330 FRANCIS B. SUMNER 



3. Effects of considerable changes in the air temperature upon both 

 warm-room and cold-room mice 



Before endeavoring to compare the mean body temperatures 

 of the warm-room and cold-room mice, under conditions which 

 were approximately normal for the two lots, it may be well to 

 consider some results illustrating the effects upon these animals 

 of sudden and considerable changes in the temperature of the 

 air. 



In table 8, we observe the effects of changing 10 male mice (5 

 young and 5 old) to a much colder atmosphere than that to which 

 they were accustomed. Through causes beyond my control, they 

 had already experienced a large and fairly rapid fall of tempera- 

 ture during the preceding night. But at the time of the first 

 test, the room had been tolerably warm for an hour or more, 

 though considerably below its average temperature. After the 

 first test, the mice were transferred to the cold room, then at a 

 temperature of about — 2.5°C. The temperatures of the mice 

 were taken about five hours later (air at 0°), and again on the 

 following morning, all but two of the animals having been kept 

 for twenty-four hours in the cold room. By the time of the last 

 test, the atmosphere of this room had risen to 6.5°. 



Considering first the figures for the adult mice,^^ we have at 

 the outset a mean temperature of 36.30°, but it must be noted 

 that the variability is very great. Twenty-four hours later, the 

 average for these five was 35.82°, all but one of the animals having 

 undergone a reduction of temperature during this interval. The 

 individual changes during this period, as revealed by the second 

 and third tests, are highly variable. 



Passing to the five nearly adult males (three and one-half 

 months old), it is to be recorded that in two cases subnormal 

 temperatures resulted, one mouse reaching the surprising tem- 

 perature of about 12.5°. These two individuals evidently lost, 

 for the time being, their power of heat regulation, and will there- 

 fore be left out of consideration in the present computations. 

 Considering the other three, we find a mean temperature, at the 



*^ 'Thirty-second' figures. 



