478 



SCIENCE. 



[N. S. Vol. XII. No. 300. 



upon the present monograplier's views as 

 to the relationship of the species with others 

 included under other concepts named by 

 writers previous or subsequent to the date 

 of the genus under investigation. Thus, 

 to take a very simple case, if there were 

 a genus A described in 1830 with three 

 species of which a is nearest related to d, of 

 genus B, 1840, h is nearest related to e, 

 of genus C, 1820, while c is nearest related 

 to /, of genus D, 1850, we have already 

 under the method of elimination a series of 

 varying alternatives : 



1. If the genera B and C be deemed 

 valid, D cannot be separated, but is con- 

 sidered synonymous with A. 



2. The systematist who decides that B 

 is invalid applies ^ to a and d and may 

 recognize J> as a good genus. 



3. If C be treated as invalid A may be 

 applied to h and e, B and D being con- 

 sidered good. 



Thus while it may be theoretically pos- 

 sible for a monographer to arrange to his 

 own satisfaction the relations of the dif- 

 ferent genera, a change of taxonomic 

 opinion affects not only the supposed limits 

 of the genera but may necessitate a totally 

 different application of the name A to any 

 one of the three groups of species. And 

 when we reflect that the complications are 

 increased in almost geometrical ratio when 

 the species are more numerous and when 

 the question of the validity of B, C or D 

 may be subject to equally great complica- 

 tions from other aspects of their real or 

 supposed relationships, it becomes evident 

 that the conceptual method of elimination 

 involves an endless chain of casuistry, and 

 is a counsel of darkness and confusion 

 rather than of stability and perspicacity. 

 Moreover, in the lower plants and animals 

 the large composite genera of the earlier 

 writers are in manj^ cases now distributed, 

 not merely to different families, but even to 

 different orders and classes, so that the 



elucidation of some of the more difficult 

 cases of residual taxonomy would require 

 months of unprofitable labor in different 

 parts of the biological field, and yet the 

 conclusions could have only individual 

 sanction, no steps in the process being 

 secure with the exception of those which 

 deal with genera described as monotypic. 

 The designation of type species by a simple 

 and uniform method would, however, 

 render the application of all generic names 

 equally definite, and would largely elimi- 

 nate the personal equations which have 

 thus far added immeasurably to the labor 

 of biologic taxonomy, and which continue 

 to hamper all efforts to popularize the 

 science. 



Although, as previously noted, the Eoch- 

 ester Rules gave a tacit adherence to the 

 method of elimination, the case is not, in 

 reality, that of supplanting one method of 

 procedure by another, since with the pos- 

 sible exception of a small proportion of the 

 flowering plants the method of elimination 

 has never been consistently applied in any 

 part of the botanical series. Most botanists. 

 Continental, English and American, have 

 continued to deal with genera in a manner 

 purely personal and arbitrary. Seldom has 

 there been any formal recognition of a type 

 much less the choice of one by any fixed 

 rule. Genera have often been deprived of 

 all their original species and made to do 

 duty for an entirely new set, with or with- 

 out modification of the original description. 



The conditions obtaining in the earlier 

 genera of ferns have been investigated by 

 Professor Underwood, and found to be much 

 the same as in the Myxomycetes and Fungi, 

 while a brief excursion among the palms 

 reveals the persistence there of the spirit of 

 lawlessness. The genus Oreodoxa, for ex- 

 ample, was based on two species, one of 

 which is now placed in Euterpe, and the 

 other in Catoblastus, while the name Oreodoxa 

 has been applied without warrant to the 



