542 THE PHYLOGENETIC SIGNIFICANCE OP THE MARSUPIAL ALLANTOPLACENTA, 



Relation of the Allantoplacentation of Perameles to that of the Monodelphia. 



This can be considered under three heads, (a) maternal preparation, (h) . 

 embryonal preparation, (c) allantoic completion. 



{a) In Perameles, the maternal preparation consists in the formation of a 

 passive syncytium. The nuclei of this syncytium, become very regularly arranged 

 in the deeper portion of the thickened epitheKum. The arrangement is that of 

 groups of nuclei, each nuclear aggregation being contained in a lobule-like pro- 

 jection of the deeper surface of the proliferated epithelium. Between these 

 lobules, capUlaries ascend into the epithelium and ramify in rich plexuses at and 

 below its surface (Hill, 1897-9). 



Hill (1900) has also described a similar maternal preparation (syncytial 

 formation) in the pregnant uterus of Dasyunis viverrinus and I can myself bear 

 witness to the occurrence of a like phenomenon in the ease of Pseiidochirus 

 eooki. 



' The development of such a syncytium in the latter two genera in which an 

 allantoplacenta does not occur is, to my mind, just as significant, phylogenetically, 

 as the extrusion of surjDlus yolk, which has been shown to occur in early cleavage 

 stages of Dasyurus and of Bidelphys (Hill, 1910; Hartman, 1916, 1919). Just 

 as the latter phenomenon points to the derivation of the marsupial ovum from 

 an ancestral, yolk-laden, meroblastic type, so the trophospongial proliferation 

 which occurs in the pregnant uteri of Dasyurus and Pseudochirus is indicative 

 of the presence of a complex system of placentation in the prototypal marsupial. 

 It is unfortunate that our knowledge of possible alterations in the mucosa of the 

 pregnant uteri of other marsupials, particularly Phascolarctus and Macropus, is 

 absolutely nil. 



The importance of this feature of marsupial intra-utcrine development be- 

 comes evident when a comparison is made with certain of the Monodelphia. An 

 essentially similar process occurs, for example, in the rabbit, as Maximow 

 (1900) and Schoenfeld (1903) have shown us; in Castor [Willey (1914, p. 215) 

 says "At the beginning, the proliferating cells retain their cell-boundaries (PI. 20, 

 fig. 70). Eventually the proliferation will become syncytial"]; and in many 

 others. But there is, in particular, a most striking agreement with the Insec- 

 tivora. Hill (1897) has already pointed this out in the case of Sorex and 

 Hubrecht's (1909, p. 357) description of maternal preparation in Tupaia would 

 serve almost word for word for Perameles. 



(6) This consists in the attachment, to a portion of the previously prepared 

 maternal syncytium, of a circumscribed area of the trophoblast, the chorion. 



In Perameles, immediately on attachment, two distinct trophoblastic layers 

 are formed; a plasmodial layer (the plasmodiblast or plasmoditrophoblast) and 

 a more internally situated cellular layer, the cytoblast or cytotrophoblast. Stmc- 

 turally and physiologically these agree with the similar layers found in the 

 allantoplacenta of many Monodelphia. And this agreement is the more striking 

 the more the details are examined. Phagocytosis on the part of the trophoblast, 

 giant cell formation (diplokaryoeytes and megalokaryoej'tes), the presence of 

 migrating leucocytes and of pigment-laden cells and partial gland necrosis are 

 features of a chorionic attachment in Perameles which in no essential way differs 

 from that found in Monodelphia. 



(e) Allantoic attachment. This has already been described elsewhere. It is 

 quite like that of Monodelphia and results in a similar close apposition of foetal 

 and maternal bloodvessels. 



