MAMMALIA 



405 



without establishing any definite structural connection between the 

 fetal and the uterine membranes, and that still others have a well- 

 defined type of placental gestation. We may quickly dispose of the 

 situation involved in the egg-laying mammals by saying that their 

 mode of development is essentially sauropsidan, and need not be 

 repeated here. 



The marsupials present a wide variety of conditions. Their eggs 

 though minute are somewhat larger than those of the monodelphian 

 mammals. The embryo has a brief period of uterine gestation, 

 though no fixed nor definite uterine attachment is established. In 



!t///A 



Fio. 209. — Diagram of the embryo and fetal membranes of the marsupial 

 Bypsiprymnus rufescens (on the left), all, allantoic cavity; amn, amnion; amn. c, 

 amniotic cavity; coel, extra-embryonic ccolom; ser, chorion or serous membrane. 

 (From Parker and Haswell, after Semon.) 



Fig. 210. — Diagram of embryo and fetal niembranes of Phascolarcius cinereus 

 (on the right). Letters as in fig. 209. (From Parker and Haswell, after Semon.) 



most marsupials a large part of the surface of the egg is covered over 

 by the compressed and expanded yolk-sac, as in Hypsiprymmis 

 (Fig. 209). In Phascolardus (Fig. 210) the allantois is in contact 

 with part of the surface. In Perameles (Fig. 211) a primitive type 

 of allantoic placenta is formed and sends out vascular outgrowths 

 into the maternal tissues, much as does the Trager or primary pla- 

 centa of the rodents and annadillos, among true placental mammals. 

 Just here it may not be out of place to recall that it is believed by 

 several leading authorities that the conditions found in the mar- 

 supials of to-day are not primitive but largely degenerate, and that 



