1912] PEIRCE—RESPIRATION 95 
tant in connection with respiration experiments in which heat pro- 
duction is being studied; for evidently the flasks, as well as the 
organisms within them, will lose heat most rapidly through the air 
when the air is decidedly cooler than they are, whether because the 
air is actually cold, or because the heat liberated by the organisms 
in the flask accumulates. The total heat liberated, therefore, must 
be the sum of the heat lost through radiation, etc., plus that 
retained by the flask. By placing the flasks in a constant tempera- 
ture, the rate, and therefore the amount, of heat-loss can be deter- 
mined for each flask and each degree of difference in temperature. 
Another factor to be reckoned before we can attempt even 
approximately to determine heat yields is the heat equivalent of 
each combination of apparatus. This work has revealed to me, as 
I never suspected before, the limitations of the ordinary thermome- 
ter as an instrument of precision. In the following determinations, 
as will be seen below, I have used the same thermometers and the 
same cotton plugs in the same flasks throughout the series of experi- 
ments, thus determining the heat equivalents of each calorimeter, . 
consisting of thermometer, Dewar flask, and plug. Indeed, I kept | 
the flasks in the same places on the wooden filter stands and even 
on the same spot on my table. These latter precautions are, 
however, strainings at gnats, for the camel of unavoidable error to 
be swallowed is very large, owing to the small size of the flasks. \ 
The method followed to ascertain the heat equivalent of each 
set of apparatus was fundamentally as follows, although I modified - 
one detail or another in the series of determinations which I 
attempted. Into a flask, the temperature within which is known 
and recorded, 200 cc. (100 cc. if the flask already contains 100 cc.) 
of warm distilled water, at a known temperature, are quickly 
poured; the temperature within the flask is again recorded, as soon 
as it has again become fairly stable. The fall in the temperature 
of the water poured into the flask indicates that the flask and the 
inclosed air, thermometer, and cotton plug, have taken up warmth 
from the water. If all the water could be delivered into the flask 
at the temperature recorded, the fall in temperature would indicate 
only the amount of heat taken from the water by the apparatus. 
But this ideal condition I have found practically unattainable. 
