Germinal Layers in Vertebrates. 303 



ception if we would believe that gastrulation has any definite 

 nieaniiif^ at all. This is the first point to be considered in 

 this important question. 



The second point concerns the discrimination of the primary 

 germinal layers. If we would derive the bilaterally symme- 

 trical Chordata from a Gastrula-like form with radial symmetry 

 we must adhere to the homology of the ectoderm and endo- 

 derm of such a Gastrula-like form with the external and 

 internal germinal layers of the Chordata. Since, however, 

 the endodei-m of the Gastrula-like form in the first place 

 gives rise to the intestine, our first step in the determination 

 of the germinal layers is to elucidate the question by what 

 elements or what layer the intestine is formed. To this layer 

 the term endoderm must be applied, it matters not whether 

 .something besides the intestine is formed from these cells or 

 not. 



I will not here enter into the question whether typical 

 gastrulation, i. e. invagination, represents a primary or a 

 secondary mode of formation of the endoderm. Nevertheless, 

 in the interpretation of the conditions which are found in 

 Vertebrates I assume that the bilaterally symmetrical Chor- 

 data are derived from a Gastrula-like form with radial 

 symmetry, since in the development of the lower Chordata it 

 is impossible not to recognize a gastrula, although a some- 

 what modified one. At the same time, in homologizing the 

 germinal layers of the Chordata we must not lose sight of 

 our thesis, that the inner layer (the endoderm) of the gastrula 

 forms the intestine, while the ectoderm constitutes the outer 

 covering ; otherwise the homologization loses all its meaning. 

 Therefore we shall designate as endoderm cells those from 

 which the intestine arises, no matter whether something else 

 is also formed from these cells or not. 



If from this point of view we compare the conditions which 

 exist in Amphioxus and the Vertebrates, we arrive at the 

 following results : — The segmentation of the ovum proceeds 

 in such a way that in holoblastic ova, as a result of the 

 segmentation, we get a blastula, one half of which is com- 

 posed of smaller blastomeres (micromeres) and the other of 

 larger blastomeres (macromeres). The difference between 

 the micro- and macromeres has arisen in consequence of the 

 fact that the former multiply more rapidly than the latter. 

 Since the more rapid multiplication of the micromeres also 

 continues after the formation of the blastula, the micromeres 

 commence to spread out over the macromeres and to grow 

 round them. In cases where we find a single-layered 

 blastula (in Amphioxus) this process takes place in such a 



