1914] Grinnell: Mammals and Birds of the Colorado Valley 137 



colors in the rudiments of their feathers to be ascertained. Nos. 

 12763 and 12764, both with reddish coloration, were a mated pair taken 

 in the overflow bottom on the Arizona side. No. 12765. of reddish 

 type, was taken on the desert on the Arizona side. 



The salient fact shown by this comparative examination is that 

 in all oihir characters the specimens aberrant in color of wings and 

 tail, are perfectly typical of chrysoides (that is, of its subspecies 

 mearnsi). None of the phenomena consequent upon hybridization is 

 evinced in other particulars, such as general size, proportional dimen- 

 sions, extent of dorsal barring, colors of body and head. In all these 

 characters there is no nearer approach of the red-shafted chrysoides 

 to collaris, than of the yellow-shafted chrysoides. 



My conclusion is that the strain of chrysoides occurring at the 

 present time in the lower Colorado Valley shows proneness to replace- 

 ment of yellow by red. without there having been any interbreeding 

 with another species. This may be accounted for chemico-physiologic- 

 ally, as in the case of the linnet of the Hawaiian Islands, where, how- 

 ever, the change has been from red to yellow. (See Grinnell, 1911, 

 p. 191.) According to this idea there has been some cause, germinal 

 or somatic, affecting the oxidation of the basic chromagen, so that in 

 the critical instance the process is going beyond the stage of yellow- 

 production and reaches the red-production stage. The phenomenon 

 at the present time may be purely sporadic ; on the other hand it may 

 be an incipiency of a character which, beginning in this center of 

 differentiation, may become fixed and heritable over a larger area, 

 thus constituting the arising of a new species. The flickers collected 

 were only a few out of the large population which probably occupies 

 the extensive tract of saguaros widening out to the eastward of the 

 river in Arizona. Observation in the same region at successive times 

 in the future will afford data indicative of the real significance of 

 the phenomenon. 



It is quite evident that the aberrant examples described by Brew- 

 ster and by Swarth from central Arizona, as referred to above, are of 

 the same nature as the Colorado Valley specimens. The chances are 

 that they were not hybrids. So far as shown by the literature at hand, 

 no unquestioned hybrids have been found between chrysoides (or any 

 of its subspecies) and collaris or cafi r. 



