ISLANDS OF LANGERHANS 695 



several laboratory mammals, and in the toad. It would be well worth 

 while to study their occurrence throughout the vertebrate scale, for quali- 

 tative studies in phylogeny are just as much needed as are the quantita- 

 tive. 



Reactions to Experimental Conditions. The islands of Langerhans 

 have proved thus far to be rather unresponsive to experimentation. Bens- 

 ley, contrary to earlier observers, has found no noticeable change in num- 

 ber following secretin administration and acute inanition. The effect, if 

 any, of chronic inanition still remains to be determined. Thyroidectomy 

 is said to produce an increase in amount of island tissue. Kojima claims 

 to have discovered that feeding rats with the posterior lobe of the hypoph- 

 ysis causes the cells of the islands of Langerhans to stain more intensely 

 with Mallory's stain, on account of the accumulation in them of granules 

 possessing an affinity for Orange G. 



Relation to Diabetes. Homans (&) alone has been able to correlate 

 structural changes in the cells with apparent variations in physiological 

 activity. He produced various degrees of diabetes in the dog by removal 

 of portions of the pancreas, with the result that the mitochondria in the B 

 cells became first accentuated and ultimately disappeared in association 

 with hydropic degeneration, while those in the other cells showed no 

 changes whatsoever. He believes that the B cells pour out a secretion 

 which facilitates the metabolism of carbohydrates, and that these changes 

 are indicative of initial hyperactivity followed by fatigue and exhaustion ; 

 and indeed this would seem to be the most likely interpretation. It is 

 possible, however, that the alterations may result from the diabetes, or 

 in other words, that the B cells are merely peculiarly sensitive to the 

 altered conditions. 



Secretory Phenomena. Intracellular secretion antecedents are un- 

 known. The mitochondria in Homan's experiments merely serve as indi- 

 cators of the activity of the cells. There is no reason to suppose that 

 they themselves produce the secretion. Thus far attempts independently 

 to influence the specific granulations experimentally have proved unsuc 1 

 cessful. In the absence of further microchemical studies, we can only say 

 that these granules are rather dense aggregates of cytoplasmic material 

 necessary to^the organization of the cell, and that they display a surpris- 

 ing constancy in phylogeny. 



