DISCUSSION OF RESULTS AND GENERAL CONCLUSIONS. 117 



averaging not far from 19 per minute, the deep breathing of impending coma 

 was very noticeable, the respirations being easily counted from across the room 

 with one's back to the subject. A gradual fall in the carbon-dioxide tension 

 will be noticed during the two days on which experiments were made. Between 

 noon of February 14 and 7 a. m. of February 15 the subject took 90 grams of 

 levulose, but this seems to have had no effect on the carbon-dioxide tension in 

 the alveolar air. 



Comparison of the Metabolism of Diabetics with that of 



Normal Individuals. 



One of the difficulties hitherto encountered in the interpretation of experi- 

 ments on the metabolism of diabetic patients has been the lack of sufficient 

 controls with normal individuals. In our previous report we were able in a 

 measure to supply this deficiency, and, we believe, presented enough data 

 regarding normal individuals to establish a satisfactory base-line for com- 

 parison. Criticism, however, has arisen not only as to the character of the 

 normals selected, but also as to the number. Lusk, 1 whose opinion on matters 

 of diabetes is not to be regarded lightly, maintained that the normal individuals 

 weighed very much more than did the diabetics and hence were not comparable 

 on the basis of per kilogram of body-weight. In the second place, Lusk laid a 

 great deal of stress upon the fact that one of the subjects, Case I, was extremely 

 high-strung, nervous, and apprehensive, and by no means an ideal subject. He 

 maintained that the results with this individual played too important a role in 

 our determination of the apparent increase in metabolism. Leaving out of 

 consideration the heavy individuals used in the normal comparison, and omit- 

 ting, also, the results obtained with Case I, he computed that the increase in 

 the heat production was but 5 per cent instead of the 15 per cent that we 

 maintained. 



At the time of going to press with the first diabetic publication, the num- 

 ber of normal individuals available for comparison was somewhat limited, and 

 it was obviously desirable to increase this as rapidly as possible. Fortunately 

 in the intervening time we have been able to study a sufficiently large number 

 of normal individuals so as to extend the list, and are now prepared to meet 

 this criticism by presenting data based on (1) an enlarged number of normal 

 subjects, and (2) a normal individual corresponding in height and weight for all 

 of the diabetics save one. We also have normals to correspond with two dia- 

 betics during the several stages of their disease when their weight changed. 

 While, therefore, in the first report the comparisons were made between 

 normal individuals collectively and diabetic individuals collectively, we shall 

 attempt in this publication to compare each diabetic with one or more normal 

 individuals with essentially the same body-weight and height, and finally, as 

 before, to compare normals and diabetics collectively. Under these conditions 

 the criticism of Professor Lusk can be much better met than it could have been 

 two years ago, when the normal data were limited. That Professor Lusk's 



'Lusk, loc. cit. 



