SEGMENTATION OF THE OVQM OF THE SHEEP. 225 



I have taken with some modification from Robinson's paper on 

 the rat. 



I take the trophoblast to be hypoblastic instead of epiblastic. 

 The true epiblast is cut out from the surrounding trophoblast 

 at an early stage (v. Robinson's [46] fig. 8). In the majority 

 of his figures he certainly shows the trophoblast cells to be 

 darker than the hypoblast and more like the true epiblast. But 

 this is not so in all. It is not so in fig. 6. I do not see why 

 the thin side walls of the " segmentation cavity " of this figure 

 should not be derived from the hypoblastic mass, in which 

 casCj since they subsequently give rise to the trophoblast 

 (figs. 8^ 9), the trophoblast may be said to be hypoblastic. 

 Selenka (49) and Duval (25) have given very different accounts 

 of the rat^s development. That there is much to be said on 

 both sides is evident from Weysse's (56) remark that, from his 

 own work upon the subject, he is " still undecided." 



Sobotta (54) describes and figures a separation of the 

 segments into groups of small and large segments, but he does 

 not trace their fate. 



In this case we see that there is no zona radiata. There is 

 a segmentation cavity, which, however, quickly becomes oblite- 

 rated. The very early loss of zona may account for there being 

 no growth round by the "hypoblast" during the early stages. 

 If Selenka's or Duval's description is adopted, it would be open 

 to a similar interpretation to that suggested in diagram C^_^ for 

 the guinea-pig. The same may be said for Mus sylvaticus 

 (Selenka, 50). In this the epiblast is seen to be cut out very 

 early as a definite knob (fig. 36). 



On the left of the plate, also diverging from D, are shown 

 two developmental histories, which seem to form a rather 

 special group. In these the envelopment of the epiblast by the 

 " hypoblast " is supposed to take place rather less quickly. 

 Figs. V to Vy represent what seems to be a possible interpre- 

 tation of the ontogeny of the bat. In this case the surrounding 

 does not occur until after the appearance of the cavity of the 

 blastocyst. 



Fig. V^ is taken from van Beneden and Juliu" (13j, the 



