120 



ON THE USE OF THE TERM " VARIETY " 

 IN SYSTEMATICS. 



BY 



L. J. Sedgwick, F.L.S. 



The remarks which follow are based on practical experience as a 

 field worker only, and the writer hopes that readers of the Journal 

 will accept this limitation when considering them. The main ques- 

 tions brought up for discussion are : (l) whether the term variety as 

 used in our Floras is applied to one natural phenomenon only, or at 

 the worst to several phenomena which are perfectly homologous, and 

 if not, then (2) whether there is any way of separating out the various 

 phenomena hitherto confused under the one term, and assigning to 

 each a separate term ; since it is clearly unscientific to use one and 

 the same term for phenomena which are heterologous, 



It is usually assumed that in all taxonomic work the personal 

 equation enters largely, and cannot be eliminated. It is open to 

 question whether the latter part of this assumption is correct so far 

 as species and units lower than the species are concerned. So far as 

 concerns genera, families and all units above the species, since these 

 taxonomic divisions are based on assumed descent, and since for the 

 descent of plants our only material is the very fragmentary palasonto- 

 logical record, it is clear that our results must largely depend upon 

 guess work. And it is for that reason that the tendency of even our 

 deepest systematic thinkers to allow their generic divisions to be 

 (sometimes at any rate) influenced by considerations of convenience is 

 a comparatively venial sin. In the case of species however and all 

 intra-specific units — excluding of course extinct species — there is no 

 possibility of pleading lack of evidence. If into our discrimination of 

 species the personal equation enters the fault is ours. We are 

 hampered by weakness of power of perception, by shortage of workers, 

 by lack of time, and by idiosyncrasies that could be eliminated. But 

 the truths are there if we could only unravel them. In the case of 

 most " critical" species the number of available individuals is legion. 

 Mendelian and other experiments could be carried out. Even if the 

 species or group of species is in an active state of evolution at 

 the moment its different forms and developmental tendencies could 

 be enumerated and charted. For instance, to take an imaginary case 

 of an Indian genus believed to be in an active condition of 



