54 MAMMALS. 



the jerboa. Peeagalea lagotis has consideratle resemblance to a hare, and carries its habits as 

 well as its ears — making a form in the grass like it. The Wombat has been compared to the 

 beaver, or capybara ; but there is not sufficient actual similarity of appearance in them to suit 

 the purpose for which I refer to these analogies. 



Now let us take two (the two most striking, of course) of these resemblances, — two will answer 

 as well as a dozen. Let us take the mice and the flying squirrels as opposed to the Anteciiini 

 and Petauei. All inhabit the same quarter of the world (New Guinea and Australia), and have 

 special ordinal structure — that is, a structure which is found in the whole of the order to which 

 each belongs, and which, therefore, may be assumed to be of primary importance and essential in 

 character. The two first have the rodent dentition, and not the marsupial structure ; the two last 

 have the marsupial structure, but not the rodent dentition. Can these so similar species have 

 descended the one from the other ? Any one looking at the frontispiece would say that the animals 

 there represented must have done so ; that such close external resemblance is impossible on any other 

 supposition. Assume it to be so, both with the mice and the flying squirrels, for if it be so with 

 the one, the same rule must hold with the other — we should then have two animals independently 

 of each other, making the same change, both from a marsupial structure to a rodent, or from a 

 rodent to a marsupial. This is, I think, impossible. Nature never repeats herself. Had it been 

 that the one changed to a rodent, and the other to something else — that might be — but both 

 from the same and to the same is opposed to all that we know either of the laws of nature or of 

 the doctrines of chance. 



Therefore, so far as the question is the origra of species from a single progenitor, I feel con- 

 strained to admit that it cannot be. But may not the whole order of marsupials, or of rodents, 

 or a part of the order embracing those instances where close resemblance exists, have received 

 in one body the impulse of change from marsupial to rodent, or from rodent to marsupial, as I 

 think there is reason to hold is the case in whole bodies of individuals composing a species ? 

 That is an explanation which appears to me to have some germ of truth. 



But the difficulties do not cease here. We have taken the two orders, Marsupials and 

 Rodents, as in pari casii, and looked at the question of the derivation of these similar species 

 from one or other, as if either might be indifierently the oldest ; but one or other must be 

 the elder — they could scarcely be twins, — or triplets if we take in the Insectivora. But might 

 they not be the children of difierent parents ? Is it absolutely necessary that all Mammals should 

 spring from one progenitor or parent stock ? May they not have sprxing from difierent stocks — 

 the bat from the pterodactyles ; the duck-bill, or ornithorhynchus, from birds ; the whale from 

 the ichthyosaurus, and the general mass of Mammals from terrestrial reptUes ? — or have the 

 whole four classes of Vertebrata (mammals, birds, reptiles, and fishes) been developed into being 

 successively in descent, not one from the other, but from some ancestor common to them all ? 

 All these inquiries are involved in the argument whether there is such a thing as an order ; and 

 if so, whether more than one part or member of it can receive simultaneouslj^ the same new 

 impress, as appears to be the case ui species. If this were possible, then I incline to think that it 

 may have been the mode in which the orders of Mammals came into being ; but if that hypothesis 

 must be excluded, then we are driven to the conclusion that each class, order, or natural group, 

 started from some one parent species, into which had been drawn, as into one focus, all the difierent 

 rays of previous form which were afterwards dispersed among the equivalent types which sprimg 

 from it. 



