254 



PROCEEDINGS OF THE NATIONAL MUSEUM. 



VOL. 49. 



posteriorly, and with practically an equal amount of mottling on the 

 legs and feet. In the reduction of light markings on the upper surface 

 the type of Strix occidentalis caurina may be closely matched in all 

 respects by some of the California birds, as well as by the type of 

 Strix occidentalis occidentalis. Moreover, the type of Strix occidentalis 

 occidentalis is an old, formerly mounted specimen, and is considerably 

 faded and apparently otherwise discolored. Another southern Cali- 

 fornia example (No. 1392, collection of G. Willett) is actually identical 

 with the type of Strix occidentalis caurina, except for being darlcer. 

 There is thus no distinctive character left for the recognition of Strix 

 occidentalis caurina as a subspecies. 



From what has already been said it is evident that there are but 

 two forms of Strix occidentalis, instead of four, although both have 

 a relatively wide geographical range. The only characters that now 

 appear to be of value for the separation of these geographical races 

 are in the size and distinctness of the white markings on the feathers, 

 principally of the upper parts, though also below, but there is great 

 variation even in this respect in the series of specimens examined. 



The following millimeter averages of birds from various regions 

 throughout the range of the species, show clearly of how little sub- 

 specific value is any difference of size : 



This reduction of the four described and currently recognized sub- 

 species of Strix occidentalis to two closely allied forms is a surprising 

 conclusion, in view of the wide geographical range of the species — 

 from central Mexico to southern British Columbia — and also of the 

 well-known tendency of owls in general to form numerous geographi- 

 cal races, but it is inevitable from an examination of the material 

 now available. The only other course possible is to reduce them 

 to a single form. The present result is, however, nearly paralleled 

 by the unexpected, but none the less apparently correct conclusion 

 by Mr. Robert Ridgway ^ concernmg the lack of any recognizable 

 subspecies in Otus jiammeolus. It shows, too, that geographic or 

 physiographic analogy, though of considerable assistance in working 

 out geographical races, is an unsafe guide alone. 



1 BuU. U. S. Nat. Mus., No. 50, pt. vi, Aprils, 1914, pp. 728-730. 



