■ 119 



every essential characteristic the structure of this Tongese specimen 

 corresponds with that of 0. complanata; the only real difference 

 being the existence, in the former, of the concentric rows of super- 

 ficial fossse, and their small canals opening into the subjacent cells. 



Fig. 14 teaches us that there is often a difference in the mode of 

 growth, which also in some measure distinguishes the Tongese form. 

 This figure represents a horizontal section of the primordial cell (14, a 

 and b), corresponding with those seen in the vertical section (17, a), and 

 analogous to those seen in 0. complanata (fig. 10, a and b) : it also 

 exhibits the arrangement of the subsequent growths. The primor- 

 dial cell (fig. 14, a), though more or less irregular in its form, is 

 usually globular. This cell is surrounded by a still larger one (fig. 

 14, b), a communicating orifice always existing between them. Here 

 too, as we observed in 0. complanata (fig. 10, c), we often notice the 

 existence of an imperfect rudimentary septum (14, c), partially divid- 

 ing the outer of these chambers. In the case of the specimen 

 represented at fig. 11,1 suspect that the section has traversed one or 

 two such septa, giving rise to the existence of three other large cells, 

 to the right of that which appears to have been the primordial one 

 fig. 11, a). 



The outer cell-wall (14, d) of the large cavity has been perforated 

 by a row of small apertures (14, e), in the same way that we found 

 the outer septum of the cell (fig. 10, d) to present three similar 

 canals, only in the former they are more numerous than in 

 the latter, and surround the whole organism. From this stage we 

 observe a considerable difference in the growth of the two species. 

 Instead of increasing by slow degrees, and only after the development 

 of many rows of cells, exhibiting one in the form of a complete circle, 

 which we have seen to be the case in O. complanata, in the species 

 from Tonga we very frequently obtain a similar circle at once, as if each 

 cell had been the result of the simultaneous protrusion of a gemmule 

 from every one of the apertures in the cell-wall (14, e). Though each of 

 the small cells communicates, through these apertures, with the parent 

 cell around which they are all arranged, I cannot succeed in detect- 

 ing any lateral communication between the contiguous cells forming 

 part of the same primary circle. Such lateral communications are fre- 

 quent enough in the circles of cells formed when the organism attains 

 to a more advanced stage of growth ; but the individual cells of the first, 

 and perhaps also of the second rows, appear to have no communica- 

 tion, excepting with the large cell which they surround, and with 

 those of the circle by which, in their turn, they become surrounded. 



After the formation of this first circle others are successively 



