58 ESSAY ON CLASSIFICATION. Part I. 



kind of investigation can hardly be overrated ; and it would be highly desirable that 

 naturalists should turn again their attention that way, now that comparative anatomy 

 and physiology, as well as embryology, may suggest so many new toj^ics of inquiry, 

 and the progress of physical geography has laid such a broad foundation for 

 researches of this kind. Then we may learn with more precision, how far the 

 sjjecies described from isolated specimens are founded in nature, or how far they 

 may be only a particular stage of growth of other species; then we shall know, 

 what is yet too little noticed, how extensive the range of variations is among ani- 

 mals, observed in their wild state, or rather how much individuality there is in each 

 and all living beings. So marked, indeed, is this individuality in many families, — and 

 that of Turtles affords a striking example of this kind, — that correct descriptions of 

 species can hardly be drawn from isolated specimens, as is constantly attempted to 

 be done. I have seen hundreds of specimens of some of our Chelonians, among 

 which there were not two identical. And truly, the limits of this variability con- 

 stitutes one of the most important characters of many species ; and without precise 

 information upon this point for every genus, it will never be possible to have a 

 solid basis for the distinction of species. Some of the most perplexing questions 

 in Zoology and Paleontology might long ago have been settled, had we had more 

 precise information upon this j)oint, and were it better known how unequal in this 

 respect different groups of the animal kingdom are, when compared with one 

 another. While the individuals of some species seem all different, and might be 

 described as different species, if seen isolated or obtained from different regions, those 

 of other species appear all as cast m one and the same mould. It must be, there- 

 fore, at once obvious, how different the results of the comparison of one faima with 

 another may be, if the species of one have been studied accurately for a long 

 period by resident naturalists, and the other is known only from specmiens collected 

 by chance travellers; or, if the fossil representatives of one period are compared 

 Avith living animals, without Ijoth faunae having first been revised according to the 

 same standard.-' 



Another deficiency, in most works relating to the habits of animals, consists in 

 the absence of general views and of comparisons. We do not learn from them, 

 how far animals related jjy their structure are similar in their habits, and how far 



^ In this respect, I would remark that most of Such cases should be altogether rejected in the 



the cases, in which specific identity has been affirmed investigation of general questions, involving funda- 



between living and fossil species, or between the mental principles, as are untrustworthy observations 



fossils of difierent geological periods, belong to always in other departments of science. Compare 



families which present either great similarity or further, my paper upon the primitive diversity and 



extraordinary variability, and in which the limits of number of animals, quoted above, in which this 



species are, therefore, very difficult to establish. point is specially considered. 



