72 BICELLARIID^E. 



opposite to the one bearing the spine. In this condition 

 they bear a general resemblance to Eucratea. This simple 

 growth, l^owever, soon ceases; after a while, from the 

 uppermost in a series two cells are given off, the second 

 taking its origin from the side a little below the angle. 

 From these two cells two series are developed, which be- 

 come more or less adnate, though never completely or . 

 very intimately united. The second line of cells may be 

 compared to a branch, in such a form as Scruparia clavata, 

 which (as sometimes happens in the latter) is closely ap- 

 pressed to the original shoot, and to a certain extent unites 

 with it. 



In B. Alderi the branching is simply dichotomous. At 

 certain intervals the terminal zorecia give off a pair of cells 

 instead of one ; and each pair becomes the first term of a 

 new series. The branches, turned outwards by the two 

 interposed cells*, diverge, forming at the point of diver- 

 gence an acute angle. So loosely, however, are the lines 

 of cells connected, that above the fork they may sometimes 

 be met with not even side by side, but perfectly free and 

 independent. 



Towards the base the shoots are involved in a bundle 

 of tubular fibres, by which the colony is attached ; the 

 fibres pass upwards for some distance along the dorsal sur- 

 face of the cells, to which they are closely adnate. 



The spines on the projecting outer angle of the aper- 

 ture are a striking feature in the present species. There 

 is usually only one to a cell; but two are sometimes present, 

 placed side by side. They are often of great length, two 

 or three times as long as the cell; they are gracefully 

 curved, and taper slightly towards the extremity. 



* The aperture of these cells is always elliptical, whereas that of the 

 others is broad and somewhat truncate above, projecting considerably at the 

 outer angle, and becomes narrow towards the bottom. 



