488 SCIENCE PROGRESS. 



in the amniote embryo. Supposing that no functional 

 importance in embryonic life belongs to them, it is to be 

 noted that the embryonic branchial blood-vessels are 

 metamorphosed into the vascular system of the adult, as 

 they are in the metamorphosis of Amphibia, and a very 

 considerable and very important part of them persists 

 in the adult structure. This does not account for the per- 

 sistence in the embryo of open branchial clefts, but even 

 these do not wholly disappear in development. The 

 first remains as the Eustachian tube, and even the posterior 

 are related to the development of the thymus gland. It 

 seems to be a general fact that a structure which in meta- 

 morphosis disappears completely may easily be omitted 

 altogether in embryonic development, while one which is 

 modified into something else continues to pass more or 

 less through its orio-inal larval condition. 



If the aquatic larval condition then was ever an adult 

 ancestral condition it must have been the ancestor of Am- 

 phibian forms, and only through these the ancestor of the 

 Amniota. We have therefore to inquire how far the larvae 

 of the Amphibia resemble the adult ancestors of that group. 

 The important resemblances of a tadpole to a fish, and the 

 evidence that fishes are older palaeontologically than Am- 

 phibia, are sufficient to show that the latter are descended 

 from fishes, and that the aquatic larva is not entirely a 

 kainogenetic adaptation. But when we come to compare 

 in detail the Amphibian larvae with all known fishes we are 

 forced to the conclusion that the larva is far from repeating 

 any possible adult condition very exactly. If we suppose 

 that the tadpole of the frog represents the adult ancestor of 

 that form, we imply that that ancestor was entirely destitute 

 of fin-rays and of paired limbs, and that the fore-limbs 

 were originally evolved as outgrowths within the branchial 

 chamber beneath the operculam. We should also imply 

 that the piscine adult ancestor had no scales. Now the 

 recent Dipnoi present an obvious approximation to the 

 Amphibia, and we know that forms belonging, according to 

 the structures capable of preservation in the fossil state, to 

 the Dipnoi existed as early as the Devonian period. We do 



