PICUS VILLOSUS, HAIRY WOODPECKEE. 281 



six) on the primaries, and none on the inner secondaries and wing-coverts. The change 

 begins iu birds from tlie eastern slopes of the Rocliy Mountains, and is effected by im- 

 perceptible degrees. Laying a few scores of specimens side by side, in linear series 

 representing the change in longitude, one may see the spots disappear one by one, or 

 by pairs, proving the complete iutergradation of the two forms. This has been known 

 to several ornithologists for some years, and hints to this effect have repeatedly been 

 given ; but, so tar as I am aware, it was not formally recognized, by reduction of liarrhu 

 to a variety, until this was done simultaneously by Mr. Allen and myself 



We will next observe that the original of I'icna h-arrisii, from the Columbia River, 

 has the under parts of a peculiar smoky-gray tinge, with also more or Itss evident thin 

 black stripps along the sides. This appears chiefly in specimeus from Oregon and 

 Washington Territories, and I think is something more than mechanical soiling of the 

 feathers with carbonaceous matter from the trees, although this agent undoubtedly 

 sometimes effects a like result. All my New Mexican and Arizonian specimens are as 

 pure white below as in the eastern variety, and usually lack all traces of stripes along 

 the sides. 



Resting upon these facts respecting the distinctions between the eastern and western 

 form, we'will take up the remarkable variation in size that both exhibit in a perfectly 

 parallel manner. Specimens of typical villosus occur ranging from 8 to U inches long ; 

 and this difference is found to depend upon latitude, size increasing ;jn/i^assit with 

 heightening of latitude, by insensible degrees. But there is nothing exceptional 

 about this; on the contrary, it is merely a strong illustration of the prevailing rule 

 ■with widely-distributed birds, the individuals of which are stationary or nearly so. 

 With the maximum size of the northern birds there is associated an increase of white- 

 ness, also in conformity with a general rule. These two particulars, size and hoariness, 

 give the bird a peculiar aspect, so readily distinguishable that the Pious " leiicomelas" 

 or " canadensis" has passed for a species with mauj? writers. But I see no necessity of 

 recognizing it by even varietal name; if we did, we should also, to be consistent, sep- 

 arate the Hylotomus pilealus into races founded on size, the variation being fully as 

 great in this case. The smaller size of the South Atlantic and Gulf States' ■villosus is 

 recognizable, on an average, but is not so strongly marked; in coloration they are jire- 

 cisely the same as the ordinary bird of the Middle States. 



In the West, the boreal extension of var. Jiarrisii is out off by the topography of that 

 region, the westward trend of the mountains bringing the eastern zoological province 

 to the very Paciiie, and giving us var. vilJosus^instead ot harrisii. Consec[uently, typical 

 var. liarrisii is itself the largest form. It about equals ordinary villosus of the Sliddle 

 States. The New Mexican and Southern Califomian specimens are slightly smaller 

 than those of Washington Territory, but not notably so, the chief distinction being 

 the color of the under parts, as already noted (liyloscojnis, Vdbams). But the Mexican 

 extension of the species, as well as I can judge by descriptions, is strongly marked by 

 its, small size, being quite as much under the average as P. " leucomelus" is over the 

 average. This is the P. jardinei of authors. The parallelism is thus seen to be com- 

 plete; in the east, a series with a marked northern maximum of size and less con- 

 spicuous minimum ; in the west, a marked southern minimum and less conspicuous 

 northern maximum ; iu both cases, the variation directly and precisely according to 

 latitude ; while the iutergradation of the two according to longitude is perfect. 



It is too late to say much about the various nominal species established upon con- 

 ditions of immaturity; they have already been satisfactorily disposed of. But, 

 although so admirable and long-experienced a field-naturalist as Audubon might well 

 he excused for proposing liarrisii and ijairdnei-i, at a time when climatic variation had 

 no place in the conceptions of naturalists, it is incomprehensible that, after years of 

 study in the field, and shooting perhaps hundreds of Hairy Woodpeckers, he should 

 never have found out that the young males have more or less of the vertex red or yel- 

 low, instead of an occipital crescent of scarlet. This crown-patch is sometimes per- 

 fectly square, and bright yellow, almost exactly as in Picoides; butit is ol'tener dif- 

 fuse, of a bronzy or coppery hue, mixed with white, iStc. P. " phillipsii" is the young 

 of the northern race ; P. " martinw '' may be so too, being from Canada, but the ascribed 

 dimensions (" length, 9^'") are those of ordinary I'illosus, to which it may be as well to 

 relegate it. " Septentrionalis" and " ruhricapillus" of Nnttall are the same as "phil- 

 lipsii" and "martinw," respectively. " Picus anduionl" is one of the curiosities of 

 ornithological literature, inasmuch as Swainson and Trudeau, without the slightest 

 reference to each other, each proposed the same name for the same bird in the same 

 plumage, and not a valid species either. 



This species moults the quills and tail-feathers very tardily ; they persist with de- 

 creasing vitality, until just before they drop they are often found faded, toward their 

 ends at least, into brownlsh-iofejte. 



This condition has crept into some of the published plates. Another individual 

 peculiarity is, that the occipital scarlet hand of the male is often discontinuous behind, 

 being thus resolved into two supra-auricular spots, almost exactly as in Picus querulus. 

 This may have induced Gray to ^at jardinei alongside P. borealis in the subgenus PUren- 



Digitized by Microsoft® 



