Family and Generic Differentiation. 343 



an unlikely eventuality, so fine are the distinctions now 

 drawn between trivial variations in the structure of the bill 

 and other organs. If, on the other hand, colour-pattern is not 

 ignored, genera-splitting is far less likely to do harm, and 

 indeed is likely to be productive of much good, for we shall 

 have got down to small groups of natural and genetically 

 related species. These minor generic groups would in fact 

 in most cases be found to consist of analytical varieties 

 grouped around some central or dominant specific type. 

 They would be really natural units which when integrated 

 with others into larger and still natural groups (super- 

 genera or what not) woukl express at a glance the phylo- 

 genetic natural relationships of the particular family or 

 subfamily we were dealing with. 



My point, therefore, is that while disintegrating within 

 justifiable and natural limits we should at the same time 

 integrate on the above lines. 



Take, for example, the Redshank association again. 

 In this group, so variable are the structural features of the 

 bill and other anatomical features that almost every species 

 could conceivably be made the type of a distinct genus; and 

 the same might be said of the Dunlin association. Colour- 

 pattern saves the situation. 



Take, again, the case of the Ground-Finches of the 

 Galapagos, famous for their association with Darwin's 

 original conception of the origin of species ! Mr. Roth- 

 schild, in conjunction with Dr. Hartert in a recent review 

 of these Finches, has integrated them into one genus, 

 Geospiza, sinking the three other genera (^Cactornis, Cama- 

 rhynchus, and Platyspiza) on account of the complete series 

 of intergradations in the form of the bill which they allege 

 to exist. Incidentally they hold that there are 34 species 

 and subspecies of these Finches, while Mr. Ridgway main- 

 tains 35 — in spite of the fact that there are only some 

 14 islands in the whole group, and that Duncan Island, 

 which has a superficial area of some 12 square miles only, is 

 found to contain no less than 10 of these species. 



To me, with my limited experience, such a state of things 

 seems to be rather shattering to one's preconceived ideas 



