The lowest Vertebrata[34] are perfectly limbless, and if, as most Darwinians would probably assume, the primeval vertebrate creature was also apodal, how are the preservation and development of the first rudiments of limbs to be accounted for--such rudiments being, on the hypothesis in question, infinitesimal and functionless?
In reply to this it has been suggested that a mere flattening of the end of the body has been useful, such, _e.g._, as we see in sea-snakes,[35] which may be the rudiment of a tail formed strictly to aid in swimming.

Also that a mere _roughness_ of the skin might be useful to a swimming animal by holding the water better, that thus minute processes might be selected and preserved, and that, in the same way, these might be gradually increased into limbs.

But it is, to say the least, very questionable whether a roughness of the skin, or minute processes, would be useful to a {39} swimming animal; the motion of which they would as much impede as aid, unless they were at once capable of a suitable and appropriate action, which is against the hypothesis.

Again, the change from mere indefinite and accidental processes to two regular pairs of symmetrical limbs, as the result of merely fortuitous, favouring variations, is a step the feasibility of which hardly commends itself to the reason, seeing the very different positions assumed by the ventral fins in different fishes.

If the above suggestion made in opposition to the views here asserted be true, then the general constancy of position of the limbs of vertebrata may be considered as due to the position assumed by the primitive rugosities from which those limbs were generated.