1917.] Chapman, Distribution of Bird-life in Colombia. 245 



The problem presented by our large series of Rupornis magnirostris is 

 so puzzling that I have assembled a large series of the northern forms of this 

 species in order to determine its range of individual and geographic variation. 



Primarily, it was essential to ascertain what were the characteristics of 

 R. m. ruficauda and to what area this race was restricted. In describing it, 

 Sclater gave no definite type-locality and although the first place named 

 under the "Habitat" assigned to the new form is Cordova, Mexico, it is 

 evident from his description and subsequent figure in Exotic Ornithology 

 (pi. 88) that among his specimens the characters on which he based the form 

 were typically developed only in those from western Panama, whence I 

 have a series of eleven specimens. From the localities named by Sclater I 

 therefore suggest David, as an appropriate type-locality for this form. 



If this suggestion be accepted, our specimens from Boqueron, near David, 

 are essentially topotypical. All have the interspaces in the rectrices largely 

 or wholly rich tawny or reddish brown, and in the adult, the back and breast 

 are mouse-gray, much as they are in true magnirostris. 



Aside from its reddish tail, ruficauda differs from magnirostris in its 

 buffy or ochraceous underparts, under wing-coverts and upper tail-coverts, 

 the bars on which are much brighter, more tawny than in magnirostris. 



Whether all western Panama specimens have the tail as fully reddish 

 brown as in those I have examined cannot be stated. Of five specimens 

 from the Canal Zone three are essentially like those from Boqueron, two 

 have much more gray in the tail which, however, still shows more or less 

 tawny. Two specimens, one each from Coiba and Iguaro Islands, have 

 the interspaces wholly tawny. A male from Marraganti, 150 miles east of 

 Panama City has about as much gray and tawny in the tail as the two Zone 

 specimens above mentioned, with which, in other respects it also agrees. 

 Although these three birds have not the wholly tawny and black tail of 

 Boqueron specimens, they agree with them in possessing the other char- 

 acters which distinguish typical ruficauda; that is, a back and breast which 

 are essentially hke those of magnirostris, in combination with buffy or 

 ochraceous underparts, etc. barred with reddish brown. 



Turning now to the west, a fine series of seventeen specimens from Costa 

 Rica makes it possible to learn with some certainty the relationships of the 

 form inhabiting that country. In only three does the amount of red in the 

 tail compare with that shown by the Boqueron specimens. The back and 

 breast average less gray and the latter is more streaked. There is thus an 

 advance toward griseocauda; but Costa Rica birds are obviously much 

 nearer ruficauda. 



Beyond Costa Rica, however, while specimens with more or less tawny 

 in the tail are not infrequent (even southern Mexico birds sometimes show 



