Page:Condor20(4).djvu/16

 142 THE CONDOR Vol. XX dividual bird goes through the changes in- dicated. What reason is there for believing that goshawks with heavier markings are yc. unger than. those with finer vermicula- tions--since the transversely barred plum- age, coarsely or finely marked, is the only test we now know for distinguishing old from young? The .author may 'be perfectly correct in refusing 'recognition to the west- ern subspecies, striatulus, but a casual state- ment of his belief, such as is cited above, cannot by itself be expected to convince others. At just o.ne point in the paper is a tri- nomial used: "Hybrid Flicker, Colapres au- ritus [sic] cafes". As no comments are made it is not clear what inference is to be drawn from this manner of entry. The bird report by Anderson (pp. 376-381) lists species collected by the Canadian Arc- tic Expedition on the coast of extreme north- western British America and northern Alaska. Sixty-one species are listed, most- ly without comment. Mr. Taverner's pecu- liar usage of names is not adopted, the more generally accepted classification of the A. O. U. Check-List being followed throughout.--H. S. SWARTH. THE GEOGRAPHICAL DISTRIBUTION OF COLOn AND ,OF OTHER VARIABLE CHARACTERS IN THE GENUS JUNCO: A NEW ASPECT OF SPECIFIC AND SUBSPECIFIC VALUES. By JONATHAN DWIOHT, M.D. Bulletin of the American Museum of Natural History, vol. xxxwi, June 1, 1918, pp. 269-309, plates xi-xiii, 5 figs. (maps) in text. Dr. Dwight's Previous studies of plumage variation, together with his known interest in the group of birds here discussed, render this publication one deserving of more than ordinary attention. The problem concerned is the classification of the juncos so that names may be applied to the various groups of species and subspecies, the method em- ployed is somewhat novel, and the resulting systematic treatment of the genus is radi- cally different from that adopted in the A. O. U. Check-List. While, however, the ar- rangement of species and subspecies here given may be taken as the author's concep- tion of their proper relationships, the out- come of the careful study of a la.ge amount of material, the treatise itself is more in the nature of an essay on a method of research, rather than the detailed exposition of ac- cumulated data bearing upon this particular problem. Thus, in the author's own words, it is not 'so much his purpose "to attempt a complete revision as it is to focus atten- tion upon them [the juncos] from a new an- gle;" and "the winter ranges are not given and other matters of indirect interest are not taken up because they scarcely come within the scope of this particular study of tile Juncos." As a result the reader is confronted with many sweeping statements, rather dogmat- ically uttered, on points regarding which he might wish to weigh the evidence for him- self before accepting the author's classifica- tion of the genus as final. Characters of the juncos are found to be "of two kinds, qualitative and quantitative, which include all differences of structure, size, proportions, pattern, and coloration.' In structure .... they are all practically alike; in size and proportions, their differ- ences are quantitative; brat, in pattern and coloration, the variations are both quanti- tative and qualitative." Color characters alone are here considered. Nine areas on the biid's body are differentiated (head, breast, back, sides, wingoveris, tail, lores, iris, and bill), and each part considered by itself. The geographical distribution of the types of coloration on the several parts is separately platted, and species and subspe- cies determined according to the extent of coordination in the several maps. In a gen- eral way, of course, this (barring the maps) is very similar to what has been done by most monographers of bird groups, though not usually with the different parts of the bird so rigidly defined, nor with such abso- lute disregard for other modifying factors. Some of the results attained by Dr. Dwight are more or less in accordance with those of one or another of previous authorities on this group, but the allocation of some forms is so widely at variance with all prior class- ifications, that, before arriving at a final conclusion, it would seem desirable to give some consideration to factors other than those of color characters, so arbitrarily de- fined. Several forms in good standing in the Check-List are here regarded as hybrids, annectens, ridgwayi, montanus, and dorsalis being disposed of in this way. The speci- mens serving as types of annectens and ridwayi had already been shown to be clearly of snch character by Rldgway, but that montanus and dorsalis are of the same category is a new idea. The contention ap- pears to be well founded, and is a point of some importance in the author's argument. The occurrence of individual birds appar- ently of hybrid origin and in sufficient num- bers to have long been regarded as repre- sentative of distinct forms, is, of course, a feature deserving of most careful considera- 'tion in any systematic treatment of the group. A new name is provided Junco nomencla- ture, Junco oregonus coneel, proposed for the race called connectens in the Check-List, and shueldti by Ridgway. Connectens is regarded as a synonym of hyemalis (in ac- cordance with Ridgway's previous conten- tion), and shuel(ti as a synonym of ore- ganus. The type specimen of shueldti is a winter collected bird from Fort Wingate, New Mexico, and if this individual is actu- ally an example of the Alaskan Junco o.