ON THE METHOD OF PAL.EONTOLOGY 435 



not necessarily concern itself, and so far as the mere restorations of 

 the palEeontologist are concerned, it is a wholly irrelevant question. 

 The empirical laws of morphology supply all that the palaeontologist 

 requires for this object. 



Let us imagine that all existing animals had perished, but that 

 their dead forms were gathered together and submitted to the 

 investigation of some intelligent being from whom the knowledge 

 that they had ever lived was concealed. He would of course remain 

 entirely ignorant of physiology and all its laws. Life, if he were 

 acquainted before only with physical and chemical phenomena, 

 would be an inconceivability, and the conception of adaptation to 

 purpose, of physiological correlation, would fail to suggest itself 

 where nothing was known of actions or functions. 



Nevertheless, by the careful comparison of one form with another, 

 he would see that in one set of specimens certain structural peculiari- 

 ties were invariably associated, in another set others, and he would 

 thus arrive at precisely the same laws of morphological correlation,^ 

 and at the same classification of these dead forms as that which we 

 have reached from our study of the living Ones. He would not term 

 Lions and Tigers and Wolves " Carnivora," for he would not even 

 know that they eat anything, but he would assuredly form a group 

 with pretty nearly the same limits as the Carnivora, simply because 

 all these animals resemble one another, and differ from the rest 

 in certain peculiarities of dentition, &c. So, again, he would group 

 Oxen and Sheep and Deer together, because they present corre- 

 sponding coexistences of structure, though, knowing nothing of 

 their digestive processes, he would not call them " Ruminantia." 



And now, after our imaginary being had made himself acquainted 

 with the whole series of forms before him, and had established his 

 great laws of morphological correlation and his classification, suppose 

 that a mass of fragments of other creatures, more or less similar to 

 those which he had first familiarized himself with, were placed before 

 him, and he were desired to put these fragments together, and to 

 reconstruct these dismembered forms, how would he proceed ? Sup- 

 pose the first bone which came to hand very closely resembled the 

 jaw of a Deer, would he not naturally conclude — could he logically 

 escape the conclusion — that in all probability the skull and limbs 

 which belonged to this jaw were like those of a Deer also ? And 

 finally, supposing that, guided by this strong probability, he had 

 selected a complete deer skeleton from the mass, all of whose parts 



^ Except so far as he would be deprived of the advantage of the study of development. 

 This, however, obviously by no means interferes with the validity of the general argument. 



F F 2 



