( 390 ) 



Another apparent difference of adult (t. c. darwini is that the under tail-coverts 

 have verv wide Imffy white borders, and that the lower atidnmen has the feathers 

 edged with gre}-. As two of onr iiKdcx have perfectly lilack bills, we cannot believe 

 that these colour differences are due to young age. 



We think that the young Iiirds formerly recorded as G. coi/irostiix subsp. ? (I.e.) 

 are certainly young Jdnciiu. 



Geospiza conirostris propinqua llidgw. 



Nov. ZiioL. VI, 1899, p. I.'i9.— '.r'. inopniqifi Ridgw., liniU .\<,rlh nml Middle Am,r. I, 1901, ji, 499. 



No fresh material, as neither Beck nor (ireen collected on Tower Island. Tliis 

 form can generally be distinguished without difficulty from typical conirontrix by 

 its more elongated iind less high bill ; but in some specimens the difference in 

 height is barely perceptible. 



Geospiza conirostris brevirostris Uidgw. 



Nov. Zool,, YI, 1899, p. 159. — G. conirostris subsip. ibidem, p. 159 note under no. 5, p. 21.11 no. 31 ; 

 Geoxj,i-.a bnrirostris Ridgw., Birih North and Middle Aiiirr. I, 1901, p. 514 (Charles, 

 (?) Indefatigable. 



We have not received any fresli material of this interesting form, but liave once 

 more gone into the question of its position. J\Ir. Ridgway states that this bird 

 belongs to the slender-billed group formerly called ('(ictoniix. All these so-called 

 Cactornis have the bill longer and much more slender ; and, moreover, in no case do 

 we tind that more than one slender-billed form occurs on one island, and on Charles 

 Island we already have (i. scanc/rns iiiteriHedia Ridgsv. It is true, as Mr. Riilgway 

 says, that we have not seen the type, but we have a specimen from Charles Island, 

 the typical locality of this form, which in our opinion agrees with the drawing and 

 measurements given by Mr. Ridgway. We have, moreover, a number of young 

 G. colli rosin's ronirostris and (lanrini, which in the shape and size of their bills 

 are almost indistinguishable from <uu' (Jharles Island bird, which we consider to be 

 brevirostris. 



It may be retorted that we have recorded our specimen from " Gardner Island 

 near Charles Island," and that true G. conirostris was recorded by Baur from 

 Gardner Island, and so an error might have arisen ; but we wish once move to 

 emphasise that Baur never visited Gardner Island near Charles, and the Webster- 

 Harris expedition never lauded on Gardner near Hood, so that an error is impossible 

 (cf. Nov. ZooL. VI, 1809, p. 144). 



Mr. Ridgway is inclined to unite an Indeftitigable specimen with brevirostris, 

 aud we also have one {ride Nov. Zooi,. VI, p. l-'iO) which seems hardly sejiarable 

 from it. 



Geospiza dubia dubia Gould. 



Nov. Zool. VI, 1899, p. \(,0.^(iaixpi-j, dubia Ridgw. B^rt/.s' North n„d Middle ,-lm<T.I (1901), p. 501. 



We have received a number from Chatham Island which agree with the types, 

 and oue/emale from Indefatigable, 16. ii. 19(Jl, R. H. Beck coll. 



The birds we united with this form in our former article, and which came from 

 Duncan Island, agree certainly in the size of the bill, but are ijuite young birds, 

 their bills being as large as our old (/ii/jia dubia. It is therefore quite possible that 



