VARIATION 295 



These observations concern one character only — the length of 

 the wing. But they are of considerable interest since they 

 reveal a remarkably wide range of variation in this particular 

 in one or two species ; while in closely allied species this range 

 is often surprising. Further, he shows that, in many cases, in 

 times of stress, as in unusual inclemency in the weather, the 

 shortest winged forms are the first to succumb. 



The Willow-wren and the Skylark he found to be the most 

 variable among British birds in this matter. In the former species 

 the wings of eighty-two specimens gave the following measure- 

 ments : — 



Willow-wren — 



Average wing . . . 2'57 



Longest „ .... 2"85 



Shortest „ .... 2'ao 



Percentage of variation . 25. 



Lark — 



Average wing .... 4*20 



Longest ,, . . . . 4'8o 



Shortest ,, . . . .375 



Percentage of variation . 25. 



The Wheat-ear and the Whin-chat are very closely allied, 

 yet the following table shows they differ greatly in the stability 

 of their measurements : — 



Whin-chat- 



Average wing .... a'gS 



Longest ,, .... 3"io 



Shortest „ .... 2-85 



Percentage of variation . 8J. 



Wheat-ear — 



Average wing .... 378 



Longest „ ... 4-15 



Shortest „ .... 3-52 



Percentage of variation . 16J. 



Among the Finch tribe the range is interesting ; thus from 

 specimens taken at Irish Lighthouses Mr. Barrington found the 

 following differences — to take but a few cases : — 



