84 W. J. DAKIN. 



membrane supposed to separate the protoplasm of tliese two 

 parts, but there is no trace of one, and the cell contents of 

 both are continuous. 



The appearance of two definite structures separated by 

 a membrane is due to an external flange or projection existing 

 on the wall of the rod-cells at their junction with the rods, by 

 means of which adjoining rod-cells are connected. 'J'his pro- 

 duces in sections the effect of a " sieve-membrane " with 

 circular holes through which the rod-cells and rods pro- 

 trude. 



It is a rather difficult point to decide. Hensen, so far back 

 as 1865, said that by reason of the rod-cells ending at the 

 same level a sharp bounding line "was formed, which could 

 easily be mistaken for a membrane, but this was not present. 

 Patten did not see it either, but, as stated above, believed 

 there was a delicate membrane, the "terminal membrane," 

 in each rod-cell. Rawitz found no membrane either inside 

 or external to the cells, but Schreiiiei- and Carriere both 

 affirmed its presence. Hesse (34) refers to a sieve-membrane, 

 and on p. 409 he remarks that in some specimens of P. 

 jacobfeus and P. maxim us the inner interstitial cells 

 can be followed up to the sieve-membrane, which is possibly 

 a product of these cells. 



In my opinion the sieve-membrane is, as above stated, due 

 to the extended walls of the rod-cells, and has no part from 

 the interstitial cells. This line is usually well marked in the 

 marginal regions of the retina, where there are no rods borne 

 by the rod-cells (PI. 6, fig. 1, M.ret.). Where necessary, the 

 well-marked line above referred to will be called a " pseudo- 

 sieve-membrane " for convenience in description. In macera- 

 tions of the retina, in -^-^ per cent, chromic acid (the pre- 

 parations being stained with picro-carmine and examined 

 with the oil-immersion) a series of very delicate parallel 

 fibres could be seen running longitudinally on the surface of 

 the rod-cells (PI. 7, fig. 13, CeUs A). It was not possible to 

 follow them proximally to the nucleus. At the junction of 

 rod-cell and rod they bear thickenings (PI. 7, fig. 13, S.m.), 



