68 J. W. JENKINSON. 



the amnion is exceedingly variable, there is always a stage in 

 which the knob of embryonic epiblast is included in the 

 trophoblast or false amnion. In some cases (Erinaceus, 

 Cavia, Pfceropus) a cavity, the amniotic cavity, which never 

 communicates with the exterior forms inside this knob ; in 

 others (Mus) the cavity is only temporarily in connection 

 with a cavity in the Trager, and in others again (Lepus, 

 Vespertilio, Sus, Ovis, Sorex, Talpa, Tupaia, Tarsius) the 

 epiblastic knob folds or flattens out, the overlying tropho- 

 blastic cells (Rauber's cells) disappear (though this is denied 

 by Assheton for the rabbit, and Heape for the mole), and 

 the amnion then forms in the characteristic Sauropsidan 

 manner. 



Most writers have, therefore, agreed that van Beneden's 

 theory of the metagastrula was premature, that the primitive 

 streak really represents the point where gastrulation takes 

 place, and that the so-called blastopore in those forms in 

 which it has been found (Lepus — though its existence here 

 is denied by Assheton — Talpa, Vespertilio, and Didelphys) 

 is a structure of unknown significance. At the same time it 

 must be admitted that Selenka's account of the development 

 of the opossum is difficult, if not impossible, to reconcile 

 with what we know of other mammals. It may be that a re- 

 investigation of the early stages in the segmentation of the 

 Marsupial ovum will throw fresh light on this subject, or 

 possibly the development of this group of animals is quite 

 different to that of the remaining Placentalia. 



More recently, however, Duval (12) has announced his 

 intention of resuscitating and defending the abandoned meta- 

 gastrula theory of van Beneden. His views are based on his 

 account of the development of Vespertilio murinus. There 

 is in this mammal, according to Duval, an epibolic gastrula, 

 of which the outer layer of clear cells is derived from two of 

 the blastomeres, the inner mass of dark cells from the re- 

 maining two in the four-celled stage, the earliest which 

 Duval was able to find. At the opposite pole of the ovum to 

 that at which the inner mass is attached (the point where van 



