THE METABOLISM OF THE CARBOHYDRATES 687 



food. The former is made up of eggs, fish, green vegetables, fat, etc., 

 and the latter, to start with, should consist of 100 grams of bread, dis- 

 tributed between the two main meals of the day, one of which is break- 

 fast. This diet should be continued until the glycosuria either disappears 

 or attains a constant level. If it disappears, the case is classified as a 

 mild one of diabetes, and the daily allowance of bread may be increased, 

 by 50 grams a day, until the sugar again makes its appearance in the 

 urine, indicating that the assimilation limit has been reached. For 

 therapeutic purposes, the patient should now be instructed to take about 

 three fourths of this amount of carbohydrate in his daily ration, and 

 he should be supplied with explicit instructions in the shape of diet 

 tables as to what variety and quantities of the various carbohydrate 

 materials his food may contain. His urine should be examined at fre- 

 quent intervals once a week and he should be instructed as to the 

 nature of his disease and the importance of his remaining aglycosuric. 

 By further treatment such so-called latent cases of diabetes may be 

 kept in perfect health for many years. 



When, on the other hand, the glycosuria persists with 100 grams of 

 bread in the daily ration, this must be reduced to 50 grams, and if after 

 some days the first reduction does not suffice to render the urine free 

 from sugar, carbohydrates must be withheld entirely from the diet. 

 If the glycosuria does not now disappear, the case is to be considered 

 severe, and it may be necessary to undertake the starvation treatment, 

 which has recently been developed in this country by Allen 18 and Joslin 19 

 with apparent success. By the reduction of carbohydrate, or by the 

 starvation treatment, it is usually possible to make even the severest 

 cases of diabetes aglycosuric. When this has been attained, the amount 

 of protein or carbohydrate food may be cautiously increased until just 

 short of the assimilation limit. 



To avoid error caused by irregular absorption from the intestines, some investigators 

 have recommended the determination of the assimilation limit after intravenous or 

 subcutaneous injections of sugar. But even this refinement in technic has not, as a 

 rule, had the effect of rendering the results of any very evident value as a criterion 

 of the utilization of glucose in the animal body. The reason for the unreliability of 

 this method is mainly that the period of injection of the glucose solution usually oc- 

 cupies only a few minutes, so that it causes a sudden instead of a very gradual in- 

 crease in the sugar concentration of the blood, the conditions being therefore quite 

 unlike that which exists during the normal absorption of glucose from the intestine. 

 The mechanism by which the body ordinarily disposes of excessive amounts of glucose 

 absorbed into the portal blood, is not adjusted to operate when the systemic blood 

 is suddenly overcharged with this substance. In the one case the glucose is a food- 

 stuff; in the other, because of its excessive concentration in the blood, it is more or 

 less of a poison. Such results, in other words, merely show us how much glucose can 

 be added at one time to the organism without any overflow into the urine, but they 



