SPECIES AND VARIETIES: THEIR ORIGIN BY MUTATION. 165 



from its allies in the garden is to be considered as an elementary 

 species." 



Systematists are often obliged to lay more stress on differences of 

 structure, while De Vries places emphasis on constancy. Any offspring 

 which show perpetual variations are individual or fluctuating varieties, 

 and so cannot form elementary, fixed, species. 



What, then, causes constancy or fixity of characters ? 



Now De Vries does not seem to realise or emphasise enough the 

 necessary conditions to secure subsequent stability of a variation of 

 structure. He seems to regard it as an innate quality, per se. But 

 practical horticulturists know the necessity, first, of isolation, then of 

 selection, generation after generation, till a high percentage is maintained 

 if absolute identity cannot be secured. Moreover, they must be grown in 

 the same conditions. Thus Professor James Buckman observed : " Experi- 

 ments with seeds of plants showing any particular tendency, and especially 

 if repeatedly grown in the same soil, will ever result in an increase of the 

 peculiarity." * As an illustration, it is now known that the " long" and 

 "short" or "turnip" forms of radishes, carrots, rape, and turnip are 

 the results of sowing the seed in loose and compact soils respectively, yet 

 by selection these forms have become hereditary. 



Though " everything tends to become hereditary," as M. Carriere 

 observed, there is every degree between a quickly-acquired constancy 

 and none at all, even with the aid of the most careful selection possible. 

 The latter are ever-varying, fluctuating forms, which seed merchants can 

 only issue as "mixed," as De Vries observes. When De Vries raised his 

 elementary species which he called gigas and rubrinervis, he probably 

 grew them year after year under precisely the same conditions wherein 

 constancy was to be expected. 



Nature affords numerous analogous instances in any genus with a 

 large number of species of which many are sure to be local and charac- 

 teristic of the places in which they grow, as e.g. the subspecies of Poly- 

 gonum aviculare as described by Sir J. D. Hooker in his " Students' 

 Flora of the British Isles." 



If a form has apparently lost some character of the parent, or it has 

 become latent, or if it exhibits some feature already known in an allied 

 species, such De Vries calls a "variety" (p. 141). As examples he 

 mentious the " wheat-eared carnation " and the " green dahlia," as both 

 have lost their flowers ; but the former, at least, can reproduce them. 



Species, therefore, are "progressive" in acquiring something new; 

 varieties are " regressive " in having lost something (p. 15). He mentions 

 Pyrola and Honotropa as examples, as they are polypetalous ; but as 

 the polypetalous state undoubtedly preceded the gamopetalous, it might 

 be regarded as atavistic or " degressive," as De Vries calls it, being 

 probably brought about by the tendency to degradation through sapro- 

 phytism. 



A difficulty here strikes one at once ; for both progressive and regres- 

 sive features may occur in one and the same flower or plant. Thus, in 

 several cultivated Composites, the florets have acquired strange forms in 

 their corollas, unknown in the wild state or when first cultivated, as in 

 * Treasury of Botany, s.v. ' Brassica.' 



