220 JOHN SUNDWALL 



cells and islet cells of the pancreas. In the former, after stimu- 

 lation in order to rid the cell of the secretion granules which hide 

 the canaliculi, they are seen — ''the apparatus is located in the 

 portion of the cell between the nucleus and the lumen but 

 branches of the canal may appear basalwards along the sides of 

 the nucleus." In the islet cells they resemble those in the 

 acinus both in topography and relation. In neither were com- 

 munications with the exterior seen. 



Cowdry ('12 b) describes them in the spinal ganglion cells of 

 the pigeon. 



That these canaliculi conununicate with the exterior of the 

 cell has been claimed by Holmgren and Retzius, among others: 

 that they do not, by Ciolgi, Kopsch, Misch, Studnicka, and 

 Bensley. 



The canalicular apparatus in the cells of the tubules of the 

 lachrymal gland of the ox can not i^e easily studied owing to the 

 general presence of secretion granules. However, in many 

 glands cells with few or no granules were seen in which this 

 apparatus could be made out readily. It was seen for the most 

 part in that part of the cytoplasm proximal to the nucleus. It 

 appears as a network of open spaces or canals with branches 

 which frequently terminate in slight nodular enlargements. A 

 layer of cytoplasm was alwaj^s seen between these terminations 

 and the cell membrane. The apparatus at one end comes into 

 close proximity to the nucleus, and processes are frequently 

 seen extending around the nucleus towards the base of the cell. 



This apparatus is seen best in sections preserved in formalin 

 Zenker's solution, where it is seen as clear spaces in the cytoplasm. 

 In the tissue stained by the vital pyronin method, which demon- 

 strates the intercellular secretion capillaries, the canalicular 

 apparatus of Homlgren can be made out as well. Here also it 

 appears unstained, in contrast to the red secretion capillai'ies, 

 and is seen as clear spaces in the cell. Consec^uently it is readily 

 differentiated from the intercellular structures. Since the 

 pyronin stains the secretion substance, one can readily conclude 

 that the apparatus is in no manner concerned with the secretion 

 substance of the cell and is indej^endent of the secretion 

 capillaries. 



