562 
H. V. WILSON 
arranged sphincter fashion. But the shape and arrangement of 
the lining epithehal cells suggests plainly that they are the closers 
of the canal. As I have said the cells lining an expanded canal are 
in general elongated transversely to the long axis of the canal. 
Very often the cell is so long and narrow that it is properl}^ de- 
scribed as fibre-like (figs. 20 and 21). Mingled with such one finds 
other cells that are not greatly elongated and still others that 
are polygonal (fig. 21). I have examined some canals in which 
contraction had very materially diminished the size of the lumen. 
In these I found that a very large number of cells were either only 
moderately elongated or were polygonal. This is what one would 
expect to find if the epithelial cells do by contraction shorten and 
so tend to close up the canal. 
In this connection it may be noted that in the case of contracted 
canals, when seen in cross section, the surrounding mesenchyme 
(collenchyma) cells are found to be greatly elongated and arranged 
in such fashion that they radiate outwards from the canal wall. 
The appearances suggest that as the epithelial cells are the closers 
of the canal, the surrounding collenchymal cells act as openers. 
COMPARISON 
It is well known that in a large number of sponges the dermal 
surface is covered and the canals lined with a single layer of cells 
(pinacocytes of Sollas) forming an epithelium. It was F. E. 
Schulze who in 1875 first established this fact. After demon- 
strating the presence of epithelia in Sycandra he showed in 
succeeding numbers of his classical ^'Untersuchungen'' that the 
same or very similar structural conditions are found in a great 
variety of sponges. Schulze's conclusions have been confirmed 
and extended, for the same and other forms, by many observ- 
ers. A review of the literature shows, however, that both 
Schulze and other observers have now and then, in this sponge or 
that, been unable to demonstrate the presence of distinct cells in 
the epidermal membrane. Possibly in some of these forms the 
epidermis is a continuous syncytium as in Stylotella and Reniera. 
With regard to the canal lining a number of recorded facts suggest 
