METAMORPHOSES OF ANIMALS. 131 



of the animal, and not on the ventral side, as in the Verte- 

 brata, that a solid basis was obtained for the natural limi- 

 tation of true homologies. It now appears more and 

 more distinctly at every step of the progress Embry- 

 ology is making, that the structure of animals is only 

 homologous within the limits of the four great branches of 

 the animal kingdom ; and that general homology, strictly 

 proved, proves also typical identity, as special homology 

 proves class identity. 



The results of all embryonic investigations of modern 

 times go to show more and more extensively, that animals 

 are entirely independent of external causes in their de- 

 velopment. The identity of the metamorphoses of ovipa- 

 rous and viviparous animals belonging to the same type 

 furnishes the most convincing evidence to that effect. 1 

 Formerly it was supposed that the embryo was affected 

 directly by external influences, to such an extent, that 

 monstrosities, for instance, might be ascribed to the in- 

 fluence of external causes. Direct observation has shown 

 that they are founded upon peculiarities of the normal 



1 This seems the most appropriate though they may agree in laying 

 place to remark, that the distinction eggs or bringing forth living young, 

 made between viviparous and ovipa- The essential feature upon which 

 rous animals is not only untenable any important generalization must be 

 as far as their first origin in the egg based is, of course, the mode of de- 

 is concerned, but also unphysiolo- velopnient of the germ.^ In this re- 

 gical, if it is intended, by this de- spect we find that Selachians, whether 

 signation, to convey the idea of oviparous or viviparous, agree with 

 any affinity or resemblance in their one another. This is also the case 

 respective modes of development, with the bony fishes and the reptiles, 

 Fishes show more distinctly than any whether they are respectively ovi- 

 other class, that animals, the devel- parous or viviparous ; even the pla- 

 oprnent of which is identical, in all cental and knplacental Mammalia 

 its leading features, may either be agree with one another in what 

 viviparous or oviparous ; the differ- is essential in their development, 

 ence here arising only from the con- Too much importance has hitherto 

 nection in which the egg is devel- been attached to the connections in 

 oped, and not from the development which the germ is developed, to the 

 itself. Again, viviparous and ovipa- exclusion of the leading features of 

 rous animals of different classes differ the transformations of the germ it- 

 greatly in their development, even self. 



K 2 



