Influence of Induced Diabetes on Tumors. 135 



time of beginning of the treatment does not exceed 20 by 25 

 millimeters, and the animal will survive the treatment for a period 

 of at least ten days. In every case where the animal lives for 

 three days a marked effect upon the growth is to be noted. Small 

 growths show no local reaction, but begin to retrogress at once, 

 while in very large growths there is rapid development of local 

 inflammation and softening. Small growths go on to complete 

 cure, while with large growths the death of the animal most 

 frequently occurs before much of the growth has been absorbed. 



The following protocol of a single experiment is given as 

 illustrative. On May 10, 1912, the growth measured 12 X 10 

 mm., on May 11, 14 X 11 mm., on May 13, 16.5 X 13 mm., on the 

 14th, 19 X 14 mm. The treatment was begun on May 14, and 

 continued for twelve days. On May 16 the growth measured 

 16 X 12 mm., on the 18th 14 X 11 mm., on the 21st, 10X8 mm., 

 on the 25th, 4X3 mm. On June 10 no trace of the growth could 

 be felt on or under the skin. Early in July the animal was 

 chloroformed. Autopsy showed no detectable growth anywhere. 

 This experiment is one of a number of practically similar ones. 

 Controls have always been employed, and many experiments 

 have been carried out upon animals where the growth was un- 

 questionably past the stage where spontaneous retrogression 

 would occur. The largest growth which we have succeeded in 

 curing by the above indicated treatment measured 45 X 47 mm. 

 We believe that this is by far the largest experimental malignant 

 growth which has so far been successfully treated. 



Upon the basis of the above summarized results, we were led 

 to apply a similar method of treatment to some human cases of 

 cancer at the General Memorial Hospital, under the supervision 

 of Dr. James Ewing and Dr. Richard Weil. In some cases the 

 results were slightly encouraging, but the results obtained with 

 rats were far from duplicated. We have so far applied the method 

 in only a very few desperate cases of human cancer, and we believe 

 that the treatment may still be worked out so as to be of value as 

 a therapeutic measure. 



We have studied the glycosuria following injection of phlorhizin 

 and withdrawal of carbohydrate from the diet in one case in detail 

 and may summarize our findings by stating that the urine picture 



