ORGANIC FORM AND ARCHITECTURE 707 



might originally have been but varieties of a few central ones; and 

 indeed these have been increasingly used to mark, and group, 

 "sub-genera" accordingly. Then, too, with increasing knowledge 

 and larger collections, forms appear with intermediate characters 

 between what at first seemed distinct and easily described species; 

 and this happens in some cases so freely that willows, roses, 

 brambles, hawkweeds, etc., have remained perplexing to syste- 

 matists from Linn^'s times and earlier ones to our own, though now 

 fresh light is forthcoming. 



Systematists, however, throughout their work, still have in the 

 main no choice but to keep by the distinct forms they can observe, 

 describe, and identify, and not speculate about ancestries they do 

 not know: so here it is worth recalling that Darwin left a legacy 

 towards continuing the Index Kewensis, the Kew Gardens list of 

 their new plants, thus substantially continuing Linn^'s herbarium 

 descriptions. We are thus ready to utilise both conceptions of 

 species; for practical purposes still that of characteristic form, and 

 generally with much the same habitat and way of life as well; and 

 secondly that of common ancestry, as we can make this out — a 

 task probably often now impossible, always extremely difficult, and 

 at best only beginning. Thus Johannsen, one of the foremost plant- 

 breeders, who first produced "pure lines", calls the taxonomist's 

 formal and apparent species phenotypic species, and the true genetic 

 ones which he and others now seek to identify, genotypic accordingly. 

 That most enthusiastic of geneticists, Lotsy, goes so far as to 

 describe the former as little better than special creations of Linn^, 

 and thus as "Linn6onts"; while the still more precisely described 

 varieties of Jordan (up to about 200 in the common Uttle cruci- 

 ferous weed, Draba verna, albeit a single species to ordinary eyes as 

 distinct as any) he similarly honours, and derides, as " Jordanonts" I 

 For other species than verified genotypes, he has thus Httle use; 

 yet we must go on as we do until we get them. Indeed, we take it 

 that the most of Linne's species, and maybe most of Jordan's 

 varieties too, will be justified anew; though cleared of mere hybrids 

 and of forms of different breeds and races merely convergent, 

 through adaptation to similar conditions. In this direction the 

 Mendelian breeders are now making remarkable advances; but more 

 especially in detail, and thus as yet doubtfully helpful in illuminating 

 the origin of species, let alone of larger variations upon the genea- 

 logical tree. Here great hopes have been aroused by De Vries's 

 Mutations, for him clearly important enough to rank as new 

 species, and even to throw Hght on their distinctiveness, as veritably 

 "explosive" changes into new forms, persistent until a fresh muta- 

 tion may occur. He thus throws aside, as practically of no evolu- 

 tionary value, all those minute variations on which Darwin and his 

 successors, even to Weismann included, have so much depended, 



