AND OVIDUCTS IN THE AMNIOTA. 819 



allantois and yolk-sac balance each other. The shell-enclosed egg 

 is for a short time transfei^red into the marsupium (c/. Hemon, 

 Echidna ; not into a bursa as was imagined by Klaatsch, Gegen- 

 baur, and othei's), which secretes fluid (the " nutritive sweat " in 

 Gegenbaur's unfortunate diction), and this can be taken up by 

 tlie embryo through the porous shell. This may be the reason 

 why the shell is soon cast off. 



In Marsupials the shell is at first still present, but soon absorbed 

 within the oviduct. The egg-membranes, etc., of the embryo 

 establish no structural communication with the iiterus. Tlie 

 ovarial yolk is much reduced. But the yolk-sac becomes en- 

 larged, as in the Monotremes, still occupies a great portion 

 of the inner egg-surface, and has established intimate contact 

 with the serosa. Tiie allantois, being independent of the expansion 

 of the coelom, which results in the driving away of the yolk-sac 

 vessels from the somatopleure, establishes a villous placenta. 

 Such must have been the condition of the Metatheria. 



In Perameles the allantois still reaches the surface, where it 

 is very vascular, and fused with the serosa, a truly respiratory 

 arrangement. The placenta being lost in most other Marsupinls, 

 the allantois reverts to its primary function of urinary receptacle, 

 although apparently late during the foetal life. 



The Metatherian stage may therefore be characterised as one 

 in which the posterior of the two bags, the allantois, has super- 

 seded the previous attenjpts of placeutation by the yolk-sac. 

 The new placenta was perhaps not advanced enough to prevent 

 the foetus from being born soon after the limited amount of yolk 

 was used up. Certainly it did not pass beyond the non-deciduous 

 stage, and it never reached the extent of even the lowest recent 

 Eutherian placenta. Yet one effect of this incipient organ must 

 have been to render the foetus less independent than that of a 

 viviparous reptile. It had therefore still to be transferred into 

 the marsupium, there to be kept moist and suckled in as 

 premature a condition as the Monotreme. 



Two opposite tendencies are inherent to this stage. One 

 palingenetic, to give birth when the yolk is used up ; the other 

 cfenogenetic, to prolong the retention of the foetus because of the 

 compensatory, respiratory, etc. advantages incidental with a 

 placenta. Obviously the Metatherian stage was a half-way 

 house at a parting of the ways to further improvements, leading 

 to Marsupials * and to Placental; 



* We have here an instance of the well-known fact that Group-names based upon 

 single anatomical features are mostly unsuitable for classificatory purposes. The 

 taxouomic value of these characters may be good enough, but they are not diagnostic. 

 If such names were used merely as labels without much intrinsic meaning, well and 

 good, but even the best of us cannot, on occasion, resist taking their face-value for 

 full value. There are Mammalia Implacentalia with a placenta, and we now know 

 that the young Echidna does not lie in a bursa. Odontornithes are a valueless, 

 heterogeneous assembly, and overconlidence in " Ratitse " was responsible for 

 having branded Hetfperorniii as a " swimming Ostrich." " Mammalia " is for- 

 tunately an excellent term, although invented before Monotremes were known. 



