MUTUAL AFFINITIES. 55 



If we must take this as the coiirse of Nature, it will then be most in accordance with the 

 preponderance of evidence to admit the Marsupials to be the parent from which all the other 

 Mammals have descended. The Rodents would come next — indeed, we are not without some grounds 

 for considering them first. But this would not affect our classification materially. If they came first, 

 they gave off the Marsupial as a side branch, which has gone no further ; and they then went on to 

 the development of the other orders. If they did not come first, but descended from the Marsuj)ials, 

 they left them behind, standing still, and proceeded as in the other case. Whether the Monotremes 

 cannot be treated as an exception, being scarcely Marsupials, and be regarded as the direct source of 

 the Edentates (ant-eaters, &c.) is another question involved in doubt. From the Rodents would then 

 spring two antagonistic organizations — the carnivorous and the herbivorous, a partition of a bifold 

 principle embodied in the Rodent itself — some of them like the rat, being both vegetable and flesh- 

 feeders. The Insectivores under our supposed case would be the first of the flesh-feeders. The lower 

 monkeys, Prosimi.e, I would, with M. Gratiolet, refer directly to the Insectivores. From them also 

 I would draw the Carnivores ; and from the Carnivores, the monkeys and man. 



Reverting to the herbivorous section of the Rodents, there seems more direct evidence of the 

 descent of the Pachyderms from them (as through the Toxodon, the Capyb.\ra or Hykax,) than is 

 accessible for most other groups. From the Pachyderms to the Ruminants and Sea Cows follows, 

 as a natural step. The whales are generally supposed to be drawn from the same source, but their 

 location is attended with peculiar difiiculty. 



In all this, however, there is a multitude of objections, puzzles, and contradictions, which can 

 only be got over by an amount of reconciling and explaining away which appears to me to be quite 

 an exception to the usual simplicity of truth. This remark, however, only applies to the orders and 

 general mass of typical forms, each taken as a whole. ^\Tien we come to deal with the individual 

 species of the difierent groups, their affinities to each other are generally simple and clear ; and this 

 inclines me to think that we have probably reached the tnith in the one case, and not in the other. 

 In the meantime, however, feeling no confidence in any plan of descent which I can suggest, I 

 shall, in the main, follow the system of classification of Mammals laid down by Professor Owen ; 

 merely deviating from it in those points on which I have formed a decided opinion of my own 

 in opposition to his conclusions, although derived from the same premises. 



That arrangement proposed by Professor Owen will be found in the Appendix, along with 

 those of Cuvier, Milne Edwards, and other eminent naturalists. The modification of it, which I 

 have adopted, has been given at the commencement of this volume, coupled with a table of the first 

 geological formations in which remains of the different orders and families have been discovered. 



Before leaving this subject I would wish to draw the reader's attention to one noteworthy 

 inference to be drawn from the course of descent, whichever view be adopted ; and that is, the 

 excessive rarity of any important change in the form or structure of animal life. VV hen we thmk 

 of the extraordinary prolificness of nature, of the vast diversity of form and organization scattered 

 all over the globe, we are apt to imagine that the changes must have been frequent. But on 

 more careful inquiry we see that at the utmost only three or four important changes have succeeded 

 each other in any one direction from the first appearance of Mammals, down to the present day. 

 Taking aU the orders together, and including the small changes in each, their number is beyond 

 reckoning ; but, looking at important changes of tj-pe, the number is as I have stated. 



