Vol. XIX 

 igo2 



1 DvviGHT, Variations of the Americaji Goldfinch. 



163 



ations whatever way we look at them, and there is still another 

 factor in the matter with which we must reckon. This is the per- 

 sonal factor, and by it I mean that no two students are likely to 

 measure the same series of birds alike. The subjoined table, com- 

 piled from published records, shows variations quite as great as any 

 of which I have already spoken, and illustrates individual variation 

 in the measurer rather than in the measured. 



Table III. 



The most obvious fact to be deduced from the foregoing table 



^ Five specimens. 



