68 PROCEEDINGS OP THE AMERICAN ACADEMY 



geschieJit spider an der uhrigen Peripherie. So hesteht der Keim aus 

 zwei Schic/t/en, welche iin verdickten Hande zusammeidiungcn" This 

 description of the Trout corresponds exactly with Ilaeckel's account of 

 the development of the Gadoid ovum : " Jetzt folgt der hochst wichtige 

 und interessante Vorgang, den ich als Einstulpung der Blastula auf- 

 fasse und der zur Bildung der Gastrula fiilirt (Fig. 63, 64). Es schliigt 

 sich nciinlich der verdickte Saum der Keimscheibe, der *■ Randwidst' 

 oder das Properistom, nach innen urn und eine diinne Zellensc/iickt 

 wdchst als directe Fortsetzung desselhen, tvie ein immer enger werdendes 

 Diaphj'agma, in die Keimhiihle liincin. Diese Zellenschicht ist das 

 entstehende Entoderm." (p. 439.) In one respect this account is 

 incorrect ; for it represents the entoderm as spreading beneath the 

 entire blastodisc and forming a floor to the cleavage-cavity. There is 

 a plain rolling-under, or involution, as an initiatory step in the forma- 

 tion of the ring ; but we believe that the process is, in the main, more 

 correctly described as an ingrowth, due both to a rapid multiplication 

 of the cells, and also to the centrifugal expansion of the ectoderm. 

 The floor cells of the cleavage-cavity are entirely periblastic, and have 

 nothing whatever to do with the ring. The inner edge of the ring 

 represents the limit of the ingrowing layer, which is thus confined to 

 a narrow arc, precisely as in the elasmobranchs, according to Balfour's 

 statements. 



As to the significance of this inflected growth of marginal cells, 

 there is every reason to believe that it is fundamentally the same phe- 

 nomenon that has been so often described in other groups of animals 

 as accompanying the epibolic or circumcrescent expansion of the 

 ectoderm: in short, it must be regarded as the equivalent of a gas- 

 trula invagination. There is, of course, no such wholesale invagina- 

 tion as supposed by Ilaeckel, nor is such an invagination a necessary- 

 consequence of the view that the ring is the homologue of the lip of 

 the blastopore in Amphioxus. Regarded as an invagination, the 

 process is an extreme abbreviation of that seen in Amphioxus, since 

 it is limited to an arc that agrees very nearly in width with the em- 

 bryo (exclusive of the yolk-sac). If we take into consideration the 

 embryonic entoderm (periblast) as well as the involuted entodermic 

 ring, it would then be perfectly correct to say that there is a complete 

 ingrowth ; for the periblastic cells begin to multiply centripetally 

 shortly before the ring appears, and reach the centre of the floor of 

 the cleavage-cavity a little after it is formed. Thus the periblastic 

 portion of the entoderm can be said to participate in, an<l to form a 

 part of, the general ingrowth of the entoderm. That this ingrowth 



