108 J. P. HILL. 



misleading-, inasmuch as it is applicable not to the Mammalia 

 as a whole, but, so far as it ret-'ers to matters of undisputed 

 fact, to one ouly of the three mammalian subclasses, viz. the 

 Eutheria. So far as the latter are concerned, practically all 

 observers, as we have seeu, are agreed that there is present 

 during at least the early stages of development a complete 

 outer layer of cells which encloses the embryonal anlage 

 or inner cell-mass (that portion of it immediately overlying 

 the latter being termed the " Deckschicht " or " Rauber's 

 layer"). It is, of course, this enveloping layer or tropho- 

 blast which Hnbrecht interprets as a larval membrane. 

 It fulfils the conditions, and were the Eutheria the ouly 

 Vertebrates known to us, the idea might be plausible 

 enough. 



Turning now to the Metatheria, and remembering that these, 

 according to Hubrecht, are descended from the Eutheria, we 

 should naturally expect to find the supposed larval membrane 

 fidly developed, with all its ancestral relations ; and so we do 

 if we are content to accept Hubrecht's interpretation of 

 Selenka's results and figures in the case of Didelphys. The 

 " urentodermzelle " of Selenka is for Hubrecht " undoubtedly 

 the mother-cell of the embryonic knob," the ectoderm of 

 Selenka is manifestly the trophoblast — a complete larval 

 layer. It is no doubt unfortunate that Hubrecht had to rely 

 on the work of Selenka as his source of infoi'mation on the 

 early development of Marsupials, but it must be remembered 

 that he reads his own views into Selenka's fiirnres. On the 

 basis of my own observations on the early ontogeny of Mar- 

 supials, I have no hesitation in affirming that a hirval mem- 

 brane, in the sense of Hubrecht, does not exist in any of the 

 forms (Dasyurus, Perameles, Macropus) studied by me. Tiie 

 observations recorded in the preceding pages of this paper 

 demonstrate, in the case of Dasyurus without the possibility 

 of doubt, the entire absence of any cellular layer external 

 to the formative region of the blastocyst, i.e. in a position 

 corresponding to that occupied by Rauber's layer in Eutheria,, 

 whilst in the case of Perameles and Macropus, they yield not 



