SECRETION OF THE DUCTLESS GLANDS. 781 



or slight and transient, or absent altogether, depending upon the 

 amount of pancreatic tissue left. According to the experiments 

 of von Mering and Minkowski on dogs, a residue of one-fourth to 

 one-fifth of the gland is sufficient to prevent the appearance of sugar 

 in the urine, although a smaller fragment may suffice apparently 

 if its physiological condition is favorable. The portion of pancreas 

 left in the body may suffice to prevent glycosuria, partly or com- 

 pletely, even though its connection with the duodenum is entirely 

 interrupted, thus indicating that the suppression of the pancreatic 

 juice is not responsible for the glycosuria. The same fact is shown 

 more conclusively by the following experiments: Glycosuria after 

 complete removal of the pancreas from its normal connections may 

 be prevented partially or completely by grafting a portion of the 

 pancreas elsewhere in the abdominal cavity or even under the skin. 

 So also the ducts of the gland may be completely occluded by liga- 

 ture or by injection of paraffin without causing a condition of per- 

 manent glycosuria. 



On the basis of these and similar results it is believed that the 

 pancreas forms an internal secretion which passes into the blood 

 and plays an important, indeed, an essential part in the metabolism 

 of sugar in the body. Moreover, considerable evidence has been 

 accumulated to show that the tissue concerned in this important 

 function is not the pancreatic tissue proper, but that composing the 

 so-called islands of Langerhans. In man these islands are scattered 

 through the pancreas, forming spherical or oval bodies that may 

 reach a diameter of as much as one millimeter. The cells in these 

 bodies are polygonal; their cytoplasm is pale, finely granular, and 

 small in amount. The nuclei possess a thick chromatin network 

 which stains deeply. Each island possesses a rich capillary network 

 that resembles somewhat the glomerulus of the kidney. 



According to Ssbolew,* ligation of the pancreatic duct is followed 

 by a complete atrophy of the pancreatic cells proper, while those 

 of the islands of Langerhans are not affected. Since under these 

 conditions no glycosuria occurs, while removal of the whole organ 

 including the islands is followed by pancreatic diabetes, the obvious 

 conclusion is that the diabetes is due to the loss of the islands. 

 This conclusion is strengthened by reports from the pathological 

 side. A number of recent observers (Opie, Ssbolew, Herzog, et al.) 

 find that in diabetes mellitus in man the islands are markedly af- 

 fected. They show signs of hyaline degeneration or atrophy or in 

 severe cases may be absent altogether. 



Several theories have been advanced to explain the action of 

 the internal secretion of the pancreas. It has been suggested that 

 the secretion contains an enzyme which is necessary for the hydrol- 

 *"Virchow's Archiv," 168, 91, 1902. 



