No trace of such a structure exists in any bird or in any reptile, and yet it crops out again in certain sharks.

There indeed it might well be supposed to end, but, marvellous as it seems, it reappears in very lowly creatures; namely, in certain of the ascidians, sometimes called tunicaries or sea-squirts.

[Page 82] Now, if we were to concede that the ascidians were the common ancestors[61] of both these sharks and of the higher mammals, we should be little, if any, nearer to an explanation of the phenomenon by means of "Natural Selection," for in the sharks in question the vascular prominences are developed from one foetal structure (the umbilical vesicle), while in the higher mammals they are developed from quite another part, viz.

the allantois.
[Illustration: Upper Figure--ANTECHINUS MINUTISSIMUS (_implacental_).
Lower Figure--MUS DELICATULUS (_placental_).] So great, however, is the number of similar, but apparently independent, structures, that we suffer from a perfect _embarras de richesses_.

Thus, for example, we have the convoluted windpipe of the sloth, reminding us{83} of the condition of the windpipe in birds; and in another mammal, allied to the sloth, namely the great ant-eater (Myrmecophaga), we have again an ornithic character in its horny gizzard-like stomach.