EPITHELIOID MEMBRANES IX MOXAXOXID SPOXGES 545 
or more large clear vacuole-like spaces. This appearance again 
is probably due to cracking of the epidermal layer, perhaps coupled 
with a violent coagulation set up by the sublimate. The appear- 
ance is certainly an artefact, although the pieces in many places 
look at first sight like cells. 
Silver nitrate. Thin pieces were sliced off from the surface of a 
living sponge, and were fixed 5-10 minutes in one-twentieth 
per cent osmic acid. After thorough washing in distilled water, 
they were transferred to one per cent silver nitrate and exposed to 
direct sunlight 5-10 minutes (Hert wig's method). After washing 
and running up through the alcohols, strips of epidermis were 
peeled off and mounted in balsam. Tangential sections were 
also made and mounted in balsam. As a control small hydrome- 
dusae were stained in the same way. The subujnbrellar surface 
of the latter showed the usual polygonal network of distinct 
brown lines, marking out the epithelium cells. The method 
was employed several times on favorable days. 
Stained in this way the surface of Stylotella frequently exhibits 
no lines that in any way suggest cell boundaries. But in places an 
appearance is got with a Zeiss D objective as if epithelium cells 
were marked out. Examination with an immersion objective 
shows that the appearance (fig. 4) is due to artefacts and not to 
the presence of epithelium cells. The facts may be summed up 
as follows. The network of lines is below the thin surface layer. 
The lines are no browner than other strands, viz., have the 
osmic and not the silver stain. In the meshes are irregular 
masses that are usually nucleated. The areas marked out by the 
lines may var}' greatly in size. It is plain that such areas cannot 
be epithelium cells. The appearance is probabl}' caused by vio- 
lent coagulation of mesenchyme cells and strands. Inter -cellu- 
lar connectives and parts of cell bodies remain as the network of 
brown strands, while the cell bodies, contracted and torn loose 
from the connectives, remain as the irregular masses that lie 
in the meshes. 
