346 FRANCIS B. SUMNER 



the fact that the graduation of this one extended to a lower point 

 (20°C.) than that of either of the others. 



In a few cases, among the older of these young mice, the rectal 

 temperature was taken (the mice were invariably killed in the 

 end, however); in all other cases, the following procedure was 

 employed. 



The mouse was cut into very deeply, in the throat or chest 

 region. Such an incision necessarily severed the large blood- 

 vessels and soon led to the death of the animal. After the cut 

 was made, the thermometer was inserted with all speed far into 

 the abdominal viscera. The mouse was held, in nearly all cases, 

 by a large piece of cotton-batting, and care was taken not to 

 touch or bring the hand very near the body. The cotton was 

 left loose, however, so as not to wrap up the young animal. 



The mercury always rose abruptly to a point not far from the 

 first stopping-place, then more slowly for a few tenths of a degree, 

 when it became nearly or quite stationary. In some cases, it 

 now remained almost perfectl}^ motionless for a minute or more, 

 in others, it began to rise again, sometimes reaching a point 1° or 

 2° higher. In others still, the final decline commenced much 

 more quickly. 



The results of my temperature determinations upon young 

 mice are recorded in tables 13 to 20 where the figures given are 

 the maximum reading for the thermometer and the first tempera- 

 ture recorded, the latter being that attained at the end of the 

 first sudden rise of the mercury, following insertion. It was 

 probably reached within 10 or 15 seconds. 



That temperature readings obtained in the way described are 

 a fair index of the body temperatures of the living animals is 

 rendered very probable by a number of cases in which the rectal 

 temperature was taken immediately before killing. The two were 

 found to agree within close limits (tables 13 and 16). 



In the course of my observations, I frequently recorded the 

 temperature changes for several minutes, at half-minute intervals. 

 These later readings, of course, throw no light upon the normal 

 temperature of the body, but are of interest in showing tempera- 

 ture conditions after somatic death. 



