84 



Dr. J. Homans. The Relation of the [Oct. 1, 



are associated with one or two slate blue cells, they are again easily dis- 

 tinguished, but, when no such granule cells are present, and especially when 

 the islets are disintegrated, a positive identification of any one islet cell, as 

 opposed to a duct cell, seems to me impossible. This alteration is illustrated 

 in figs. 5 and 6. Fig. 6 shows an islet of the 26th clay fragment (Experi- 

 ment 3). Several of its cells are obviously B cells, well stained, but the 

 greater part of the islet is not to be distinguished from the masses of duct 

 cells shown in fig. 5, which is drawn from the same specimen. 



I have been unable to determine whether this peculiarity is a result of 

 degeneration or of over-activity, though I am inclined to attribute it to the 

 latter. For, inasmuch as a very small amount of pancreatic tissue, and that 

 portion containing the fewest islets, is left after operation, the islets, whatever 

 their function, must be working at their maximum capacity. That their 

 characteristic specific granules should therefore, after a time, be absent, is not 

 surprising. Moreover, in the duct ligation specimens, though there is 

 considerable formation of scar tissue, the islets, which are apparently not 

 diminished in numbers, and therefore under no physiological strain, present 

 their normal granule content. Even by a neutral gentian stain, with which 

 I have never succeeded in showing the specific granules in the graft islets of 

 the dog, the granules of the duct ligation islets are easily seen. It is 

 suggested, then, that when only a small portion of the pancreas is left, the 

 characteristic granules of the B cells are exhausted. In a badly cicatrised 

 specimen such cells are hard to distinguish, but there is reason to believe 

 that more are present than are immediately evident, and it is possible that 

 the ducts may even be forming islet tissue. 



A consideration of the sugar function in connection with these changes 

 shows that none of the animals became severely diabetic. A considerable 

 amount of sugar appeared sooner or later in the urine of nearly all, and 

 in the dog killed four days after operation it was present from the start. 

 Less pancreas was left in this instance than in the other observations, 

 probably too little to support life. Though the islet cells in this specimen 

 were altered in the manner already noted, no conclusions can properly be 

 drawn as to the relative importance of the acinous or islet tissue to carbo- 

 hydrate metabolism. 



The animal killed 16 days after operation was mildly diabetic from the 

 start. Here the pancreatic acini were highly charged and in good condition, 

 while the islets were composed of feebly stained shrunken cells. It would 

 be reasonable to suppose that the latter were exhausted from carrying too 

 heavy a load. The animal killed at the end of 26 days was not diabetic 

 at first and never appeared at all ill. In this case, as in the 33-day graft, 



