( 162) 



The type of " G. j/cn-viila" according to Darwiu, liad been collected on James 

 Island. Sbarj)e's locality, Chatham Island, for the same must therefore be 

 incorrect. Salvia describes and tignres specimens from Bindloe Island, which are 

 different, but cannot be called G. parnda. Ridgway (juotes as islands inhabited by 

 G. parndii, Chatham, James, Bindloe, Abingdon, but he is evidently not very 

 confident witli regard to the value of his G. parcula, saying that he thinks "it can 

 be demonstrated that the line between the two supposed species cannot be sharply 

 drawn." We have examined, of what we consider typical G./uliginosa : — 



22 skins from Charles Island. Wings, 01 — 65 mm. ; culmeu 12 — 14. 



104 from Chatham Island. Wings 60 — 66 mm. : culmeu 12 — 14'2. 



10 from Hood Island. Wings about 62 mm. ; cnlmen 12 — 14. No black- 

 billed adult male among them. 



46 from Albemarle Island. Wings, 60 — 65 mm. : calmen 13—14. 



11 from Narborough Island. Wings, 63 — 65 mm. ; cnlmen 12 — 13. The 

 wings average rather long, but the number of specimens is so much smaller than 

 from other islands that this of no importance. 



28 from James Island. Wings, 62 — 64-5 mm. : cnlmen 12 — 13-6. We are not 

 able to separate the .specimens from James Island in the least from typical 

 G. /ulii/i/wm. 



26 from Harrington Island. Wings, 60—65 mm. ; cnlmen 12 — 14. 



46 from Duncan Island. Wings, 62 — 65 mm. ; cnlmen, 12 — 14. 



21 from Indefatigable. Wings, 61 — 65 mm. ; cnlmen, 13 — 14. 



12 from Gardner (near Charles) Island. Wings, 60 — 65 mm. ; cnlmen, 13 — 14. 

 26 from Jervis Island. Wings, 00 — 67 mm. ; cnlmen, 13 — 14. 



We have thus examined over 350 specimens, but in the measurements very 

 young birds or such with the wing in moult are not included, as they may only 

 help to give a wrong impression. 



15. Geospiza fuliginosa minor subsp. nov. 



We have examined forty-three skins from Bindloe Island, with the wing 

 58 — 62 mm., the cnlmen, 11 — 13-5 : seventy-three from Abingdon Island, wing, 

 58 — 63 mm., cnlmen, 12 — 13'5. In addition to these small differences, however, 

 the bill is much slenderer and more compressed laterally. The number of adult 

 males which we were able to measure is very small, but the material is quite 

 suflicient to show beyond any doubt that the Abingdon and Bindloe form deserves 

 the rank of a subspecies. 



10. Geospiza acutirostris Ridgw. 



Geosjnsti acul'iroetris, Ridgway in Pr<K. U. S. Nat. Mua. XVII. p. 363 (1894) ; Ridgway, op. cil. XIX. 

 p. 531. PI. LVII. fig. 21 (1896). 



This species differs very distinctly from G./uliginosa and G./uliginosa minor 

 in its rather long, thin, straight and pointed bill. We have been able to examine 

 sixty-five specimens, including the type. In no other sj)ecies have we seen such a 

 proportion of black-billed adult black males, while black males with yellowish beaks 

 are very scarce. All the birds of the Harris' expedition were killed in December. 



The measurements of G. acutiro.Uris are given very correctly by Ridgway. 



This species is only known from Tower Island. 



