1914] Grinnrll : Mammals and Birds of the Colorado Valhij 175 



no specimens of it were actually secured. But elsewhere abundant 

 material testifies to its status. 



The occurrence of this sparrow appeared to be correlated closely 

 with the extent of the arrowweed association, or of the almost as 

 attractive young willow growth. The lower ends of the large valleys 

 thus appeared to offer the best conditions. A series of lagoons below 

 Ehrenberg margined with tules and surrounded by arrowweed wa.s 

 notable for its song sparrow population, as was also the "drowned- 

 out" area just above the Laguna dam. Below Potholes, in a growth 

 of mixed cane and young willow, song sparrows were notably numer- 

 ous; also wherever there was a pure growth of cane, as along "Cane- 

 brake Caiion" and on the California bank near Pilot Knob. 



All the above areas are yearly sub.jected to overflow, and this may 

 be held to account for the fact that even up to ]\Iay 15 not a single 

 new nest or young bird of this species was found. In the vicinity of 

 Salton Sink young have been taken as early as March 30, and are 

 common by the middle of April. It would appear that the birds have 

 in the former region habitually come to defer their nesting time until 

 after the high-water period, the last of May and early June. It was 

 further observed that most of the old nests (which were considered 

 wath fair certainty to have belonged to song sparrows) were in arrow- 

 weeds or willows above the highest mud mark. The coating of mud on 

 vegetation up to a regular height was indicative of the pre\'ioas sum- 

 mer's flood level. So that, again, the song sparrows have here accom- 

 modated themselves to the extraordinarj- conditions by building their 

 nests much higher above the ground than is asual elsewhere. 



The fifty-one specimens (nos. 13246-13296) of Melospiza melodia 

 saltonis obtained by our party, together with three other skins (nos. 

 4178^180) taken by J. G. Cooper at Fort :\Iohave in January, 1861, 

 are in no perceptible degree different from the topotype series from 

 Salton Sink. (See Grinnell. 1909, p. 268.) The larger .series now 

 available increases the range of variation in all characters, but not so 

 much that any single specimen, even when considered by itself, would 

 raise a question as to its identity. The Colorado Valley birds, as far 

 as I am able to judge, are not in the least divergent towards fallax, 

 it being understood that the latter name applies to the Rocky Moun- 

 tain song sparrow. It would appear that saltonis is very sharply 

 defined both geographically and specifically. It is of further note that 

 Cooper's .specimens, taken fifty years airo. are identical with the 

 average saltonis as it is todav. 



