THE EAULY DEVELOPMENT OF THE MARSUPIALIA. 109 



the slightest evidence for the existence of any such layer. 

 The formative region of the Marsupial blastocyst, which is 

 undoubtedly the homologue of the inner cell mass of the 

 Eutheria, forms from the first part of the unilaminar blasto- 

 cyst wall, and is freely exposed. The remainder of the latter 

 is. constituted by a layer of non-formative cells, the destiny 

 of which is the' same as that of the so-called trophoblast of 

 the Eutherin. I have therefore ventured tosnggest that they 

 are one and the same. If, then, the trophoblast is really a 

 larval membrane, we must assume, in the case of the Mai-- 

 supial, either that its '' Deckschicht " portion has been com- 

 pletely suppi-essed (but why it should have been I fail to 

 understand, unless, perhaps, it is a result of the secondary- 

 acquisition by the Marsupials ot" a shell-membrane, these 

 mammals being even now on the way to secondarily assume 

 the oviparous habit !), or that the non-formative region of the 

 Marsupials is not the homologue ot" the trophoblast, in which 

 case the Marsupials must be held to have entirely lost the larval 

 membrane, since there is no other layer present which could 

 possibly represent it. These considerations may well give us 

 pause before we calmly accept Hubrecht's conception of the 

 trophoblast as a larval membrane present in all mammals 

 without exception. 



Coming now to the Prototheria, v/e find, according to 

 Hubrecht, ^^ the trophoblastic vesicle . . , yet compara- 

 tively distinct," and so it is if we accept the interpretation of 

 Hubrecht of the observations and figures of Semon, Wilson 

 and Hill. The unilaminar blastoderm of these authors is 

 unmistakably the trophoblast. The cells situated internally 

 to that in the region of the white yolk-bed are not ento- 

 dermal, as suggested by Semon, but constitute for Hubrecht 

 " the mother cells of the embryonic knob." I need only quote 

 again the opinion of Assheton thereanent and express my 

 agreement therewith; he writes ('09, p. 233) : "For this view 

 I can see no reason derivable from actual specimens described 

 and figured by those four authors" (Caldwell, Semon, Wilson 

 and Hill). It would appear, then, that the assumption of 



