86 



SCIENCE. 



[N. S. Vol. XXin. No. 577. 



ly numerous, but at least one is not dis- 

 turbed by any lurking doubt whether, after 

 all, he has got the right genus. 1 can see 

 no real simplification in having Mariscus 

 and sundry other vaguely marked sections 

 of Cyperus treated as separate genera, so 

 that in identification one must first struggle 

 with the generic key and then, haunted by 

 some misgivings as to his success in this 

 matter, proceed to the specific key. As I 

 have said, minds may act quite differently 

 in this regard, but clarity of classification 

 is after all a very artificial consideration 

 and it is dearly bought if secured at the 

 expense of misleading statements. There- 

 fore, I can not believe that a writer when 

 dealing with a limited flora is justified in 

 stating that two genera are distinct if he 

 is wilfully or carelessly neglecting their 

 more or less complete intergradation in 

 some other flora. 



It may be thought that the sort of world- 

 wide perspective which is here advocated is 

 a thing well nigh impossible; that a writer 

 who attempts a flora, for instance, of the 

 United States will have his hands quite full 

 with the difficulties of his already hercu- 

 lean task and that his progress would be 

 slow indeed were he forced to estimate each 

 generic difference by the large standards 

 of the world's flora. Here I must use a 

 much-dreaded word, quite a bug-bear of 

 many a self-reliant scientist, namely, au- 

 thority. It is indeed quite impossible for 

 any one of us to repeat the acciimulated 

 work of many decades which has led to 

 the recognition of hundreds and thousands 

 of genera. We must accept at least a large 

 part of them on authority. The absolute 

 necessity of this is so self-evident that it 

 needs no justification. I am not advo- 

 cating any slavish submission to the opinion 

 of others, nor minimizing the importance 

 of verifying and in every possible way cor- 

 recting previous work, but merely urging 

 a ■vyjiolesome, resjpect for the opinion of 



those whose intensive monographic work 

 or broad cosmopolitan outlook has given 

 them an exceptional opportunity to see 

 just where generic limits may be most nat- 

 urally and profitably drawn. 



Fortunately, the genera of flowering 

 plants have been comprehensively treated 

 in two or three works of high excellence. 

 These sources of information are suffi- 

 ciently complete and recent to be of the 

 greatest service to writers on restricted 

 floras. In addition to such general guides 

 there exist for many families very detailed 

 monographs embodying the expert opinions 

 which are the result of long and critical 

 study of all members of the particular af- 

 finity concerned. It is not to be denied 

 that such monumental works exhibit a con- 

 .siderable measure of diversity in the inter- 

 pretation of genera, but this is not wholly 

 a misfortune since it permits a certain 

 elasticity in classification and enables a 

 certain selection according to varying judg- 

 ments. Furthermore, however great the 

 divergence may be in these large mono- 

 graphic and cosmopolitan works, it is rela- 

 tive uniformity compared with the state 

 which would soon obtain were genera to 

 be defined in each flora solely according to 

 its local representatives. 



One of the most unhappy tendencies ob- 

 servable in modern classification is a grad- 

 ual letting down of standards, a feeling that 

 if a few ill-defined genera are to be found 

 in a particular family, the others should in 

 the interests of a sort of specious symmetry 

 be cut up until all are about of the 

 same degree of vagueness and uncertainty. 

 When an author who tends to excess in 

 dividing genera feels called upon to assign 

 a ground for his action, it is in .nearly all 

 cases that the segregates he is making are 

 quite as good genera as many which already 

 exist. 



This process of taking the poorest exist- 

 ing work of others for a guide or as a 



