As another example, let us take the origin of wings, such as exist in birds. Here we find an arm, the bones of the hand of which are atrophied and reduced in number, as compared with those of most other Vertebrates. Now, if the wing arose from a terrestrial or subaerial organ, this abortion of the bones could hardly have been serviceable--hardly have preserved individuals in the struggle for life. If it arose from an aquatic organ, like the wing of the penguin, we have then a singular divergence from the ordinary vertebrate fin-limb. In the ichthyosaurus, in the plesiosaurus, in the whales, in the porpoises, in the seals, and in others, we have shortening of the bones, but no reduction in the number either of the fingers or of their joints, which are, on the contrary, multiplied in Cetacea and the ichthyosaurus. And even in the turtles we have eight carpal bones and five digits, while no finger has less than two phalanges. It{107} is difficult, then, to believe that the Avian limb was developed in any other way than by a comparatively sudden modification of a marked and important kind. [Illustration: SKELETON OF AN ICHTHYOSAURUS.] How, once more, can we conceive the peculiar actions of the tendrils of some climbing plants to have been produced by minute modifications? These, according to Mr. Darwin,[101] oscillate till they touch an object, and then embrace it. It is stated by that observer, "that a thread weighing no more than the thirty-second of a grain, if placed on the tendril of the _Passiflora gracilis_, will cause it to bend; and merely to touch the tendril with a twig causes it to bend; but if the twig is at once removed, the tendril soon straightens itself. But the contact of other tendrils of the plant, or of the falling of drops of rain, do not produce these effects."[102] But some of the zoological and anatomical discoveries of late years tend rather to diminish than to augment the evidence in favour of minute and gradual modification. Thus all naturalists now admit that certain animals, which were at one time supposed to be connecting links between groups, belong altogether to one group, and not at all to the other. For example, the aye-aye[103] (_Chiromys Madagascariensis_). {108} was till lately considered to be allied to the squirrels, and was often classed with them in the rodent order, principally on account of its dentition; at the same time that its affinities to the lemurs and apes were admitted. The thorough investigation into its anatomy that has now been made, demonstrates that it has no more essential affinity to rodents than any other lemurine creature has. [Illustration: THE AYE-AYE.] Bats were, by the earliest observers, naturally supposed to have a close relationship to birds, and cetaceans to fishes. It is almost superfluous to observe that all now agree that these mammals make not even an approach to either one or other of the two inferior classes.