VALUE OF COLOR CHARACTERS IN BIRDS. 315 



place without change in color pattern. Were there as many intermediate forms 

 among these cuckoos as there are in the genus Geospiza, they would of necessity 

 be retained in one genus and they might still be, with as much right as are the 

 three color types in the genus Merops, if we were accustomed to regard modi- 

 fication of bill and variation of color pattern as characters of equal importance 

 As a matter of fact, however, ornithologists do not regard style of coloration as 

 a generic character unless it emphasizes some tangible structural character of 

 bill or feet or some difference in the relative proportions of remiges or rectricee. 

 Color plays a most important part in the differentiation of species and subspecies 

 and since we must admit that the genera of birds are groups of very much lower 

 rank than genera in other classes of the animal kingdom there seems no reason 

 why more weight should not be given to color characters in the study of generic 

 differences and phylogeny. Furthermore the tendency is constantly toward a 

 still greater refinement of genera, differences are being magnified and resemblances 

 neglected, and search is always being made for slight so-called structural differ- 

 ences and not for characters which will bind several genera together into one 

 genetic phylum. It seems high time that more attention were given to this 

 phase of the study and I am inclined to think that in this connection a careful 

 study of coloration will aid materially; giving us not only an idea of the inter- 

 relationship of the species of a genus but also a clue to the phylogeny of the 

 genera. 



It seems probable also that certain species for which a distinct genus has 

 been established upon trivial characters, such as the comparative length of the 

 tail feathers, are really more closely allied to similarly colored species of an allied 

 genus, than they are to the differently colored species with which they are arbi- 

 trarily associated. 



Such a case we find in the Meropidae. The family comprises some forty-one 

 species or subspecies. Three of them stand distinct from the rest both as regards 

 coloration and structure — Meropogon forsteni and the two species of Nyctiornis. 

 The other thirty-eight are closely related to one another and while they exhibit 

 remarkable diversity of coloration, they present very slight structural differences. 

 In the attempts to subdivide the group the only character that systematists have 

 been able to find is the relative length of the tail feathers. In the genus Dicro- 

 cercus the tail is deeply forked, in the fifteen species of Melittophagus it is square, 

 while in the twenty-one species of Merops the central pair of rectrices are narrowed 

 and project beyond the others. 



Dicrocercus is essentially monotypic and confined to Africa. Melittophagus 

 is also African with the exception of two species which occur respectively in 

 Java and in the Indo-Malay region. Merops on the other hand occurs through- 

 out Africa as well as in southern Europe, Arabia, the Indo-Malay region, and 

 Australia. In studying the coloration of these genera we find that Dicrocercus 

 closely approaches one of the African groups while the two other genera resolve 

 themselves into several groups. The red-throated species of Melittophagus are 



