(277 ) 



frontal l]iiii(l and the wbole lower surface of a much brighter yellow. Otherwise, 

 it agrees iu colour and size with F. c. brasilienxis. On examining the series in 

 the British Museum, I find that the two specimens from Pernambuco (coll. 

 W. A. Forbes) have the black frontal band, which is altogether wanting in the 

 ten other exaruides from Bahia and Hio, and the lower parts rather brighter 

 than the latter. I propose, therefore, to separate the form of Pernambuco as 



Pitylus canadensis frontalis n. subsp. 



Similis subspeciei P. ,-. brasiUcn.sis dictae, sed vitta fioutali nigra et colore 

 subtus distiucte clariore facile distinguendus. 



Tyjie in Mns. Tring, No. 1742. A. Robert coll. " ? " ad. S. Lourenco, 

 Pernambuco, 28 to m metr. elev., July 29th, 19(1.3. Wing 94, tail SI, bill UU mm. 



llalj. Pernambuco, N.E. Brazil. 



21. *Guiraca rothschildii Bartl. 



Guiraca rnllisch/ldii E. Bartlett, Ann. Mar/. Xat. llisl. (G) vi. (Aug. 18'J0) p. Iij8 (River 

 Carimang, Brit. Guiana. Types now in Tring Museum, examined). 



(?. cyaiiea (uec Linni' I) Chapman and Riker, Auk vii. (1890) p. '2(')8 (Santarem). 



CyanocomjisH ci/anohlrx (nee Latresnaye .') Ridgway, Bull. I'.S. Nat. Mas. No. 50 (" Bird.s of 

 North and Middle America," Part I.) (1901) p. 509 (Brit. Guiana to Lower Amazon). 



Guiraca ci/anea roihsrhildi Berlepsch and Hartert, Xoi\ Zuol. ix. (1902) p. 24 (Orinoco regiou). 



One adult ?, taken March 4, 1904. No. 2U13. "Iris brun." 



In the shape of the bill and coloration this specimen agrees exactly with 

 the type female from River Carimang, and is, of course, very different from the 

 female of <t. ci/uiien. Both the upper and lower surface are very much darker 

 than in the latter bird, and the bill is altogether ditferent, being much longer 

 and straighter. 



Some time ago I examined iu C!ount Berlepsch's collection one male, collected 

 by Mr. W. A. Schulz near Para. It agrees perfectly with examples from British 

 Guiana and N.E. Pern. Although recorded from Santarem, this is, I believe, the 

 first record for Para. 



G. rotlisclnldii seems to me to be decidedly distinct from G. ci/anea, and in 

 my opinion it is but the southern representative of G. concreta ci/anoidcs, with 

 which it agrees iu the shape of the bill. The females of both forms, too, are 

 very much alike, that of G. r. cyanoides being only distinguishable by its rather 

 brighter rusty-brown color.ation. On the other hand, G. ci/anea has a much 

 shorter and much more curved bill, and the female is very different, as said 

 above. Moreover, in the vicinity of Puerto C'abello, N.W. Venezuela, there occurs, 

 side by side with G. ri/anea, a slightly modified form of G. rothschildii which 

 seems to be an intermediate link between the latter and G. c. cyanoides. These 

 facts seem to point towards specific distinctness of G. cyanea and G. rothschildii. 



The latter is the species to which Mr. Ridgway — quite erroneously— ajiplied 

 the name "cyanoides." M. de Lufresnaye, when describing his "Corcoborus 

 cyanoides," gave, in the Latin diagnosis, only the characters of the " female or 

 junior avis," which,, consequently, must be regarded as the actual type. It has 

 no bearing whatever on the case that the bird suj)poscd by Lafrc^suaye to be the 

 male of G. cyanoides, now turns out to belong to G. roth>ichil<lii. Tiierefore 

 Mr. Ridgway was quite in error iu considering the male specimen in the Boston 

 Society collection as Lafresnaye's type. A glance at the description of the latter 



