VALUE OF COLOR CHARACTERS IN BIRDS. 317 



genetic, and that the species are better left in one genus, associated according to 

 color characters even though we confess an inability to arrange them all in definite 

 color groups. 



The Alcedinidce furnish an exceedingly interesting series of color types, with 

 comparatively little structural differentiation. The widespread genus Ceryle 

 with a generally admitted distinctive form of bill, exhibits several quite different 

 types of coloration in Africa, India, Japan, and America respectively, but all 

 of them different from anything found among the remaining species of the 

 family. 



Other so-called genera are much less satisfactory. Alcyone and Ceyx are 

 established upon the fact of their having but three toes instead of four. The 

 species of Alcyone ranging from northern Australia to New Guinea and adjacent 

 islands have the upper parts plain blue above, neither banded nor spotted. 

 Ceyx on the other hand presents several types of coloration, many species being 

 tawny above with a violet wash and little or no blue, while others are largely 

 black or blue above. C. cyanipectus and its allies would unquestionably be 

 regarded as members of the genus Alcedo were it not for the reduction in the 

 number of toes. Indeed C. cyanipectus is a perfect counterpart in coloration and 

 bill structure of Alcedo moluccana. In the same way a little tawny kingfisher 

 of Madagascar, a typical Ceyx in coloration, is placed in the differently colored 

 African genus Ispidina on account of the presence of the fourth t oe. (PI. XXV 1 1 , 

 fig. 9.) No authors seem to claim any close relationship between Alcyone and 

 Ceyx by virtue of the suppression of one of the toes and the character docs 

 not seem to be a very important one, therefore I see no reason why Ceyx 

 cyanipectus should not be regarded as a three-toed offshoot of the Alcedo stock 



or that Ispidina madagascariensis (PI. XXVII, fig. 9) may not be quite as 



closely related phylogenetically to Ceyx (PL XXVII, fig. 8) as to Ispidina. 

 The species of the genus Halcyon present such slight structural variations that 

 efforts at generic subdivision have proven futile. It is the grand residuum 

 after a few well marked groups have been removed; such as Dacelo f Todi- 

 ramphus, Syma, Tanysiptera y etc. These seem to be perfectly natural asso- 

 ciations of species but even so, the coloration of some of them is so strikingly 

 like certain species of Halcyon that one wonders to which genus the latter are 

 more closely allied. Dacelo cervina for example is a perfect duplicate of Halcyon 

 chelicuti and Todiramphus recurvirostris of H. sanctus while species of Pelargopsis 

 follow so closely the pattern and tints of H. cinnamomeus that one wonders after 

 all whether the resemblance in bill structure between Pelargopsis and Ceryk is 



not merely parallelism, and whether the former is not more nearly allied to 

 Halcyon. 



It is not the aim of this paper to propose any changes in classification or 

 nomenclature suggested by the examples which have been cited. They are 

 presented simply as suggestions. 



It is quite possible that color resemblances do not necessarily denote close 



