Page:Condor19(1).djvu/19

 18 THE CONDOR Vol. XIX though prosecuted quite intermittently. Repeatedly, after long intervals, I have gone back to verify earlier impressions, and by this method I feel sure that my appreeiation of the facts has been clarified; characters have been detected which were not seen at the start, and fortuitous variations have come to be recognized as such. Hesperiphona is subject to but one molt per year; yet there is much variation in certain respects owing to wear and possibly, also, to age. To weed out these variations of non-phylogenetic significance is of course a necessary process in working up any group of birds. The characters of Evening Grosbeaks which I have found serviceable in sub- specific discrimination are as follows: (1) Proportions of bill; (2) width of frontal yellow band in male; (3) color tone of sides, and lower surface generally, in female; (4) color tone of top of head and back in female; (5) color tone of upper and under parts in male. With the' color tones, the student has to keep continually in view the probable degree reached in the process of wear, and even then some variation is to be found which he must deal with in mass-effect so as to determine the mean condition. As a result of my study, it is found that Hesperiphona vespertina montana, the "western" subspecies reeognlzed in current literature, is a composite of four distinguishable races. To provide names for these, correct according to the rules of nomenclature, it has been necessary to restrict the current name within nar- rower limits. The following discussion goes into more detail than might other- wise be needed, because my conclusions as to the basis of the name montana are at variance from those of other authors (e.g., Mearns, 1890, and Chapman, 1897), and I wish to anticipate as fully as I can any queries that may arise on the part of future students. The original characterization of montana by Ridgway (in Baird, Brewer and Ridgway, 1874, p. 449) unmistakably fits the extreme Mexican race, in com- parison with vespertina, as does also the figure on plate XXII of the same work, even if the source of the specimen from which the drawing was made had not been indicated (as it was). In fact, birds from the "Pacific Coast to Rocky Mountains", as well as from "North America east to I, ake Super/or" (loc. cit., p. 450), are all included under "var. v espertina"; while montana is stated to be found in "Guatemala, Mexico, and the southern Rocky Mountains". It was only long subsequent action that extended the range of montana to cover the whole of "Western North America" wherever Hesperiphona occurs at all. I am aware that a specimen in the United States National Museum (no. 11,960), from Cantonment Burgwin, New Mexico, has been considered the type of montana; but this specimen was not published as the type until 1890 (Mearns, p. 247) } and, as just pointed out, a virtual type had already been designated, no. 35150. This latter specimen, from which the drawing of the bill in both Cooper's Ornithology (1870) and Baird, Brewer and Ridgway (1874) was made, bears the following data: Mirador, Mex. (near Vera Cruz); pine forests; June 64; Dr. C. Sartorius; , 180 (punctuation different on tag). This, it now seems clear, is the real type of ttesperiphona vespertina montana Ridgway; and the name mexi- cana Chapman, based also on a bird from the state of Vera Cruz, becomes a pure synonym of montana. Through the courtesy of the custodians of the National Museum I have been permitted to examine both of these specimens, of such criti- cal importance. The one from Fort Burgwin, near TaGs, in northern New Mex- ico, belongs to the subspecies here called wartertl. It is clear from canons xxxm and xLi of the A. O. U. revised (1908) Code of