Reproduction 245 



connection between these structures in sharks and the somewhat 

 similar structures in reptiles or mammals, nor between the "placenta" 

 of a lizard and that of a higher mammal. The exaggerated filamentous 

 gills of the intrauterine larvae of some viviparous salamanders and 

 the much-expanded, bell-shaped gills of the larvae of the "marsupial" 

 frog, Gastrotheca, suggest that the larva may obtain nutriment as well 

 as oxygen from neighboring maternal sources — practically a "branchial 

 placenta." 



Viviparity is commonly thought of as something peculiarly mam- 

 malian. Yet there are viviparous fishes, amphibians, and reptiles. Vivi- 

 parity in fishes is certainly a secondary rather than a primitive condi- 

 tion. This is attested by the fact, among others, that in some viviparous 

 sharks the large egg is encased in a shell whose vestigial character is 

 indicated not only by the extreme thinness and delicacy of the shell 

 but by its possession of filamentous extensions which correspond to 

 the curling "tendrils" (Fig. 198) whereby the deposited eggs of other 

 sharks are attached to external objects. Similar evidence indicates 

 that viviparity has been secondarily acquired by some amphibians 

 and reptiles. The only vertebrate class which contains no viviparous 

 members is Aves. 



In view of the fact that all birds and the most primitive mammals 

 that we know are oviparous, and the further fact that oviparity pre- 

 dominates among the lower classes of vertebrates, it is highly probable 

 that the earliest vertebrates were oviparous and that the animals 

 which constituted the main trunk of the vertebrate genealogic tree 

 were oviparous. But viviparity has appeared on twigs of various lower 

 branches of the tree as well as at its mammalian top. 



The marsupial structures of vertebrates afford another example of 

 convergence in evolution — that is, the independent origin of func- 

 tionally similar but genetically unrelated things. Defining a marsupium 

 as a brood-pouch developed on the external surface of the body-wall, 

 there are marsupial fishes (sea horse; pipefish), marsupial frogs, and 

 marsupial mammals. 



The chordate ancestors of vertebrates must have been small ani- 

 mals and presumably produced small eggs with little yolk. It is likely 

 that primitive vertebrates had small eggs and that large yolk-masses 

 have been secondarily acquired. But even within a small group of 

 vertebrates, the yolk content of eggs may be highly variable, being 

 apparently easily susceptible to evolutionary change. In point of size 

 and content of yolk, the vertebrate egg has evidently had many ups 

 and downs. 



In spite of the diversity of vertebrate methods of reproduction, an 

 evolutionary trend is clearly to be seen. There is a certain extravagance 



