ugh the construction of the apparatus is such
that this presents some little difficulty. It is necessary, for
example, to accurately measure about 16 cubic centimeters of pure
nitrogen, pass it into the potassium pyrogallate pipette, and then
(having taken a definite sample of oxygen) gradually absorb the oxygen
in the potassium pyrogallate and measure subsequently the accumulated
nitrogen. The analysis is tedious and not particularly satisfactory.
Having checked the manufacturer's analysis of a number of cylinders of
oxygen and invariably found them to agree with our results, we are at
present using the manufacturer's guaranteed analysis. If there was a
very considerable error in the gas analysis, amounting even to 1 per
cent, the results during short experiments would hardly be affected.

ADVANTAGE OF A CONSTANT-TEMPERATURE ROOM AND TEMPERATURE CONTROL.

A careful inspection of the elaborate method of calculation required for
use with the calorimeter formerly at Wesleyan University shows that a
large proportion of it can be eliminated owing to the fact that we are
here able to work in a room of constant temperature. It has been pointed
out that the fluctuations in the temperature of the gas-meter affect not
only the volume of the gas passing through the meter, but likewise the
tension of aqueous vapor. The corrections formerly made for temperature
on the barometer are now unnecessary; finally (and perhaps still more
important) it is no longer necessary to subdivide the volume of the
system into portions of air existing under different temperatures,
depending upon whether they were in the upper or lower part of the
laboratory. In other words, the temperature of the whole ventilating
circuit and chamber, with the single exception of the air above the acid
in the first sulphuric-acid absorber, may be said to be constant. During
rest experiments this assumption can be made without introducing any
material error, but during work experiments it is highly probable that
some consideration must be given to the possibility of the development
of a considerable temperature rise in the air of the potash-lime
absorbers, due to the reaction between the carbon dioxide and the solid
absorbent. It is thus apparent that the constant-temperature conditions
maintained in the calorimeter laboratory not only facilitate
calorimetric measurements, but also simplify considerably the elaborate
calculations of the respiratory exchange formerly