NATURAL SELECTION, ETC, 97 



patent facts tliat all species vary more or less ; tliat 

 domesticated plants and animals, being in conditions 

 favorable to the production and preservation of varie- 

 ties, are apt to vary widely ; and that, by interbreed- 

 ing, any variety may be fixed into a race, that is, into 

 a variety which comes trne from seed. Many such 

 races, it is allowed, differ from each other in structure 

 and appearance as wfdely as do many admitted species ; 

 and it is practically very difficult, even impossible, to 

 draw a clear line between races and species. Witness 

 the human races, for instance. Wild species also vary, 

 perhaps about as widely as those of domestication, 

 though in different ways. Some of them apparently 

 vary little, others moderately, others immodei-ately, to 

 the great bewilderment of systematic botanists and 

 zoologists, and increasing disagreement as to whether 

 various forms shall be held to be original species or 

 strong varieties. Moreover, the degree to which the 

 descendants of the same stock, varying in different di- 

 rections, may at length diverge, is unknown. All we 

 know is, that varieties are themselves variable, and that 

 very diverse forms have been educed from one stock. 

 2. Species of the same genus are not distinguished 

 from each other by equal amounts of difference. 

 There is diversity in this respect analogous to that of 

 the varieties of a polymorphous species, some of them 

 slight, others extreme. And in large genera the un- 

 equal resemblance shows itself in the clustering of 

 the species around several types or central species, 

 like- satellites around their respective planets. Ob- 

 viously suggestive this of the hypothesis that they 

 were satellites, not thrown off by revolution, like the 



