190 University of California Piihlicafions i)i Zoologi). [Vol.7 



operative to an extent to establisli a limit of population. There 

 is nothing to indieate that removal of adverse intliienees results 

 in weakening the vigor of the species. 



Of course, general decadence of a genus whether or not 

 obviously overspecialized, is an entirely different proposition. 

 But this does not appear to be at all probable in the case of 

 Carpodacus, a genus of very wide geographical range, through 

 greatly varying conditions. 



The fact that the off -color character appears to be most preva- 

 lent on islands (as those off Lower California as well as the 

 Hawaiian) is to me of greatest significance. Close breeding is 

 known to be in some cases accompanied by various abnormal 

 manifestations. The assumption of the orange or yellow dress 

 in the insular linnets may, after all, be at bottom, of intrinsic 

 or germinal origin. 



This is emphasized by the conspicuous way in which the color- 

 manifestations of the Hawaiian linnet are in direct harmony 

 with the tyrosin-oxidation idea, as expressed by Riddle. In order 

 that this chemical operation may proceed, the presence of a cer- 

 tain enzyme is essential at the time of feather-growth in the 

 dermal tissues involved. The amount of enzyme produced is 

 quite likely to be an inherited character modified through ger- 

 minal variation. It is within possibility that those individual 

 linnets originally introduced on the Hawaiian Islands happened 

 to be fortuitously of a character deficient in enzyme. This char- 

 acter, as with conspicuous physical characters in many observed 

 cases, might have been intensified through close breeding, until 

 sufficient in degree to bring about an arrested tyrosin-oxidation, 

 yellow or orange pigment resulting instead of red. This explana- 

 tion appears to me somewhat more deserving of credence than 

 any other that has come to mind. Yet this series of processes are 

 of as yet so inscrutable a nature as to make the explanation 

 offered tentative to the last degree. 



Summary. — In estimates of color characters account nuist be 

 constantly taken of the complex relations existing between the 

 colors themselves, feather structure, and the process of wear. 



In all available specimens of the linnet from its United 

 States range the usual color (in the parts of the plumage which 



