EMBRYOLOGICAL CRITERION OF HOMOLOGY. 109 



or even necessary one. Morphologists can, however, no longer 

 close their eyes to the fact that the primary germ-layers do 

 not have that fixed and absolute value that has so long been 

 attributed to them. This is shown by contradictions both in the 

 origin and in the fate of the germ-layers. It has long been recog- 

 nized that the primary layers are not, as Haeckel originally 

 endeavored to maintain, always differentiated by the same 

 process. On the contrary, there are many processes, some 

 of which differ radically from the very beginning of develop- 

 ment, such as the multipolar delamination of Geryonia as com- 

 pared with the embolic invagination of Amphioxus or Echinus. 

 It is true that here, as elsewhere, a nearly complete series of 

 intermediate forms exists ; but this does not affect the truth 

 of the general proposition or remove the difficulty. Despite 

 these differences of origin, however, the primary germ-layers 

 have been generally regarded as homologous; for they are 

 considered as homologous with the respective layers of the 

 Coelenterata (in which the primitive diblastic condition has 

 been retained), a homology which has persisted through all 

 the transformations of the higher types, and through all the 

 secondary modifications of the process of gastrulation. 



But even this faith is being shaken, since it is becoming 

 more and more clearly apparent, both on general and on special 

 grounds, that even in a prospective sense the inner and outer 

 layers of the diblastic embryo do not always have the same 

 value. Balfour recognized this truth on the general ground 

 that their relation to the " mesoblast " is not always the same, 

 and later researches have, I think, abundantly confirmed his 

 view. As I have elsewhere urged, we certainly cannot regard 

 the layers of the diblastic embryo of LopadorhyncJms as the 

 precise homologues of those of Amphioxits, when in the former 

 case it is the outer layer and in the latter case the inner layer 

 that is to give rise to the entire mesoblast.^ 



1 A curious example of the lengths to which embryologists are driven in the 

 attempt to meet this difficulty is shown in Lwoff's recent attempt to show that 

 the mesoblast and chorda of Amphioxus are derived from the outer layer. My 

 own observations on this point (in Amphioxus) differ widely from Lwoff's in the 

 matter of fact ; but even by accepting his conception of the gastrulation we do 

 not escape the contradiction. 



