I0O Francis B. Sumner 
same air, being heated to a temperature of 20° or 30° C., its hu- 
midity (degree of saturation) fell to 50 percent or even considerably 
less. ‘There is a popular fallacy to the effect that moist air may 
be dried by heating, after the manner of one’s damp clothes. ‘The 
notion contains this element of truth, of course—namely that heat- 
ing, except when the same degree of saturation is maintained by 
evaporation, does increase the capacity of the air for taking up 
moisture from other objects. Accordingly, in the present experi- 
ments, the sawdust upon the floor of the cages and the cotton 
waste used for bedding were much damper in the cold room than 
in the warm one. Indeed there was commonly no perceptible 
dampness at all in the latter, while in the former it became dis- 
tinctly moist if the changing were long neglected. ‘The relative 
humidity was determined at rather irregular intervals by the use 
of an ordinary psychrometer or wet-and-dry-bulb thermometer. 
The percentage values were obtained from the scale attached to 
the “hygrophant,” manufactured by J. S. F. Huddleston of Bos- 
ton, though the instrument actually used was a similar one of 
another make. ‘These percentages are probably to be regarded 
as rather rough approximations. ‘They are, however, believed to 
suffice for present purposes. [he temperature, in each room, 
was recorded continually by means of a thermograph, and thus the 
daily and hourly fluctuations of temperature could be followed.’ 
Since the conditions and methods of treatment differed con- 
siderably in the different series of experiments, a further account 
of these is deferred to the separate discussions of the latter. No 
account of the feeding nor of the general care of the mice is regarded 
as necessary here.® It is sufficient to state that, except for the 
differences in temperature and humidity above mentioned, it was 
my endeavor to maintain all conditions as similar as possible for 
the two contrasted lots of animals. 
™ The laborious task of computing the mean temperatures, etc., from the tracings upon the thermo- 
graph sheets has been mainly performed by my wife and my mother. 
®J must acknowledge my indebtedness to Prof. W. E. Castle and Prof. T. H. Morgan for valuable 
suggestions relative to the care of the mice. The stock was all obtained from Miss Abbie E. C. 
Lathrop, the well-known animal breeder, of Granby, Mass. 
