temperatures of 70° if the distribution of air and blood were
effectually secured, and I also found a proper standard of oxidation
from a proper temperature. Afterward I proceeded to test for
combination at lower temperatures, and discovered a gradually
decreasing scale until I arrived at 40° Fahr., when efficient
combination ceased. Of course, my method was a very crude imitation of
nature, but it was sufficient to show this fair and reliable result,
that the oxidation of blood decreases as the temperature of the oxygen
decreases.

From this point I went to animal life itself. I exposed animals to
pure cold oxygen and to cold atmospheric air, and compared the results
with other experiments in which animals of similar weight were exposed
to warm air and warm oxygen. The facts gleaned were most important,
for they proved conclusively that the products of combustion, that is
to say, the products resulting from the union of oxygen and carbon,
were reduced in proportion as the temperature of the oxygen was
reduced. In the course of this inquiry another singular and
instructive fact was elicited. It has been long known that at ordinary
temperature, say 60°, pure neutral oxygen does not support animal life
so well as oxygen that is diluted with nitrogen. In the nitrogen the
molecules of oxygen are more freely distributed under the influence of
motion, that is the meaning of the observed fact. What, then, would be
the respective influence of low and high temperatures on the
respiration of pure oxygen? To settle this question, animals of the
same size and weight were placed in equal measures of oxygen gas and
common air at a temperature of 30° Fahr., and with the inevitable
result that the animal in the pure oxygen ceased to respire one-third
sooner than did the animal in common air. Carrying the inquiry
further, I found that if the oxygen gas were warmed to 50° Fahr., the