( 159 ) 



72 specimens from Hood Island, together with one from Gardner near Hood, 

 there does not remain a shadow of doubt that they are one somewhat variable 

 species. This species is characterised by its bill being mnch more elongated and 

 narrower than in G. strenua, and by the prevalence of blackish colour in the fem'iles 

 and young birds, which are much darker than in 0. strenua, being brownish black 

 above and below, on the underside with whitish edges from the chest downwards. 

 The culmen varies in males from 24 to 18*5 mm., the basal width of maudilile from 

 13 to 9 mm., the height at base from 18 to 14 mm., the gonys from 13-5 to 

 10'5 mm. 



5. Geospiza conirostris brevirostris Ridgw. 



Cactomis hremroslrh, Ridgway in Proc. U.S. Nat. Jfus. Vol. XII. p. 108, Fig. 4 (1890). 

 Geospiea brevirostrii, Ridgway op.c. XIX. p. 541. 



In 1800 Mr. Ridgway described from a single immature specimen of the 

 " Albatross'" expedition as new a finch which he called Cactomis brevirostris, which 

 in 1896 he also places among the Cactomis group of Oeosjnza, comparing it with 

 Geospiza barringtoni. After a close examination of the figure and onr one skin 

 from Gardner Island (near Charles Island), we are forced to remove it altogether 

 from the Cactornis grouji, it being barely distinguishable from the smaller-billed 

 specimens (= media Ridgway cf. antea) of Geospiza conirostris from Hood Island. 



There is some uncertainty about the locality of the type specimen, as Ridgway 

 in both 1890 and 1890 first quotes it as coming from Chatham Island, and then 

 several times as being found on Charles Island, again on p. 512 in Vol. XIX. of 

 the Washington periodical he omits it on Charles (resp. Chatham) and places the 

 number for it by mistake on James Island. A\'^e think the real home must be 

 Charles, because we have one specimen which we take to belong to brevirostris from 

 Gardner (near f Uiarles), which has the same fauna as the larger T'harles Island. 



The only difference we can see between G. conirostris conirostris and G. coni- 

 rostris brevirostris is that the bill of the latter is slightly smaller and narrower. 



We possess only one probably adult female from Gardner Island (near Charles 

 Island), collected by the recent expedition. 



(We have one young Geospiza from Indefatigable which agrees with Ridgway's 

 description of Dr. Habel's specimen from the same island. These birds are too 

 light to belong to G. conirostris brevirostris, and all the measurements are distinctly 

 smaller. 



This form, unfortunately, cannot finally be named or described, on account of 

 the absence of adult s])ecimens in collections, as far as we know ; but it will probably 

 be an unnamed subspecies of G. conirostris. See also Salvin's remark in Trans. 

 Zool. Soc. Lond. IX. p. 481 under G.fortis.) 



6. Geospiza conirostris propinqua Ridgw. 



Gcos2)iza propimiua, Ridgway in Prur. U.S. Nat.Mns. XVU. p. 3G1 (1894) and .7). cil. XJX. p. 543 

 (1896). 



Described from Tower Island, whence we have a series from Dr. Banr, 

 including the type and a good many from the new expedition, among them a fair 

 number of adult males. They agree so well with G. conirostris conirostris, that 

 one has to examine them very closely to discover the points of difference in the 



