36 THE CONDOR Vol. XXIII 
During July the birds were apparently absent, until July 21, when a female was 
discovered feeding in a gravelly riffle about one-fourth mile east of the Sentinel Bridge. 
The water was only three to four inches deep and the current strong. The bird seemed 
to be industriously turning over the rocks to obtain food between and beneath them. 
Often she was wholly immersed for from six to ten seconds by count. At the end of 
about ten minutes she drifted down the river and dove several times in still water. In 
the afternoon about 5:45, this female Harlequin returned to the same feeding ground 
and was watched again. For a full half-hour it continued feeding in the same manner, 
continually ducking its head under the swift current and always working up-stream. 
Apparently the Harlequin does not procure all of its food by diving, but at times 
feeds in shallow water. The occurrence of these birds during the nesting season and 
the disappearance of the male during the middle of the summer would indicate nesting 
of the species in the Valley or close at hand, but no direct evidence in this regard was 
secured.—HaroLp C. Bryant, Berkeley, California, November 10, 1920. 
Distribution of the Townsend Fox Sparrow.—lIn studying Swarth’s Revision of the 
Avian Genus Passerella (Univ. Calif. Publ. Zool., vol. 21, 1920, pp. 75-224), the attention 
of the writer was drawn to some apparently erroneous conclusions of the author regard- 
ing the migration and distribution of Passerella iliaca townsendi, particularly as to its 
winter range. 
On page 145 of the paper under discussion, Swarth states “The ‘Townsend fox 
sparrow is a notable example of a bird- with a winter habitat nearly as sharply defined 
as its summer home.” On page 105 he states further that “townsendi in turn leap-frogs 
over fuliginosa’, the breeding bird of the Puget Sound and Vancouver Island region, 
the impression being given here and by the map on the following page that the “sharply 
defined” winter habitat of townsendi lies entirely south of that of fuliginosa. The author 
further, on pages 145-146, calls the attention of the reader to the apparently discontinu- 
ous distribution of townsendi in the southern part of the Alexander Archipelago. The 
1909 Alexander Expedition failed to find it at the localities visited in that region, but 
Swarth (loc. cit.) mentions the fact that it is known to be a common summer visitant to 
Forrester Island, near the southern extremity of the archipelago, and records summer 
specimens taken by other collectors at Howkan and Wrangell. 
The following data from notes of the writer accumulated during six summers 
(1914-15-16-17-19-20), and one winter (1919-20), spent in the region under discussion, 
fill some of the gaps noted by Swarth and modify some of the conclusions which he 
reached. The greater part of the six summers were spent on Forrester Island, but vc- 
casional short visits were made at this season of the year to nearby points on Dall and 
Prince of Wales islands. During the winter of 1919-20 the writer resided at Craig, 
Prince of Wales Island, but frequent trips were made to nearby sections, to Suemez, 
Dall and Long islands, and to other points on Prince of Wales Island. 
That the Townsend Fox Sparrow breeds more plentifully on Forrester Island than 
at any other point in the southern end of the Alexander Archipelago is very true. But 
that it fails to breed on Prince of Wales, Dall and Long islands, the writer doubts. 
Although the 1909 Alexander Expedition failed to find it in the region in summer, the 
writer has found it at that season at Craig and Waterfall, Prince of Wales Island, at 
several points on Dall Island, and at Howkan, Long Island, and, though no occupied 
nests were examined at these points, several nests entirely typical of the bird were 
noted in the fall. 
On Forrester Island the species was always present at the time of the writer’s 
arrival, the earliest date being April 21 (1915). In this locality probably fifty nests’ 
were examined during six summers. The majority of the eges are laid between May 29 
and June 20, extreme nesting dates being April 29 (1915), a nest with one fresh. egg, and 
July 9 (1916), a nest with three eggs. 
During the latter part of August there is apparently a movement up the sides of 
the mountains, and for some time after this date townsendi is quite rare near sea-level. 
This movement is shared to a considerable extent by the Varied Thrush (Ixvoreus naevius 
naevius) and Oregon Junco (Junco oreganus oreganus), the three species being fre- 
quently found in close proximity in the woods on the mountain sides from about 1000 
feet altitude to timber line. As the weather becomes cooler they work back down the 
mountains to the shore. At Craig, in 1919, the Townsend Sparrow became common in 
