XXIX] MY FRIENDS AND WORK, 1871-1886 95 



systematic arrangement and uniformity of treatment through- 

 out. But here the immense extent of the subject, the over- 

 whehning mass of detail, and above all the excessive diversities 

 in the amount of knowledge of the different classes of animals, 

 rendered it quite impossible to treat all alike. My pre- 

 liminary studies had already satisfied me that it was quite 

 useless to attempt to found any conclusions on those groups 

 which were comparatively little known, either as regards the 

 proportion of species collected and described, or as regards 

 their systematic classification. It was also clear that as the 

 present distribution of animals is necessarily due to their past 

 distribution, the greatest importance must be given to those 

 groups whose fossil remains in the more recent strata are the 

 most abundant and the best known. These considerations 

 led me to limit my work in its detailed systematic ground- 

 work, and study of the principles and laws of distribution, to 

 the mammalia and birds, and to apply the principles thus 

 arrived at to an explanation of the distribution of other 

 groups, such as reptiles, fresh-water fishes, land and fresh- 

 water shells, and the best-known insect-orders. 



There remained another fundamental point to consider. 

 Geographical distribution in its practical applications and 

 interest, both to students and the general reader, consists of 

 two distinct divisions, or rather, perhaps, may be looked at 

 from two points of view. In the first of these we divide the 

 earth into regions and subregions, study the causes which 

 have led to the differences in their animal productions, give a 

 general account of these, with the amount of resemblance to 

 and difference from other regions ; and we may also give lists 

 of the families and genera inhabiting each, with indications 

 as to which are peculiar and which are also found in adjacent 

 regions. This aspect of the study I term zoological geography, 

 and it is that which would be of most interest to the resident 

 or travelling naturalist, as it would give him, in the most 

 direct and compact form, an indication of the numbers and 

 kinds of animals he might expect to meet with. 



But a large number of students now limit themselves to 

 a study of one of the classes, or even orders, of the higher 



