DISCUSSION OF RESULTS AND GENERAL CONCLUSIONS. 89 



in the rectum are accompanied by similar fluctuations in different parts of the 

 body. For purposes of computing the heat production, it is most necessary 

 to know if an alteration in temperature of 0.1 degree in any one part of the 

 body shows a like alteration in the temperature of the body as a whole, the 

 absolute temperatures not being so significant. The investigation referred to 

 has shown that when the rectal temperature falls 0.1 degree, there is a corre- 

 sponding fall in all other parts of the body ; hence it is logical to assume that a 

 change of temperature measured in the rectum may be ascribed to that of the 

 whole body. In computing the heat production, therefore, correction can be 

 made for the heat lost from the body when there is a lowering in the tempera- 

 ture by multiplying the specific heat of the body by its weight and by the 

 loss in temperature as measured in the rectum. 



The importance, therefore, of this table is that it establishes the fact that 

 the temperatures of diabetics as a rule range well within what are commonly 

 supposed to be the normal limits. We wish, however, to point out here the 

 difficulties experienced by Benedict and Slack in securing adequate and proper 

 records of the body-temperature by means of the mercurial thermometer in the 

 mouth. As a means of indicating the presence or absence of a high fever this 

 method is certainly rational, but for all physiological purposes and especially 

 for the measurement of small differences in temperature, observations taken in 

 the mouth are wholly unsatisfactory and no reliance should be placed upon 

 them. On the other hand, rectal temperatures should be taken deep in the 

 rectum, the thermometer being inserted at least 5 cm. if the true internal 

 temperature of the body is to be recorded. 



BODY-WEIGHT. 



Perhaps no one gross observation made during the course of diabetes melli- 

 tus is of greater significance and causes greater alarm, both to patient and 

 physician, than the persistent loss in body-weight. On the other hand, slight 

 changes in body-weight which may accompany dietetic alterations or the inges- 

 tion of sodium bicarbonate are looked upon as material gains and of diagnostic 

 value, and are thus liable to be misunderstood by the patient. To interpret 

 intelligently these changes it is necessary both for the physician and for the 

 patient to realize the factors affecting the body-weight of normal as well as 

 pathological cases. 



Few realize that the normal individual is continually undergoing changes 

 in body-weight throughout the 24 hours. The extent and rapidity of these 

 fluctuations may best be shown by means of a chart obtained from observations 

 made every hour for 24 hours. (See fig. 1.) The subject was weighed on a 

 sensitive platform balance ; the total weight of food eaten at different times 

 was recorded, also the total weight of drinking-water when it was taken, and 

 the weight of urine and the weight of feces when passed. In addition to being 

 weighed every hour, the subject was weighed immediately before and after 

 eating or drinking, passing urine, or defecating. Two series of observations 

 were made, the results of which are shown in curves I and II. Curve I gives 



