Page:Condor16(5).djvu/26

 216 THE CONDOR Vol. XVI will attract attention, than that a statement of probable affinities will be re- membered from one out of a mass of papers. Although it is of course a matter of regret that there is but one specimen of this island subspecies at hand, I believe that the appearance of this single bird justifies the naming of the race to which it belongs. Although there is great variation shown in series of Salpinctes obsoletus, both as regards shades of gray or brown, and character of spots, bars, or streaks, on various parts of the plumage, this specimen stands absolutely outside of this range of var- iation, so that I do not believe that there is any question of its belonging to this species. From guadloupnsis it is not so readily distinguished, at least as regards color, but the measurements lie outside the limits reached by that form.- As in its variation from typical guadloupsis there is an apparent approach toward the characters of obsoletus it might be considered as illus- trating intergradation between the two, but for the present at least, in view of the many peculiarities of distribution observed in the genus, it seems best to consider obsoletus and guadeloupcnsis as distinct species. We are probably safe in assuming that the Rock Wrens of all the islands off the coast of California, as well as those of most of the Lower California islands, are derived from the mainland form Salpinctes obsoletus. In fact, in most cases they are not to be distinguished, though it does seem to me that in the island birds throughout there is to be detected a slight general tendency toward lengthening of culmen. In the Santa Barbara group this tendency has reached, on San Nicolas, most remote from the mainland, a stage where we are perhaps justified in recognizing the variation in nomenclature, and considering the San Nicolas Rock Wren as a separate subspecies. There is another slightly differentiated island form of obsoletus, S. o. exsul (not seen by me), from San Benedicto Island, of the Revillagigedo group, off the coast of western iIexico. Of additional island ocalities there are at hand specimens from most of the Santa Barbara islands, and from the following Lower Cali- fornia islands: The Coronados, San Benito, Cerros and Ildefouso. None of these are to be distinguished with certainty from typical S. obsoletus. Thus there is on the mainland coast of California and Lower California, and on most of the adjacent islands, the Rock Wren, Salpinctes obsoletus, in its three very slightly distinguished races, obsoletus pulverius and exsul. In the midst of this general range there is found on two islands, Guadalupe and San iIartin, a sharply differentiated form, Salpinctes guadeloupensis, apparently divided into two races, guadeloupensis and proximus. Bearing in mind the above facts as regards distribution, and also the degree and kinds of difference distinguishing the forms, it seems to me that in the light of our present limited knowledge of the subject, it is best to regard Salpinctes obso- letus and S. guadeloupensis, as distinct species, the first composed of several, the second of two, different forms or subspecies. In other words, it is the treatment accorded these forms in the A. O. U. Check-List (1910, p. 336) that seems to me the more reasonable, rather than the view expressed.by Ridgway (1904, pp. 643-653) in his recent study of the group. At the same time recognition must be accorded the possible significance of the peculiar juvenal plumage of Salpinctes obsoletus hofius (not seen by me). In this iIexican form the young is described by Ridgway (1904, p. 648, foot- note) as being similar to the corresponding stage of guadeloupensis, and this may be an indication of close relationship between these two forms, though the geographical position of hofius adds no emphasis to such a theory. The