NOV. 22, I915 
Improved Respiration Calorimeter 
339 
same as that of the air leaving the chamber. This rheostat is mounted 
on the end of the observer's table beside those for controlling the temper¬ 
ature of the zinc wall. 
allowance: for conditions affecting the heat of the chamber 
Any passage of heat into or out of the chamber through the walls or 
in the ventilating air current being prevented, the sum of the quantity 
of latent heat in the water vapor of the outgoing air and that of sensible 
heat removed by the water circulating in the heat absorber would equal 
that actually produced in the chamber if there were no change in the 
temperature of the walls or in that of any objects confined within them. 
Under ordinary conditions, however carefully the rate of abstraction of 
heat from the chamber has been regulated to accord with that of pro¬ 
duction, temperature changes can not be absolutely avoided, so they 
must be measured and allowance made for them. 
Change in Temperature of the Metal Walls 
If the temperature of the copper wall is lower at the end of a given 
period than it was at the beginning, and the temperature of the zinc wall 
has been kept identical with that of the copper wall throughout the period, 
a certain amount of heat has been imparted to the air of the chamber by 
the copper wall during the period; or, conversely, if the copper wall is 
warmer at the end of the period, some heat has been absorbed from the 
air by the wall. To ascertain how much allowance must be made for the 
heat involved in such changes, it is necessary to determine the tempera¬ 
ture of the copper wall at the beginning and the end of the period and to 
know how much heat is necessary to raise the temperature of the calo¬ 
rimeter a given amount—i. e., its hydrothermal equivalent. 
The temperature of the copper wall is determined by means of an 
electric-resistance thermometer somewhat like that described on page 
313 for determining the temperature of the air. In this thermometer, 
however, each of the six coils of nickel resistance wire is wound on a thin 
fiber strip about 12 cm. long and 1 cm. wideband is covered with a thin 
layer of silk and lacquered, the completed bulb being about 1.5 mm. thick. 
A strip of brass, slid into guides soldered to the copper wall, presses each 
coil firmly against the wall so that there is close thermal contact with 
metal on each side of the coil; hence, changes in the temperature of the 
wall affect the resistance wire very quickly. These six coils, joined in 
series by well-insulated No. 16 copper wire, are distributed on the side 
walls and ceiling in such manner as to show the average temperature of 
the total mass of copper. The terminals of the series of coils are con¬ 
nected with the special switch, mentioned on page 334, and through that 
with the temperature indicator (Wheatstone bridge) on the observer’s 
table. The bridge and galvanometer are sensitive to resistance changes 
