EPITHELIOID MEMBRANES IX MONAXONID SPOXGES 541 
it as an extension of that layer. If the sponge has been kept in an 
aquarium a short time, preparations show that the pores are for 
the most part about half closed. In fig. 3 two such pores are 
shown as they appear in a thick tangential section similar to that 
from which fig. 1 was made. The pore membrane, pm., here 
extends well over the pore canal. If the sponges have been kept 
some time in the aquarium, regions will be found in which the 
pores are closed. Fig. 2 represents the dermal surface of a thick 
tangential section. The region shown lies over a subdern.al 
cavity and the pores are closed. The outlines of the pore canals, 
p.c, are visible on focussing just below the surface. The thin 
sheet, p.m., covering in the pore canal is the pore membrane. A 
comparison of such preparations shows plainly that the pores 
are closed by a thin extension of the epidermis over the pore canal. 
Owing to the peculiarities of the species, especially unevenness 
of surface, abundance of spicules, and abundance of amoebocytes, 
it is well nigh impossible to observe the closure of the pores in 
living preparations of the dermal membrane as made from the 
normal sponge. Free hand tangential sections of the living 
sponge were sliced off, but these proved of no value. Pieces were 
cut out from the upper part of the oscular lobes in very trans- 
parent regions and where the wall of the lobe is thin, but these 
again were useless for the purpose. I did succeed, however, in 
observing the closure of the pores in life by practicing the follow- 
ing method. 
Free hand sections about one-eighth inch thick were made 
transversely through an oscular lobe, and therefore directly 
across two or three of the main efferent canals. Such a section 
when cut is a circular piece of sponge tissue perforated by the 
segments of these canals. The segments of the canals are of course 
open above and below at each surface of the section. If such a 
section be kept a day in an aquarium a new dermal membrane de- 
velops over both surfaces and over the open ends of the canals. 
The new membrane closing in the canals is smooth and contains 
but few spicules. If the section now be mounted in sea water 
under a cover glass the new membrane over the canals may be 
studied with a high objective. It will be found to contain pores 
