236 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 85 



A male and a female shot near Beaverdam Creek in the valley bot- 

 tom at Shady Valley on June 5 and 10 are canaster, though birds 

 from low in the Holston Mountains a few miles west are referred to 

 erythrophthalmus. This is the farthest north and east that canaster 

 is recorded. An immature female taken on September 18 at 5,900 

 feet elevation on Roan Mountain has the tail spot only 27.2 mm, 

 though the flanks are dark. It is considered intermediate but nearer 

 canaster. This may be an area of intergradation. Two from the 

 Great Smoky Mountains, a male taken on June 19 at 5,000 feet on 

 Cosby Knob and a female on June 29 at 6,100 feet on Old Black 

 Mountain, are both canaster. These two indicate that the southern 

 form extends through these mountains and on to the south. 



PASSERCULUS SANDWICHENSIS SAVANNA (Wilson): Eastern 

 Savannah Sparrow 



Two eastern Savannah sparrows were taken at Bartlett on April 

 19, at the same time as one of the paler Churchill form. At Rockwood 

 W. H. Fox secured specimens on March 18, 21, and 31 and April 7, 

 1885. These are all dark in general appearance, with the lighter 

 edgings of the dorsal feathers distinctly brownish. 



PASSERCULUS SANDWICHENSIS OBLITUS Peters and Griscom:" 

 Churchill Savannah Sparrow 



Two females collected by Perrygo and Lingebach, at Ellendale on 

 April 17 and near Bartlett on April 19, are marked by the pale 

 gray margins and heavy black centers of the dorsal feathers, gray 

 and black being the predominant colors, with little or no buff or 

 brown. They are considered migrants of this race, which is recorded 

 in the original description ^= from the Great Smoky Mountain region. 

 The form is well marked and easily distinguished. The abundance 

 of this subspecies and of the true Savannah sparrow in Tennessee 

 has still to be ascertained. 



In the paper containing the description of this new form, a treat- 

 ment of geographical variation in the Savannah sparrow, the 

 authors'* list the Ipswich sparrow as Passerculus sandicichensis 

 princeps, saying that "there is no absolute difference of any kind 

 between this form and one or more races of P. sandwichensis. In 

 size it is not only no larger than P. s. sandwichensis^ but the smallest 

 specimens are smaller than the largest specimens of P. s. savanna. 

 The pallor of its coloration is not very marked when compared 

 with P. 8. nevadensis, and is exceeded by certain races of the rostratits 



^ Passerculus sandwichensis ohlitus Peters and Griscom, Bull. Mus. Comp. Zool., vol. 80, 

 Jan. 1938, p. 454 (Fort Churchill, Manitoba). 

 ^lUd., pp. 456, 458. 

 ^Ibid., pp. 447-448. 



