1922.] the Xear East and Tropical East Africa. 15 



Now tlie races into which the Wheatears under reA'iew 

 have been divided are : — 



cenanthe from Europe and western Asia generally. 

 argentea from central Asia. 

 rostrata from Syria and Palestine. 

 nivea from soutliern Spain. 



The characters on which these races depend for their 

 separation are the size of the culmen and win^, the degree 

 of colour on the mantle, the extent of white on the forehead, 

 and the colour of the wing-margins in autumn plumage. 



The validity of these geographical races seems to depend 

 on the proportion, necessary within a given area, of those 

 individuals which conform to the characters on which the 

 race is based. 



As I understand a subspecies, absolute constancy is un- 

 necessary and is indeed rarely seen. But what degree of 

 constanc)' is required? It must be a matter of opinion, 

 but I will arbitrarily take 75 per cent, as the necessary pro- 

 portion of birds which agree with the characters which 

 separate the race within a given area. 



Among the Wheatears in question I find that considerably 

 less than 75 per cent, have the characters assigned to the 

 various races, and as I believe that some of these characters, 

 such as density of colour on the mantle or amount of white 

 on the forehead, depend more on individual variation (or 

 perhnps age) than on geographical distribution, I am com- 

 pelled to unite them all under (Enanthe (xnanthe (icnanthe. 



Males from Europe. — Wings of males usually between 

 93 and 98 mm., rarely 99 or 100 mm. ; culmen between IR 

 and 17 mm. Birds from Great Britain average slightly 

 smaller. 10 males from Norway 95-97^ 100: 8 from Sweden 

 93-5-98; 11 from western Russia 91-99. 52 from Mace- 

 donia and Greece (teste Stresemann, Avif. Macedon.) vary 

 from 89 to 99 mm. Breeding birds from Greece do not 

 have an unusually pale mantle. Four birds from Portugal 

 have wings 93-98 and culmens 17"5-18 mm. One has a very 



