710 LIFE : OUTLINES OF GENERAL BIOLOGY 



the hooded or grey crow (C. comix), for while the ash-grey of part 

 of the plumage is very distinctive, we know that the bird interbreeds 

 freely with the carrion crow, and that the hybrids are fertile. So 

 ornithologists ask one another whether the hooded crow is really a 

 true species by itself or only a race within the carrion crow species. 



In many cases there is no practical difficulty in regard to the 

 limits of a species. Thus in the Northern Hemisphere there is only 

 one species of solan goose or gannet, Sula bassana; and although its 

 plumage changes considerably in the course of its life, no one pro- 

 poses to establish sub-species. Similarly, there is in Britain only 

 one species of kingfisher. But with many other "kenspeckle" birds, 

 the difficulty arises that the species often includes sub-species, as 

 in the case of starling, song-thrush, and golden eagle (to take three 

 diverse instances) ; or that a country has several related species of 

 the same genus, which are not always very easily distinguished, as 

 might be illustrated by gulls, crossbills, and sandpipers. Since 

 evolution is going on, there is nothing surprising in the occurrence 

 of variations within a species, but what we are now referring to is 

 the frequent occurrence of sub-species or races which breed true 

 and form recognisable smaller groups within the species. There is 

 nothing clear-cut which enables a naturalist to say that such and 

 such distinctions warrant a separate species name, while others do 

 not warrant more than a separate race or variety name. 



Yet there are several common-sense criteria of a "good species", 

 which work out well when taken together as far as opportunities 

 permit. First of all, the species characters should exhibit a con- 

 siderable degree of constancy from generation to generation. Thus 

 very variable characters, like colour and markings, are not to be 

 relied on confidently. It is difficult to get two ruffs or two buzzards 

 which are anything like identical; and yet there is the counter-fact 

 that in other cases a minute feature may be quite decisive. It is 

 sometimes possible to distinguish one species of fish from another 

 by a few scales, or one species of bird from another by a few feathers. 

 A fox can be distinguished from a wolf by its blood-crystals. 



The classifier is ;Uways pleased when he discovers some particular 

 feature that can be relied on as a test, but, apart from these reliable 

 clues, he knows that the characters on which he takes it upon him- 

 self to give a group of similar animals a new species-name must be 

 bigger than those which distinguish the members of a family, using 

 the word here to mean the progeny of a pair. It would never do to 

 emphasise as a species-character some little detail of plumage, 

 e.g. in the Red Grouse, for this might lead to the absurdity of 

 putting two members of one family into different species. Absurd, 

 we say, yet it has repeatedly happened in a study of newly dis- 

 covered animals that the male has been made the basis of one new 

 species, and the female of another I The cautious classifier would 



