ECHINODERMATA. 573 



of the blastopore (p. 551), becomes surrounded by an anal plate, 

 which is interradial in position, and lies on the surface of the 

 oral disc between the orals and radials. On the oral plate in 

 the next interradius is placed the opening of a single funnel 

 leading into the body cavity, which Ludwig regards as equiva- 

 lent to the opening of the madreporic canal (vide p. 55 i) 1 . 



From the edge of the vestibule the arms grow out, carrying 

 with them the tentacular prolongation of the water-vascular ring. 

 Two additional .rows of radials are soon added. 



The stalked Pentacrinoid larva becomes converted, on the 

 absorption of the stalk, into the adult Antedon. The stalk is 

 functionally replaced by a number of short cirri springing from 

 the centro-dorsal plate. The five basals coalesce into a single 

 plate, known as the rosette, and the five orals disappear, though 

 the lobes on which they were placed persist. In some stalked 

 forms, e.g. Rhizocrinus Hyocrinus, the orals are permanently 

 retained. The arms bifurcate at the end of the third radial, and 

 the first radial becomes in Antedon rosacea (though not in all 

 species of Antedon) concealed from the surface by the growth of 

 the centro-dorsal plate. An immense number of funnels, leading 

 into the body cavity, are formed in addition to the single one 

 present in the young larva. These are regarded by Ludwig as 

 equivalent to so many openings of the madreporic canal ; and 

 there are developed, in correspondence with them, diverticula of 

 the water-vascular ring. 



Comparison of Echinoderm Larvce and General Conclusions. 



In any comparison of the various types of Echinoderm larvae 

 it is necessary to distinguish between the free-swimming forms, 

 and the viviparous or fixed forms. A very superficial examina- 

 tion suffices to shew that the free-swimming forms agree very 

 much more closely amongst themselves than the viviparous 



1 I have made no attempt to discuss the homologies of the plates of the larval 

 Echinodermata because the criteria for such a discussion are still in dispute. The 

 suggestive memoirs of P. H. Carpenter (No. 548) on this subject may be consulted by 

 the reader. Carpenter attempts to found his homologies on the relation of the plates 

 to the primitive peritoneal vesicles, and I am inclined to believe that this method of 

 dealing with these homologies is the right one. Ludwig (No. 559) by regarding the 

 opening of the madreporic canal as a fixed point has arrived at very different results. 



