Januaet 19, 1906.] 



SCIENCE. 



85 



scaly rootstocks and bulbose base are actu- 

 ally found in the same individual. 



The genus Hexastylis of Rafinesque has 

 been recently restored and separated from 

 Asarum on the ground that the former has 

 a superior ovary, distinct styles, and fila- 

 ments shorter than the anthers, while in 

 the latter it is said that the ovary is in- 

 ferior, the styles united, and the filaments 

 are longer than the anthers. This might 

 seem to be an excellent generic division 

 were it not for Asiatic species of Asarum 

 in which an inferior ovary is combined 

 with distinct styles and nearly sessile an- 

 thers. Examples of this sort might be 

 almost indefinitely multiplied. 



In all these cases the generic segrega- 

 tion is weak, ineffective and unscholarly 

 because it rests on incomplete observation. 

 It is, moreover, distressing to see what a 

 strong prejudice there is on the part of 

 many investigators against taking a broader 

 and more cosmopolitan view of plant classi- 

 fication. If genera appear to be good in a 

 particular region, that is regarded as quite 

 justifying their maintenance in floras treat- 

 ing of that part of the world. This tend- 

 ency toward insularity is to some extent 

 perceptible in the work of different Euro- 

 pean nations. It is far more noticeable in 

 the growing separation of the taxonomie 

 work of the eastern and western continents. 

 It is obviously the outcome of the narrow- 

 ing influence of specialization and should 

 be vigorously combated, for it is seriously 

 threatening not only the uniformity and 

 harmony, but the normal progress and dig- 

 nity, of our subject as a .whole. For it is 

 quite evident that were it to be once ad- 

 mitted that generic limits rest not upon 

 species of the world but are to be fixed in 

 each flora merely according to their chance 

 representatives in that flora, it would in- 

 troduce a chaos into classification as ridicu- 

 lous as it would be unnatural. Two genera 

 might be justifiable in the limited flora 



of Connecticut, which would completely 

 merge were the other New England species 

 considered; or two genera might present 

 marked difi^erences in New England, which 

 would intergrade in the middle or southern 

 states. It is obvious that if we are to ob- 

 tain any measure of uniformity or stability, 

 genera must be founded not on local, but 

 on a cosmopolitan basis. 



It has often been suggested that excessive 

 splitting of genera has been due to personal 

 vanity of authors who enjoy creating the 

 new binomials involved. However this 

 may be, I much doubt whether any such 

 unworthy motive has played great part in 

 the matter, but infer that many authors 

 share the view of Rafinesque that large 

 genera are unwieldy things; that it is a 

 very difficult matter to prepare a good key 

 for such large numbers of species and that 

 classification is rendered simpler and clear- 

 er if the genus can be broken up into frag- 

 ments of convenient size. In regard to 

 this it may be said that the subdivision of 

 the genus into subgenera or sections ac- 

 complishes precisely the same end with no 

 confusion of generic nomenclature. Dif- 

 ferent minds may work in unlike manner 

 when confronted by the difficulties of iden- 

 tifying plants. Personally, I should very 

 much prefer to have the difficulty at. one 

 point rather than at two ; that is to say, I 

 should rather have generic lines drawn so 

 widely that genera would be pretty definite 

 and readily recognized, in the manner, let 

 us say, of Cypenis, Astragalus or Eu- 

 phorbia in the broader and long traditional 

 sense. The recognition of such genera re- 

 quires little or no mental efi'ort on the part 

 of a botanist of any training. The atten- 

 tion is left free for the specific identifica- 

 tion and this may be undertaken with a 

 happy confidence that all the species likely 

 to come into question will be found in the 

 same group and under the same eenerie 

 name. These species may be inconvenient- 



