46 DE. J. F. GEMMILL ON THE DEVELOPMENT OF 



mulleri Sars I find it to be within the radial cseca that the yolk-granules occur in 

 greatest abundance during and after metamorphosis. In Asterina, from sections 

 which Prof. Macbride kindly showed me, I should judge that a good proportion 

 of the yolk was aggregated within the cells of the enteron as far back as the gastrula 

 stage. Indeed, Asterina and Solaster stand in sharp contrast to one another as regards 

 the relative prominence of the enteron in early stages. In Solaster the development 

 of this part may be said to be post-dated not only with I'eference to its broad features 

 of form and size, but also in the more intimate detail of the accumulation of yolk- 

 granules within its cells. 



A Statolith-like Body. — An interesting modification of the hypenchyme within the 

 posterior coelom now falls to be mentioned. This is a spherical, almost homo- 

 geneous body, about '075 mm. in diameter, which occurs free in the cavity of the 

 posterior coelom, in a large proportion of cases, during the middle and later free- 

 swimming stages (PL I. fig. 10 ; PI. III. fig. 32). As far as I could make out, it 

 takes origin from a small mass of hypenchyme which gradually loses its cellular 

 character, the cell-contents becoming massed together to form the body which has 

 just been described. I venture to suggest that this body may function as a statolith. 

 In the hinder wall of the posterior coelom there is found at this stage an area often 

 pocketed outwards, where the cells are more elongated than over the rest of the 

 ccelom-wall (PI. I. fig. 10). This may possibly be a sensitive area or pit in connection 

 with the body above described. The. changes in the direction of movement which the 

 larva can exhibit, while itself retaining a definite position in the water, and the long 

 duration of the free-swimming stage, are circumstances that seem to emphasise the 

 advantage to the larva of possessing some kind of balancing mechanism. 



It is true that, in general, statoliths and statocysts are of ectodermal origin. But 

 in the case of an echinoderm larva it must be remembered that the greater part, if 

 not the whole, of the muscular tissue is developed from coelomic epithelium, while in 

 the adult it seems to be certain that this epithelium can give rise to nerve-structures 

 such as those which form the entoneural system. 



Synapta ^srovides us with at least one example of an echinoderm larva possessing 

 statocysts and statoliths. Here it would seem that they are of ectodermal origin 

 (Semon, 22), although their development is not fully known. 



In the Synapta itself the presence of " otocysts " has long been recognized 

 (Thomson, 1862, 24, p. 136). More recently they have been found also in various 

 other genera and families of Holothurians, and their function as statocysts has been 

 ascertained (Becker, i). 



'Nervous System. — In the larva perhaps the most definite indication of a nerve- 

 structure is to be found in a network of fine horizontal fibrils lying between the 

 inner ends of the elongated sucker-cells, but superficial to the basement-membrane. 

 These fibrils disappear altogether during the atrophy of the sucker. 



