128 METABOLISM AND NUTRITION 



its maximum percentage of glycogen ; but this can scarcely have been the case 

 in all the experiments. 



[For example, Lusk and his co-workers have shown that producing- phloridzin 

 diabetes in fasting dogs may cause the proteid metabolism to rise 333 to 560 

 per cent. They have also shown that in these diabetic dogs, whether fasting, 

 or fed on meat alone or on fat alone, no more fat is burned than in the same 

 dog when he is normal and fasting. Thus, in one experiment a dog weighing 

 11 kg. burned on the second fasting day 20.19 g. proteid and 55.87 g. fat with 

 a total of 606.81 Cal. Made diabetic, he lost on the fifth day 39.4 g. of dextrose 

 in the urine. He burned on this day 67.38 g. proteid and 51.15 g. fat, a total 

 of 605.77 Cal. The calories lost in the urinary sugar, therefore, are very exactly 

 compensated for in the increased proteid metabolism. Since this dog had re- 

 ceived only 115 g. of fat as food in the seven days, it is of course impossible that 

 the great amount of sugar in the urine should have come from glycogen stored 

 in the body (cf. also page 125). It must have come largely from proteid. ED.] 



Recently the question of a final conversion of fat into glycogen has been 

 actively discussed. As proof of this cases of diabetes occurring naturally or 

 produced artificially have been cited, in which on a carbohydrate-free diet the 

 quantity of sugar in the urine was too large to be accounted for by the proteid 

 destroyed. 



The first question to be asked in this connection is, how much sugar can 

 proteid yield? In a number of experiments on dogs with pancreatic diabetes, 

 Minkowski found the ratio of N to dextrose in the urine to be 1 : 2.8 ; in phlo- 

 ridzin diabetes, as v. Mering has shown, the ratio may be 1 : 5. 1 Remembering 

 that for every 1 g. N in the human urine, on the average 0.72 g. C are elimi- 

 nated by the same channel, and that proteid contains 3.28 g. C for every gram 

 of N, the carbon remaining in the body which might go to form sugar would 

 amount to 2.56 g. (3.28 0.72) for each gram of N ingested. Since dextrose is 

 forty per cent carbon, 2.56 g. C would correspond to 6.49 g. dextrose and hence 

 the utmost yield from proteid alone would be 6.4 g. dextrose for each gram of N. 



In order to prove a formation of glycogen from fat it is necessary there- 

 fore that the proportion of dextrose to N in the urine should be greater than 

 6.4 to 1. [Now Lusk has shown that in phloridzin diabetes in dogs and in 

 the most acute form of diabetes mellitus the ratio is fairly constant at 3.65 

 to 1 ; and nobody has ever positively observed a ratio so high as 6.4 to 1. ED.] 

 For the present then we must say that fat is not to be reckoned among the 

 sources of glycogen or sugar in the body. On the other hand it is fairly certain 

 that glycerin is a mother-substance of glycogen. 



Proceeding on the assumption that fat is not a producer of glycogen, Lander- 

 gren has endeavored to explain the ability of carbohydrates, mentioned at page 

 121, to spare proteid to a greater extent than does fat. In his opinion the body 

 has a specific need for carbohydrates. If they are not supplied in the food, they 



1 [In the experiments of Lusk and others a ratio as high as this is only obtained imme- 

 diately after the injection of phloridzin, when there is a preliminary sweeping out of 

 sugar. ED.] 



