( 163 ) 

 IT. Geospiza dentirostris Gould. 



Gmspizn dentirndris, Gould in P. Ziml. Soc. Loud. 1837, p. 6 : Znol. fleaifh-^ III. Birds, p. 102 

 (1841) ; Salvia in Trims. Zool. Snc Luml. IX. p. 483 (partim) ; Sharpe, Cat. li. Brit. Mux. XII. 

 p. 1 1 ; Ridgway in rroc. U.S. Nat. .Vm. XIX. p. 532. 



There are iu the British Museum si male and & female, which are identified by 

 Dr. Sharpe as (j. dentirostris. One is evidently the type. It is in our opinion an 

 immature l)ird, probal)ly a, female. Its culmen measures 14'7 mm.; the hill 

 from nostril to ti]), IU'8; wing, OS. The other specimen is a black male, and 

 marked on its original label "Charles Island." Its culmen measures 14'8 mm.; 

 the bill from nostril to tip, !)'8. The former specimen, the type, has the " tooth," 

 from which it has received its name, much in front of tlie middle of the cutting edge 

 of the upper jaw, only a few millimetres from the tip, and the " tooth " is well visible 

 from above. In Admiral Markham's bird, however, the " tooth " is exactly in the 

 middle of the cutting edge of the upper jaw, and not to be seen from above. The 

 type has no exact locality marked, and no original label. Except for the presence 

 of the " tooth " it might almost be an immature (J. forti.s. 



It is curious that neither Mr. Townsend, Messrs. Baur & Adams, nor Harris' 

 expedition came across such a bird, and the dissimilai'ity of the two only known 

 skins in the British Mnsenm forces one to suspect that they might be aberrations ; 

 but we must await further observations before coming to a definite conclusion 

 about it. 



IS. Geospiza spec. inc. 



We have one black male of a Geosjnza, skinned from a bird in spirits brought 

 home by Dr. Baur, from Chatham Island, whore it had 1 leen killed on September 8th, 

 1891. Its bill is almost but not (juite black. Its upper jaw has no " tooth," but in 

 dimensions this bird agrees wonderfully with (J. dentirostris. The culmen measures 

 14-.5 mm.; the bill from nostril to tip, '.»-8 ; the wing, 69. Without further evidence 

 we are not able to say whether this species is the same as G. dentirostris, whether 

 it is a h3'bri(l between one of the larger and one o'f the smaller species of Geospiza, 

 or whether it belongs to a hitherto unknown and unnamed species. 



19. Geospiza difficilis Sharpe. 



Geospiza dentirostris partim (non Gould !) Salvin in Trans. Zool. Soc. IX. p. 48.S. 

 Geospiza difficilis Sharpe, Cat. B. Brit. Mus. XII. p. 12 ; Ridgway in Proc. U.S. Nat. Mus. XIX. 

 p. 532.' 



The colour of this species is generally very dark, and very brownish. Not a 

 single specimen has the abdomen whitish. The bill has no " tooth," the culmen is 

 remarkably elevated and arched near the base, and depressed again in front of the 

 nostril. The length of the culmen in our old males — though none of them has 

 a perfectly black bill — is 14 — 14-8 mm.; the wing, 63 — 64. The iris is brown. 

 ^Ve have a large series from Abingdon Island collected in August. The only 

 authentic locality for this species is Abingdon Island. The locality Charles Island 

 for a skin from Markham must be erroneous, such a distribution being unheard of. 



2it. Geospiza debilirostris Ridgw. 



Geospiza dehilirontris, Ridgway in Proc. U.S. Nat. Mitn. XVII. p. 363 (1894) ; XIX. p. 533. 



Messrs. Baur & Adams collected several of this species iu spirits, and our 

 recent collection contains a large series. Most of our specimens show also a 



