be again gently retracted, water 
taking its place; and the movement may thus be alter- 
nated, till the whole of the remaining gas is excluded. 
In order to apply this principle to Volta’s process of as- 
certaining by explosion the quantity of hydrogen or oxy- 
gen gas present, in a mixture, the gas measurer is made as 
much stronger, as eudiometers are usually, when intended 
to be so used. It isin like manner drilled so as to receive 
wires for passing the electric spark. The instrument being 
charged with the gases successively in any required propor- 
tion, closed by the screw, and an explosion accomplished ; ta 
fill any consequent vacuity, the orifice is to be opened just be- 
low the surface of water or mercury. The quantity destroy- 
ed by the combustion is then ascertained by the sliding wire. 
This experiment is more accurately performed by means 
of mercury than water. From this fluid, concussion, or 
even the partial vacuum produced by the gaseous matter, 
may extricate air, and thus vitiate results. There ought 
always to be a considerable excess of gas not liable to be 
acted on. ‘The activity of the inflammation is lessened, 
and the unconsumed air breaks the shock. 
T have found the galvanic ignition produced by a small 
calorimotor preferable to the electric spark. Suppose a 
piece of iron wire to be filed down in the middle for about 
one half of an inch to about one third of the original diam- 
