THE SPLENO-PANCREATIC INTERNAL SECRETION. 



401 



cells, but these, besides being pyriform, terminate as blind 

 pouches, with their orifice directed toward the glandular lumen. 

 Saviotti also found these minute, delicate, pouch-like chan- 

 nels. Eamon y Cajal, using the Golgi method, found that 

 they sent offshoots into the cells themselves, but that these 

 also ended as blind pouches, their ampullar dilations never 

 reaching beyond the line which separates the granular portion 

 of the cell from the clear area. It is evident that these con- 

 stitute collecting channels for the products of glandular metab- 

 olism: a further indication that no free channel between the 

 blood-stream and the ducts exists. 



Indeed, experimental proof to this effect is available: 

 When the histological examinations of Kiihne and Opie were 



TERMINATION OF SMALL BLOOD-VESSELS IN THE SPL,EEN. (Gray.) 



a, Small artery. 6, Vessel undergoing lymphoid change, c, Vessel 

 continuous with supporting cells, d, Supporting cells. 



mentioned on page 394, it was noted that material injected 

 into the duct of the gland did not penetrate the islands of 

 Langerhans. If, on the other hand, Opie's observation that 

 injections of Berlin blue often filled the vessels of the island 

 per se, leaving the majority of surrounding capillaries empty, 

 is considered in this connection, it is evident that there can 

 be no direct communication between ducts and blood-stream 

 through either the islands of Langerhans or the glandular 

 lobules that contain them. This involves an all-important 

 deduction, however, viz.: that the splenic ferment, as well as 

 the oxidizing substance, merely passes through the pancreas 

 in transit, the latter, in its usual capacity of reagent, being 

 intended only to activate function. 



