No. 496] ASPECTS OF THE SPECIES QUESTION 241 



do not fully agree with any one of his predecessor's de- 

 scriptions, or he finds that some of his specimens agree 

 about as well with one of the groups recognized by the 

 previous author as they do with another. This result 

 shows conclusively that for practical taxonomic pur- 

 poses it is not desirable to attempt to define a great many 

 of these minor groups. The tendency has been brought 

 about, I believe, by the instinct of many investigators 

 that everything in nature must be named and described, 

 but nothing is to be gained by permitting this laudable 

 purpose to run to extremes. 



It is evident, I think, that our taxonomy has been 

 based on the fundamental error that the plant world 

 is to be regarded as divisible into smaller and smaller 

 groups, rather than following nature and proceeding on 

 the theory that it is built up of greater and greater ones ; 

 the science should be synthetic rather than analytic. 

 The synthetic theory will give our observation and ex- 

 perimentation a different significance and enable us to 

 comprehend some of the phenomena now masked by the 

 analytic method of attack. 



If, as now seems more probable than a few years ago, 

 sx>ecies are made up of elementary species, or races, and 

 that these are being increased by mutation, there can be 

 no end to the number of such groups produced. As 

 to the designation of these groups, I suggest that the 

 term race be employed. This has long been used to 

 designate what have been called self-perpetuating vari- 

 eties, which appear to me to he identical with the present 

 conception of elementary species, and its application may 

 readily be widened. The term variety loses its signifi- 

 cance, because it is usually quite impossible to tell* how 

 any given individual or group of individuals has arisen, 

 or from which species it has sprung. The term form could 

 be used instead of either race or elementary species, but 

 it has had such a trivial significance in literature that 

 race seems to be preferable. Subspecies implies divisi- 

 bility, and is, therefore, an undesirable term. 



