( 307 ) 



The JamarysiiiUo bird has the foreneck tinged with pale yellow, and the 

 uj)per tail-coverts deep orange with dark red tips ; while in the Marraellos specimen 

 the former is white like the throat, and the upper tail-coverts are snlphur-yellow 

 w'ith the apical portion light orange-yellow. Similar variation is to be observed 

 in Peruvian birds. 



[343. Ramphastos osculans (ionld. 



Raiiijihaslos oscuhms Goukl, Proi: Ziol. So: Low/, in. l.-illj. p. I.^G (18:i;.— " in Bra-iilia," typ3 in 

 the Vienna Museum, coll. Natterer [cf. Mu/miji: Riunpli. 1st edit. pi. u], so. Rio Madeira: 

 Ribeinlo, Borba*) ; Pelzeln, I.e. p. 2.3-i (Ribeimo, Borba). 



Right bank : Borba, Rilieirao (Natterer). 



Three specimens, including the type, all from Borba, were kindly lent me 

 by the anthorities of the Vienna Museum. After comparing them with several 

 skins of R. culmiiuitas, I must confess that I am rather donbtful about the dis- 

 tinctness of 7?. oscidaii-i. The three examples of the latter differ only by having 

 the throat, foreneck, and cheeks more or less deeply yellow, while these parts in 

 li. culminatus are jmre white, sometimes with a pale yellow tinge on the lower 

 foreneck. lu other respects, notably in the shaj)e of the bill (culmen bounded 

 by a distinct furrow on each side), the two supposed " species " are practically 

 identical. Pelzeln (l.c. footnote 1) has already called attention to the great 

 variation in the coloration of the throat, etc., and upper tail-coverts, as shown by 

 Natterer's series. In the three skins before me the colour of the upper tail-coverts 

 varies from sulphur-yellow to scarlet with orange bases, and a similar variation 

 in /'. culminatuii is mentioned above. 



What R. osculans really is cannot be decided with the limited material at 

 our command. The so-called " o.'sei/laii.i'' from the "Rio Negro" and Muuduapo 

 (Orinoco; should also be comjiared with those from the Rio Madeira.] 



[344. Pteroglossus aracari aracai'i (Linn.). 



IluiiiphiiHlMS Amriiri Linnaeus, Si/st. Nal. x. p. 104 (175><. — based on Marcgrave's " Arat:ari," Hist. 



Nal. Bms. p. 217 : N.E. Brazil ; cf. Nov. Zont. .xii. iy05. p. :iOO). 

 Pterixjlosms Wieclii Pelzeln, l.c p. 23.0 (Borba). 



Right bank : Borba (Natterer). 



Specimens from Borba, as well as others from I'ara, Maranhao, and Itaituba 

 (Rio Tapaji')z), agree with South Brazilian exam[iles. The northern form, R. aracari 

 alr/'co/li.s (P. L. S. Mlill.) (aracari of Gould, Sturm, Sclater, etc.), with brnad black 

 cnlminal stripe, is apparently confined to the Guianas.] 



343. Pteroglossus castanotis australis (!ass. 



[Plcrnijlnsxiis CdSlinioOs Gould, I'lnr. Z'inl.Snr. I.,, ml. i. If^.Vi. p. 119 (1K34.— " Bi'asilia" ; we fix 



Rio SoUmni'iiH as type locality).] 



P. mslanotis b, australis Cassin, Pro:: Ani.il. Xal. Sa. Philad. 1807. p. 112 (18G7.— Rio Parana— 



Capt. Page coll.). 



No. 478. (? ad., Calama, 20. viii. 1907.— Wing 15.S ; tail 104 ; bill 133 mm. 



Nos. 451, 4S0. ? ?, Calama, 2r,, 29. viii. 1007.— Wing loS, 102; tail l.>], los; 

 bill 115, 125 mm. 



Nos. 091, 700. iS ad., S. Isabel, 13, 10. x. I'JOT.— Wing 100, 155; tail 157 ; 

 bill 140 mm. 



* licrlcpsch «; Llavleil (.\uc. /Cool. ix. I'.Wl. p. IW) enoncously substituted •' llio Negro" as type 

 locality, Gould, however, based his descriptiuu iipou Xattcrer's skins. 



