l62 



DwiGHT, Variations of the American Goldfinch. 



TAuk 

 LApr. 



The remiges and retrices being worn for a twelvemonth, it is 

 obviously unfair to contrast a series that might be chiefly worn 

 birds with one consisting of fresh-plumaged specimens, so I have 

 prepared the following table which contrasts fresh with worn 

 birds. 



Table II. 



14 tristis, 



5 salicamans 



8 tristis 



8 salicamans 



8 tristis, 



20 salicamans 



7 tristis 



5 salicamans 



9 tristis, 



6 salicamans 



3 tristis 



3 salicamans 



3 tristis, 



1 1 salicamans 



2 tristis 



2 salicamans 



Males. 

 before ist midwinter 



" 2nd " . 



a u a 



after ist midsummer 



u u a 



" 2nd " . 



U il u 



Females. 



before ist midwinter 



u a u 



" 2nd " . 



u u u 



after ist midsummer 



u u 



2nd " . 



Wing. 



Tail. 



r ^ 



The two tables indicate that adults average a little larger than 

 young birds and suffer more from wear, both losing between 

 periods of moult an appreciable amount of the ends of the wangs 

 and tail. It further appears that saHcanums averages slightly 

 smaller in wings and tail and slightly larger in bill, but the small- 

 ness of all of these average differences is apparent if we analyze 

 the figures. I have carried them to tenths of a millimeter in order 

 to be able to show the differences and there is such an overlapping 

 of dimensions in individuals that the ruler gives slight information 

 as to the age, season or race to which a specimen may belong. 

 The average difference in length of wings and tail between tristis 

 and salicamans is about two millimeters, a like difference existing 

 between first and second year birds of either race, and a similar 

 amount of wing and tail length being lost through wear in each 

 race. It is obvious that we are dealing with extremely small vari- 



