188 ESSAY ON CLASSIFICATION 



is concerned; that all the members of the same order are homologous 

 in the complication of their structure; that all the representatives of 

 the same family are homologous in form; that the different genera 

 of one and the same family exhibit homologous peculiarities in the 

 details of their structure; and that even within the narrow limits of 

 species we may still trace homologous features, among the genera 

 which have numerous representatives, even when such resemblances 

 do not extend to the species of closely allied genera. It is plain from 

 this that the categories of homology are as numerous and diversified 

 as the essential kinds of differences which we may trace in the struc- 

 ture of animals; or, in other words, we have branch homologies, class 

 homologies, ordinal homologies, family homologies, and specific ho- 

 mologies. Examples of the more comprehensive kinds of these homol- 

 ogies will occur to every practical geologist. As to specific homologies, 

 they are particularly traceable in those structural features which de- 

 termine the proportions among the parts; as, for instance, when all 

 the species of one genus are either long-necked, short-tailed, long- 

 legged, etc, while those of closely allied genera may present reverse 

 proportions, etc. 



Let us now see what are the categories of analogy and how far it 

 is possible, under all circumstances, to distinguish homological and 

 analogical features. If analogy is a resemblance arising from a com- 

 bination of features characteristic of one group with those charac- 

 teristic of another group (such as class characters of one class with 

 those of another class, or those of families of another class), then the 

 investigation will only require the recognition of the different cate- 

 gories of structure already considered (such as branches, classes, or- 

 ders, etc.), and a correct appreciation of the mode of their combina- 

 tion with those of another group. It will, for instance, be sufficient 

 to ascertain in what manner the features resulting from a certain 

 mode of execution of the homologies of one type are combined with 

 structures of another type; or, in other words, to recognize any fea- 

 ture wherever it appears, and not merely within the limits within 

 which structures are strictly homologous. The study of analogies is 

 therefore limited to the investigation of more or less distinct features 

 that are naturally characteristic of one kind of group, in their com- 

 bination with features of groups of another kind. For instance, the 

 similarity between an insect wing and the wing of a bird is based 



