INTERSTITIAL CELLS OF TESTICLE IN DIDELPHYS. 183 



places by small capillaries. I conclude, therefore, that the intra- 

 cellular processes are formed by the substance filling the inter- 

 cellular spaces, a substance which, in my opinion, represents 

 the product of secretion. If this interpretation be correct, as 

 I believe it to be, here is a remarkable instance of a gland with 

 internal secretion, in which the product of secretion can be 

 followed, owing to its staining qualities, from within the cell 

 into the passage of excretion. Favorable sections, like that 

 represented in Fig. 13, carry one even a step farther and show the 

 same product in a vessel. 



What, then, is the real nature of these intracellular and inter- 

 cellular spaces, and of these vessels with which the intercellular 

 spaces seem to be in some way connected? The penetration of 

 lymphatics into a cell, not to speak of the much discussed and 

 very disputable conditions described by Adamkiewicz (1886, 

 1900), Browicz (1902, i and 2), Schafer (1902, 1903), Schlater 

 (1902) and others, has been repeatedly reported. At first Holm- 

 gren interpreted his trophospongium as such. Among the more 

 recent observations I may mention those published by Ciaccio 

 (1903), Felicine (1904), Kumita (1909) and Matsunaga (1909). 

 The first three authors studied the suprarenal. Ciaccio, after 

 the application of Golgi's method, and Felicine and Kumita, 

 as a result of injections, come to the conclusion that there exists 

 a pericellular net of capillaries which send processes into the cells. 

 Matsunaga has published similar results after his study of the 

 thyroid by the injection method. Kumita, however, wonders 

 whether these intracellular processes are really vessels, or "ob 

 sie nicht durch einen sekretorischen Zellzerfall zustande gekom- 

 men sind" (p. 325). This last hypothesis, however, in his 

 opinion can hardly be reconciled with the fact that the injected 

 mass never spreads over all the cell, and that the intracellular 

 space has a definite shape, ending often in a swelling in the neigh- 

 borhood of the nucleus. No matter what the interpretation 

 may be in these cases, it appears to me that the intracellular 

 processes in the interstitial cells of the opossum are not well 

 defined canals of any kind neither processes of lymphatics nor 

 intracellular ducts. They are always exceedingly irregular in 

 appearance and are never seen in an empty condition, but always 



