( l-"^--^ ) 



shorter bills, as the case may be; but names which are not to be nsed are, in onr 

 opinion, an unnecessary bnrden in snch cases. Onr reason for uniting Camarhi/nchus 

 with Gfiospiza is, that we see abont the same intergraclation in the form of the bills 

 of that supposed genns with the genus Qcospiza in Eidgway's sense, as we saw 

 between Cactnrnis and Geospiza in the old sense. We cannot ndmit the coloration 

 as a generic character, not only because in our case it would bring the yonng of 

 some of the species in another genus than the adult males, but we cannot see that, in 

 ornithology, it can be used as a generic character in any case, as the sexes and ages, 

 and sometimes seasonal plumages, differ so materially in many Ijirds, and abnormal 

 colorations, such as albinoes, are rather frequent. Mr. llidgway admits that he 

 "indeed finds himself quite unable to give precise characters for the genus 

 Camarhynchus" and that " the differences in the form of the bill presented by 

 Carnarhi/nchus, as defined by him, from Gcospiza, while perfectly obvious on com- 

 parison of specimens,* are extremely diflScnlt to describe, since they result chiefly 

 from variations of curvature in its outlines and relative proportions of various minor 

 details hardly susceptible of exact definition." This certainly does not sound very 

 convincing. Mr. Ridgway further admits that the transition from " Cactornis " 

 pa/l/f/tis to " Camdrhynchus " ])sitt<tciili(s is complete, but we find this to be the 

 same with " Cactornis" pal.l/f///s and Geospiza ; in fact, we think that "Cactornis'''' 

 pallidiis is still closer connected with Geospiza than with Camarlti/nckus in the old 

 sense. Ridgway's figures (in P. U.S. Nat. Mus. XIX. Plates XVI., XVII.) are 

 very instructive, and will explain our meaning to those who have no specimens 

 to refer to. See also our Plate VI. 



Mr. Harris makes the following observation : " The long-billed Geospizae (or 

 Cactornis') were observed to be more cactus-feeders than the thick-billed forms — 

 for example G. strcnua. Such Geospizae as strenua, jmchyrhyncha, and conirostris 

 were observed to be more seed-eaters than the slender-billed ones {Cactornis). The 

 smaller species, as fuliginosa, fratercula, and fortis were observed to frequent 

 the shores, of the more northern islands especially, in search of food, whereas the 

 slender-billed Geospizae, and those with very large beaks, were never seen feeding 

 on the rocky shores." 



The genus Geospiza is, in the Catalogue of Birds, Vol. XII., placed at the head 

 of the Fringillidae, followed by Chloris and separated by ten genera from Guiraca. 

 The South American members of the genus Guiraca, in Dr. Sharpe's sense, are, 

 nevertheless, the nearest relations to Geospiza, the form of the bill being indeed 

 very similar between the thick-billed Geospizae, such as Geospiza strcnua and 

 Guiraca cyanoidcs, especially the backward extension of the cnlmen (nasal bones in 

 skeleton) is very much the same in these genera ; while the wing is less pointed, 

 the feet larger and stronger, the tail very much shorter in Geopiza. The first 

 primary in Geospiza is rudimentary, not visible from below, and adapted to its 

 covert ; the second (first long) jirimary is shorter than the third. The resemblance 

 with certain thick-liilled Hawaiian finches, such as Telcspiza, cannot have any 

 serious consideration, as indicative of real relationship. 



Most of the species are well defined, others less, and some vary greatly in 

 dimensions and proportions. It is evident that adult black males are not equally 

 frequent in certain species. While black individuals are surprisingly rare in many 

 species, they are very frequent in Geospiza conirostris Uidgw. There can be no 

 doubt, we think, that all the black specimens are adult males. In the members of 



* We should think uot in the case of " CactcrnU '' paUUhm 



