No.lUG. PROCEEDINGS OF TEE NATIONAL MUSEUM. 467 



gest relationship in structural characters, the form of the bill in Loxi- 

 oides, Telespiza,^ and Psittirostra being not very dissimilar in character 

 to that of some species of CamarJiynchus. 



(5) Nesopelia. This genus is closely related to the genus Zenaida, of 

 general Neotropical distribution. In fact, it is doubtfully distinct from 

 the latter. 



To sum up : Of the five peculiar Galapagoan genera of birds, only 

 two {Nesomimus and XesopeUa) are of evident American relationship. 

 The remaining three have so obvious a leaning toward certain Hawaiian 

 dicieidine forms' that the possibility of a former land connection, 

 either continuous or by means of intermediate islands as " stepping 

 stones," becomes a factor in the problem. It may be that the resemblance 

 of Cocornis, " Cactornis,''^ and Camarhynchus to the above-mentioned 

 Hawaiian forms is merely a superficial one, and not indicative of real 

 relationship. I do not by any means claim, on the strength of such 

 evidence, a common origin for them, but merely present the facts as 

 "food for retlection." ' 



It will doubtless seem to some that I have gone to an undesirable if 

 not reprehensible extreme in naming so many forms of the genera Cer- 

 thidea, Geospiza. Camarhynchus, and Pyrocephalus. Whether such is 

 true or not, I have certainly not been actuated by any desire to add to 

 the number of species. On the contrary, several names, chiefly of my 

 own, have been relegated to synonymy in consequence of what seemed 

 to be good evidence of their untenability; and in naming new ones I 

 have in all cases been guided by definite principles without regard to 

 the character of the criticism which might result. Some of these new 

 names may, when additional material has been secured, i)rove also to 

 be untenable, and will then have to be "degraded;" but the decision 

 of such questions should always be a matter of evidence, never of indi- 

 vidual opinion or prejudice; and I am sure that all who have had equal 

 experience in the laborious and time-consuming task of dissecting and 

 reconstructing synonymies will bear me witness that the real promoter 



'At least what I take to be Telespiza flavissima, Rothschild, but, not being able to 

 refer to the ' ' Avifauna of Lay san," the identification is doubtful. The bird was taken 

 ou the island of Laysan by Mr. W. T. Brigham, and is No. 128455, U.S.N. M. 



^ Whether the geuera Loxioides, Telespiza, Psittirostra, and other linchlike Hawaiian 

 forms are true Friugillid;c or thick-billed Dicseidte can scarcely be said to be yet 

 satisfactorily decided. 



■^ While the prevailing fades of the Hawaiian avifauna is unquestionably Poly- 

 nesian, a small but by no means insignificant American element is present. How it 

 came there has not yet been explained. Of land birds, the jjeculiar genus Pha'ornis 

 is most like Myadestes, of America (see Stejueger, Proc. U; 8. Nat. Mus., X, 1887, p. 

 92, and XII, 1889, pp. 383,384). The American C/rc^/s /( (((Zso«i»s occurs also in Hawaii; 

 the Hawaiian GalUnuJa is barely separable from the American species, G. galeata, 

 and the Plegadis seems to be identical with P. giiarauna. The Hhnantopus is nearly 

 related to H. niejcicanus, and Anas wyvilUana is very similar to A. aberti of western 

 Mexico. Two of the Hawaiian Procellariida*, JEstrelata phwopygia and Oceanodroma 

 cryptoleucnra occur also in the Galapagos ! 



