228 Universitij of California Publications in Zoology [Vol.12 



Potholes, five miles northeast of Yuma, and near Pilot Knob; Arizona 

 side: Mellen, Ehrenberg, twenty-five miles below Ehrenberg, five miles 

 northeast of Laguna, near Yuma. 



The species thus proved to be regularly distributed all along the 

 Colorado Valley. Yet we failed to trap it anywhere on the nearby 

 desert outside of the mesquite belt. In other words, this Peromyscus 

 has appropriated the river bottom, which, in turn, is tabooed by the 

 two desert species of the region, eremicus and stephcnxi. It is to be 

 inferred that sonoriensis used the Colorado Valley as a highway of 

 immigration through the region, and found it suitable for permanent 

 occupancy. The maniculatus division of the genus Peromyscus is 

 notorious for its success as an invader. Single subspecies representa- 

 tive of this group range through the habitats of dozens of other species 

 of small rodents, from the hottest. Lower Sonoran. Colorado desert to 

 the Boreal zone on the San Bernardino Mountains. It is thus a 

 remarkably hardy animal; yet, as far as known, it nowhere thrives 

 to the total exclusion of other small rodents, save in the case of slightly 

 different subspecies on certain islands. 



The aggressiveness of Peromyscus m. sonoriensis was illustrated 

 by our experience along the Colorado. It was the only wild mammal 

 which found its way on to our boats; individuals were twice routed 

 out of the cargo of supplies on the scow. This boat was most of the 

 time moored alongside the shore, either touching the bank at one corner, 

 or with a gangplank out, so that any venturesome animal could easily 

 go aboard. 



At a point five miles northeast of Yuma, Sonora white-footed mice 

 were trapped April 30 on a river bar covered with a very young grovrth 

 of willow. This was at a time when the rising water had already cut 

 the place off from connection with the main shore. Footprints of mice 

 were plentiful on the dry. fine silt on the higher part of the island 

 thus formed. In several other places this rodent was trapped on 

 portions of the flood-bottom elevated in such a degree that they would 

 have been first cut ofif from the shore as the water rose and then set 

 adrift. In some of these cases safety could have been secured by 

 swimming short distances through relatively quiet water, or by taking 

 refuge on floating drift. It is highly probable that in the rush of 

 flood waters a large mortality must annually occur. 



It is probable also that the paucity of snakes and other enemies 

 in the flood-bottom gives this mouse relative immunity from those 

 dangers which beset the small mammals out on the desert; so that 



