BIRDS OF WORTH AND MIDDLE AMERICA. 683 



variations are more or less marked and constant; but occasionally 

 specimens occur in a given area which are with difficulty, if at all, 

 distinguishable from the form inhabiting another — sometimes dis- 

 tant — ^geographic area. When to these perplexities are added the 

 difficulty of conveying by means of description a clear idea of the 

 coloration of any particular form on account of the confused pattern 

 of coloration, which consists largely of fine vermiculations, zigzag 

 lines "herring-bones," etc., the matter of satisfactorily describing 

 the differential characters of allied forms becomes peculiarly difiicult. 

 Indeed, few groups of birds present greater perplexities to the sys- 

 tematist than the present one, chiefly on account of the difficulty of 

 bringing together sufficient material to satisfactorily illustrate the 

 extent of geographic and individual variations. Some species are 

 very uniform in their color characters throughout a vast extent 

 of territory; for example, 0. cholila, which from Paraguay and 

 southern Brazil to Costa Kica seems to be everjTvhere practi- 

 cally the same. On the other hand, other species, like 0. asio, 

 change remarkably in coloration within relatively short distances, 

 their organization being so sensitive to conditions of environment 

 that often relatively small contiguous areas of only slightly dif- 

 ferent physical character are characterized by recognizably differ- 

 ent forms. When a study of these owls is attempted with scant 

 material, that is, specimens from widely scattered localities and few, 

 often only one, from each, the impression is received of excessive 

 individual variation of one widely distributed form. This has espe- 

 cially been the case with the bare-toed group, containing forms related, 

 more or less closely, to 0. cJioliba; and some thirty-five years ago, 

 when the writer prepared his "Keview of the American Species of 

 the Genus Scops, Savigny,"" this view was adopted, even 0. cassini 

 being referred to 0. choliha as a subspecies. Although material for 

 the study of this group has not increased very greatly even after the 

 lapse of so many years, there has nevertheless been added to collections 

 enough specimens to show clearly that the view formerly held (and, 

 to a certain extent, held as late as 1897 by the authors of the Biologia 

 Centrali-Americana, Aves, (iii, pp. 15, 21, 22), is incorrect, and that 

 not only are the variations noted to a great extent geographical but 

 that often two so-called "styles of plumage" foimd among specimens 

 from the same country (sometimes even from the same locality) in 

 reality represent distinct species. Thus, in the vast extent of South 

 America inhabited by 0. choliha are a number of other forms, several 

 of which are imdoubtedly specifically distinct, and in Costa Rica and 

 Panama that species is associated with 0. vermiculatiis; and in the 

 State of Vera Cruz the large and relatively light colored 0. guatemalm 



a Proe. U. S. Nat. Mus., i, 1878, 85-117. 



