1869.] and Membranijjora Bengal ens is. 57 



In live specimens, the cells are so thoroughly transparent, that 

 their detailed structure is very difficultly noticed ; but in dead cells the 

 membranaceous covering generally disappears, and the solid radii or 

 spines are seen to project towards the median line. It is, however, not 

 always the case that they can be observed, even when the cells appear 

 well preserved. On account of their great tenderness, they not only 

 become in dried specimens variously contorted, but are often very easily 

 broken off ; such is the case in almost all the fossil Membraniporce. In 

 some, even very old cells however, they remain rudimentary, or do 

 not develop at all, with the exception of one posterior, median spine 

 which is always present. Again other very old cells become en- 

 tirely incrusted, even at the aperture. All these variations of the 

 form of the cells and the differences in the arrangement of the 

 marginal spines are amply exhibited in figures 1-4 of plate XII, and 

 these will give a better idea of those changes than* any lengthened 

 description. 



I hardly need to notice the great importance of the study of those 

 structural differences of the cells in one and the same species. In the 

 present case, I find that the cells which spread over a large flat surface 

 usually are short and broad (fig. 3), those which incrust small, thin 

 stems of water plants, and the like, are much elongated and narrower 

 (fig. 4). Were these forms not passing one into the other, and had 

 the animals not in each case been observed, one would certainly may 

 think to have a good reason for acknowledging these forms as distinct 

 species. How different would this be in the case of their being fossil 

 ZiepralioB or Membraniporce ! It is certainly true that we often de- 

 scribe merely fossil forms, and not species. 



Only the sides and the base of the cells are, as I have previously stated, 

 solid ; they are chiefly composed of carbonate of lime, forming a thin po- 

 rous layer. Each cell communicates through a large pore with each of 

 the six adjoining cells. Two of those larger pores are found on each 

 side and one in front and one behind. Sometimes, however, in younger 

 cells the number of large pores is greater. When the polyzoarium is 

 partially free, for instance in growing round a quantity of alga3, each 

 cell usually has at the base a long membranaceous tube, through which 

 a muscle, originating at the lower side of the mantle, is protruded, 

 attaching the cell to the plants, (figs. 7 and 8). The round opening 



