xi. c, 4 Copeland: Natural Selection 159 



of those distinctive forms that have endured for a long time, 

 and to recognize these, and these only, as species, then natural 

 selection would be responsible for them all. 



Obviously, our opportunities, our judgment, and our conven- 

 ience all combine to make us adopt a middle course. A very 

 considerable proportion of new species are described from single 

 plants. It is far from unheard-of for two species to be de- 

 scribed from a single plant. In the groups that vary widely, of 

 which I presume that the genus Crataegus is one, it is probably 

 not unusual for more than one species to be described from 

 the progeny of a single plant. Up to this time, such a practice 

 has not been adopted, except in cases where the parentage is 

 unknown. If the same freedom of species-recognition and de- 

 scription were practiced with various cultivated plants, for ex- 

 ample, tobacco, where the parentage is known but the offspring 

 vary beyond the bounds that would be recognized as specific 

 among wild plants of unknown parentage, there would be almost 

 no limit to the number of species; but systematic botanists 

 have so far mercifully abstained from doing this. 



The responsibility of natural selection for the species recog- 

 nized at any moment in a given place depends then very largely 

 upon what we recognize as a species. The possible origin of 

 species by the summation or selection of variations, whether 

 slight or great, is another question to which the answer is fixed 

 chiefly by our choice of definitions. In so far as the species 

 originates by a single variation — which may always be true, if 

 we define species in that way — natural selection is never respon- 

 sible for its immediate appearance. Even if we go as far as 

 the veriest determinate-variationist might, and assume that 

 species N has been derived from the species A, through B, C, D, 

 E, etc., all steps being in the same direction, and the most of 

 the steps being individually short enough to escape our notice, 

 still each of these intermediate forms by the definition just sug- 

 gested is itself a species, and N, in its turn, originates by one 

 variation from the different parent species M. 



That the species which occupy the world to-day have originated 

 by the selective accumulation of relatively small differences, it 

 is not worth the mutationist's while to deny. None of them will 

 waste his time looking for an Angiosperm as the mutant of a 

 Flagellate, nor will any mutationist be disposed to deny that 

 between these extremes there has been a large number of inter- 

 mediate steps. If we agree with Doctor Willis that the only 

 difference between little steps and big steps is one of degree, 



