7G2 bitllp:tin oo, united states national museum. 



hoi'izontally oviil or subcuiK^iite, witli ni('ni])i'an(> abovo, behind, and 

 below, but ))roadest above. Rictal l)ristle.s obsolete. Wino- rather 

 short, rounded (seventh, sixth, and fifth i)riniaries longest and nearly 

 equal, the eij^hth and fourth l)ut little shorter, ninth not longer than 

 second); wing-tip shorter than length of culmen. Tail short (less than 

 twice as long as tarsus but more than two-thirds as long as wing), 

 rounded, the rectrices l)r<)ad and rounded at tip. Tarsus long (about 

 twice as long as exposed culmen, much more than one-third as long as 

 wing), slender, its scutella indistinct (sometimes obsolete laterally) ; mid- 

 dle toe, with claw, decidedly shorter than tarsus, its basal phalanx united 

 for more than half its length to outer toe, for less than half to inner; 

 lateral toes with claws reaching about to liase of middle claw; hallux 

 about as long as lateral toes but conspicuously stouter, its claw decidedly 

 shorter than the digit. 



Colo rat! on. — Plain brownish gray, grayish brown, or olive above 

 (the wings with narrow' whitish bands in one species); under parts 

 much paler than the upper, som<'times dull white, the throat and a 

 superciliary streak sometimes butfy or I'ufescent. 



Nl<llJic<(tlov . — Unknown. 



R(mgv. — Galapagos Archipelago, where rei)resented on all the 

 islands of the group. 



I have long been convinced that Cert It idea l)elonged to the Mniotil- 

 tida? rather than the Ccerebida% where it had been placed l)y Messrs. 

 Sclater and Salvin. This view of its relationships has bec^i confirmed 

 by an examination of its anatomical structure, made at my suggestion 

 by Mr. F. A. Lucas.' 



Owing to their extremely plain coloration, it is very difiicult to con- 

 struct a key to the species of this genus, a difficulty greatly enhancc^d 

 by the fact that I have at the present time only four of the nine known 

 forms before me, while two of the remaining five I have not seen at 

 all. The following attempt is therefore far from satisfactory, but may 

 assist somewhat in the identification of the various forms. 



KKY TO I'llE SI'KCIES OF CEKTIIIDEA. 



o. No wliitisli wing-l)ar.«. 



h. Adult males with throat and supt'cciliafy streak ochraceous-buff or tawny, 

 c. Darlcer and more olive al)ove; rectrices very narrowly tipped with pale 

 Ijrownish. 

 d. Less olive a])ove, ]>aler and less olive below; hill never hlaek. (.Tervis, 

 Narborough, .Tames, fndefatii^'ahle, Albemarle, and Dnnean islands.) 



Certhidea olivacea ( p. 7(33) 

 dd. More olive al)ove, darker and more olive below; bill often blark. ( Chatham 



Island.) Certhidea luteola (p. 7(14) 



cc. Paler and more grayish; rectrices rather broadly tipped with whitish. 

 ( Charles Island. ) Certhidea ridgwayi ( p. 765 ) 



' See Lucas, The Anatomy and Affinities of Certhidia (.5?c); Proc. U. S. Nat. Mus., 

 xvii, 1894, 309, 310. 



