VARIED RESULTS OF CROSSING 381 



is confessedly quite relative it is a term of convenience when we 

 wish to include under one title all the members of a group of 

 individuals who resemble one another in certain characteristics. 

 A species is often simply a segment of a curve of closely related 

 forms. It is a statistical conception, and as there is no abso- 

 lute constancy in specific characters, as one species melts into 

 another, with which it is connected by intermediate varieties, 

 by frequent or casual variations, we have to confess that it is 

 a human device, the validity of which varies greatly according 

 to our knowledge or ignorance of the forms in question. A specific 

 name is sometimes, when we are very ignorant, as unmeaning as 

 the name of a constellation in the starry heavens. But it is 

 equally convenient. 



At the same time, since science is systematised common sense, 

 it is usually admitted oftener, perhaps, as a pious opinion, than 

 as a practice that the characters on account of which a naturalist 

 give,s a specific name to a group of similar individuals should be 

 more marked than those which distinguish the members of any one 

 family, should show a relative constancy from generation to genera- 

 tion, and should be associated with reproductive peculiarities which 

 tend to restrict the range of mutual fertility to the members of the 

 proposed species (see the author's Outlines of Zoology, 5th ed., 

 1910, pp. 14-16). 



The popular impression that crosses between " distinct 

 species " are rare is erroneous ; for, apart from the familiar 

 mules, fertile pairing is known between lion and tiger, dog and 

 jackal, wild and domestic cat, brown bear and polar bear, 

 American bison and European wild ox, horse and 'zebra, hare 

 and rabbit, duck and goose, canary and finch, thrush and black- 

 bird, capercaillie and blackcock, carrion crow and hooded crow, 

 pheasant and fowl, and the list soon becomes very long if we 

 include backboneless animals and plants (see Evolution of Sex, 

 revised ed., 1901, p. 163). 



The popular impression that fertile crosses between " distinct 



