THE CONDOR Vol. XlII Western Savannah Sparrow (Passerculus sandwichensis alaudinus). Still present at Earli- mart, Tulare County, up to May 4. This to my mind constitutes late tarrying of winter visitants, and in no wise indicates a breeding station. The behavior of the birds at no time was such as to lead one to suspect nesting; and a specimen shot on April 30 was just completing a partial pre- nuptial molt, being in the consequent plumage a duplicate of Alaskan specimens. It would ap- pear quite unsafe to base breeding records of any of these migratory sparrows upon anything short of actual discovery of nests and eggs or small young. Intermediate Sparrow (Zonotrichia leucophrys gainbell).' Observed at various points all through April. At Earlimart, Tulare County, several were noted on the 30th. One was shot on May 1, and none were noted thereafter, thus establishing a date of departnre for this season and place. Nuttall Sparrow (ZonotricMa leucophrys nullalii). A number of this form were noted in mixed flocks of sparrows in rose thickets along levees five miles northeast of Tracy, San Joaquin County, March 11. Two specimens shot were preserved, and comparison in the Museum shows them to be unquestionably of this race, thus establishing an eastward extension of the .known winter range of Z. l. nullalii. Heermann Song Sparrow (A?irelospiza rnelodia heerrnanni). During the travels of myself and assistants, we kept a constant lookout for song sparrows. The result was that contrary to previous notions large parts of the San Joaquin Valley were found to be absolutely without any represen- tative of the genus. Neither in the vicinity of Tracy, Los Banos, or Raymond could song sparrows be found. A few were noted in the neighborhood of Fresno; and on the Fresno County side of the San Joaquin River near Lane Bridge (ten miles north of the city of Fresno) four specimens were secured. Mr. John G. Tyler, of Fresno, who was with me at this point, helped me to secure these and also contributed a nest with four slightly incubated eggs which he found in the river bottom close to our camp on April 7. Another nest with three fresh eggs was found on the 8th. In each instance the nest was located in 16w vegetation, against which drift-trash had lodged; in one case the nest was 2? feet above the ground, in the other four feet. Song sparrows were found again only at Earlimart, Tulare County, where a male and two females were taken April 30 and May 2. These were the only individuals observed at this place, and were inhabiting a willow-margined reservoir. The seven song sparrows secured, as just specified, are as uniform as usual, considering the normal range in individual variation; and they are very nearly duplicates of topotypes of At. rn. heerrnanni from Fort Tejon, Kern County. The known range of heerrnanni, as lately restricted (see Grinnell, Univ. Calif. Publ. Zool. V, April 1909, p. 266), is thus extended north to include parts of the Tulare basin. lYeerrnanni is distinct from A?ir. rn. rnaillairdi (Univ. Calif. Publ. Zool. VII, February, 1911, p. 197). The differences lie in the much paler "ground color" dorsally of heerrnanni;the narrower,black-streaking both above and below and in the slightly smaller bill. There is still a great stretch of country--between Fresno and Modesto--from which we have no Melospizine returns. There may be an actual hiatus between the ranges of heerrnanni and rnailliardi. For, as our San Joaquin work has demonstrated, it is a grave mistake to as- sume that song sparrows range uniformly all over the bed of the Valley and up into the foothills. Rather are there only narrow belts of occupied ground, coinciding with sections of riparian strips. Vast areas of dry prairie intervene, unsuited to this bird. However, a circumstance accompanying human invasion will tend to obliterate these original conditions: Song sparrows were seen in the Fresno district along irrigation canals. These canals thus serve to divert a stream of riparian plants and animals, including the song sparrow, out over the plains between the rivers, by whidx process the fauna of the originally arid levels becomes metamorphosed. The ranges of the song sparrows of interior California may thus be expected to shift to a consider- able extent from what they were or even are at the present stage of events. An incident of interest though not of definite significance was that at the Earlimart reser- voir referred to above there were two females, each with a nest, but only one male, at least at the time of my arrival, April 30. One of the females, shot together with the male on that date, contained very large ova (one egg would have been laid probably the next day), and her nest was apparently completed. The other female was taken on May 2 together with her nest and fonr fresh, or infertile, eggs. This bird was incubating, as the subdermal layer in the abdominal region was glandular to an extreme degree. This state of affairs (one male, and two females with nests) might be accounted for by any one of three explanations: (1) that there. was another male at the reservoir, bnt destroyed by some means before my arrival; (2) that an excess female without a mate had gone ahead and built a nest and produced infertile eggs; or (3) that where there were more females than males, polygamy had occurred and the male. had mated with two
