666 PROF, SCHULZE ON THE STRUCTURE AND ARRANGEMENT 
where the skin is stretched above between the sword hilts and the floricomes. 
The width of the dermal pores corresponds generally with the stricture of the 
dermal septa which surround them, and it appears, as in other sponges, to alter 
frequently in life. 
The water first passes through the dermal pores into subdermal spaces, 
which are here represented by a wide-meshed lacunar network between the 
external skin anda fenestrated layer of tissue extending in the ua of the 
lateral cross-rays of the radiating rapier-like spines. 
In his “Untersuchung tiber Hexactinelliden,” Marsuauu described and 
figured (fig. 62) as skin a membrane which extends between the lateral cross- 
rays of the rapier-shaped spines in such a way that one quadrate area limited 
by these cross-rays contains a single round hole, which may be called a dermal 
pore. Though I am willing to believe that in young individuals the outer layer 
may, under certain circumstances, contain only such distant and isolated pores, 
certainly in fully developed animals the outer film does not extend between 
_the cross-rays of the rapier hilt, but, as above described, further out between 
the ends of the rapier hilts and the floricomes appended to them. What 
really stretches between the cross-rays of the rapier hilts is not so much a true 
membrane as a very wide-meshed network of delicate threads, from which 
similar bands or cords of tissue proceed not only towards the exterior to be 
connected with the true outer membrane, but also mid-cords in the form of an 
open, very irregular meshwork. They traverse wide lacunar spaces, extending 
from the cross-rays of the rapiers between their long processes, which cor- 
respond with the sword-blade, and here and there, by extension to deep fissures 
and passages, at length acquire the significance of water canals. In fact the 
periphery of this very irregular system of lacunze and canals becomes towards 
the interior a complete system of flagellate chambers (ampullaceous sacs, 
CarTER), which I now proceed to describe. 
Each individual chamber has the form of a sac more or less deep, nearly 
circular in transverse section, with the blind end semicircularly arched, and the 
almost circular opening placed transversely to the long axis of the sac. The 
length is on an average about 100y., but may vary between 60 yp. to 150 p. 
The breadth is usually about 60y., varying, however, between 40 to 80 p. 
Besides accidental flaws, which are frequently caused by the hardening and 
further treatment of the preparation, there are often actual variations in form. 
I will only draw attention here to one of these, which occurs so repeatedly that 
it can hardly be regarded as a mere abnormality. There are sometimes 
chambers which are bisected at the bottom by a larger or smaller constriction, 
while the remaining portion and the aperture remain single. I leave it an open 
question whether or not in such cases we have to deal with an actual process 
of division. 
