426 PROFESSOR LANKESTER. 



regarded justly as modifications of the branchiotroch in con- 

 sequence of their position and structure alone. The two 

 portions of the zygotroch, once differentiated from the archi- 

 troch, may have acquired a considerable independence of one 

 another in their development in some races, whilst in the 

 archaic group of Echinoderms they retain a consentaneous 

 growth. 



In many Chsetopoda, Platyelmia, and Eucephalous Mol- 

 lusca, the embryo, when still almost spherical in form and 

 diblastulous in structure, acquires the cephalotroch, which 

 takes at this early period the position of an equatorial girdle 

 (PI. XXV, fig. 6). Such a larval form I have called the 

 " trochosphere.'' It is clearly not a primitive form, but is 

 derived, by the steps I have just indicated, by a series of 

 adaptations from the telostomiate architrochopor. It is an 

 adaptational larval form common to many marine organisms, 

 and indicates that its ancestors must at one time or another, 

 in larval or adult life, have exhibited the folloAving condi- 

 tions — (1) telostomiate architrochal, (2) metaxial architro- 

 chal, (3) zygotrochal, (4) cephalotrochal — that is, suppression 

 of the branchiotroch. 



Professor Semper has recently made the attempt to set up 

 this much modified larval form, with its premature cephalic 

 circlet, as an important ancestral form, and has announced a 

 " Trochosphsera theory.'^ 



If the views which I have here expressed are well founded, 

 Semper's theory of the trochosphere is not more valuable 

 than his theory of Amphioxus. 



Many ciliated larvae have been called " telotrochic " (figs. 

 8, 13), in consequence of their possessing a perianal circlet 

 of cilia. Gegenbaur is inclined, though not decisively, to 

 refer this " telotroch " to the architroch, regarding it as the 

 equivalent of the branchiotrochal moiety. This I cannot 

 consider to be justified. The telotroch appears to be a 

 metameric repetition of the architroch, or of its branchio- 

 trochal moiety. That such is the case is suggested by 

 the condition of the Tornaria-larva of Balanoglossus. It 

 is possible that the ciliated circlets which are posterior to 

 the area of the architroch or its derivatives, are to be re- 

 garded as altogether secondary structures, and indeed 

 they are so upon any view of the case, in so far as 

 metamerism is a secondary condition. Such secondary 

 circlets as the telotroch and the other more or less numerous 

 circlets of polytrochic larvge, I propose to call " epitrochs," 

 and, accordingly, a ciliated larva, whilst either architrochic 

 or zygotrochic, may be " anepitrochic," or "monepitrochic," 

 or " polyepitrochic." 



