EPITHELIOID MEMBRANES IX MOXAXOXID SPOXGES 567 
apertures at the surface of the body, dermal pores or ostia, and 
apertures in the walls of the flagellated chambers, chamber pores 
or prosopyles. The accounts leave it uncertain as to whether the 
latter always have the same character (Minchin, loc. cit.). The 
assumption is usually made that the chamber pores are inter- 
cellular gaps. My own observations on this point are limited to 
the tetractinellid genus Poecillastra (Wilson, '04, p. 107, pi. 15, 
fig. 2). The material seemed to be well preserved and in sections 
it could easily be seen that the collar cells were wide apart and 
rested upon a bounding membrane which connected them, and 
which itself showed no cell boundaries. The chamber pores 
appeared as perforations in this membrane. If we look on the 
membrane as formed of thin extensions from the bases of the 
collar cells, the chamber pores here are equivalent to intercellular 
gaps. On the other hand there are investigators who think the 
chamber pores may be intracellular structures. Thus Evans 
('99, p. 419, figs. 32, 33, 34) is inclined to believe from his observa- 
tions on Spongilla that true porocytes exist in the walls of the 
chambers in this sponge. 
In our ideas of the dermal pores too a certain vagueness prevails, 
which can only be cleared up by further investigations. The dis- 
tinction between the actual aperture and the pore canal should, 
it seems to me, be borne in mind, although where the dermal 
membrane is excessively thin it may be that such distinction is in 
practice impossible. Usually the dermal pores are thought of as 
perforations of the dermal membrane, the two layers of epithe- 
lium being continuous round the margin of the pore. Where cell 
boundaries exist in the epidermis, such a pore, i.e., the actual 
aperture, would have the nature of an intercellular gap and would 
not differ in its fundamental structure from an osculum. I con- 
ceive the pores in Stylotella and Reniera to be of this nature. 
Were the epidermis in these sponges divided up into distinct 
cells, I take it that the pores (comp. figs. 1, 2, 3, 11) would each 
be surrounded by several cells. The observations of several inves- 
tigators, however, have inclined them to believe that the dermal 
pores are intracellular structures, perforations of cells that are 
comparable to the porocytes of Calcarea. In the very young, 
