INTRODUCTION. 



The number of genera of the family Picidce is 50, every known 

 genus being represented in the British Museum. The number of 

 species and subspecies recorded in this volume is 3S5, of which 36 

 are still wanting to complete the Collection. Of these 385 species 

 the Museum possesses the types of 75, and the number of specimens 

 in the Collection at the present date is 7894. 



As a rule, the species of Picidce are very clearly defined, but in 

 Bendroco^jus, Picoides, and Colaptes there is a decided tendency to 

 subdivide into races. Most of the latter have an exact range, and 

 it is only in the case of the North- American Colaptes, 0. auratns 

 and C. mexicanus, that there seems to be any definite appearance of 

 hybridization (see my remarks, infra, p. 22). This leads us once 

 more to the consideration of the question of " what is a species " ? 

 It is generally admitted that birds not iutergrading into an allied 

 form are species ; but where intergradation takes place the allied 

 form is a subspecies or race. The regions, however, likely to 

 produce these intermediate forms may not have been explored, and 

 such forms may, and most probably do, exist in certain localities ; 

 therefore, if this be the case, are these birds to be regarded as 

 distinct species merely because these links are not in evidence ? 

 Island forms may or may not possess some slight dififerences from 

 typical birds, but not sufficient to separate them ; yet some authors 

 take it for granted that with insular separation there can be no 

 intergradation : therefore we find insignificant islands made to 

 father a host of indifferent species or subspecies ; but I fail to see 

 (if perfect intergradation be insisted on in determining the position 



