E. MAYR 381 



tioned by Grant (this symposium). The suggestion overlooks the 

 fact that the word species has not only the biological meaning of 

 a reproductively isolated population but also the purely formal 

 meaning "kind of," simply a classifying unit. The term "agamo- 

 species" has been used to designate "totally asexual populations" 

 but, as stated above, an "asexual population" is a biological 

 impossibility. 



The most satisfactory solution in taxonomic practice has been 

 a frankly dualistic one. It consists in defining the term species 

 biologically in sexual organisms and morphologically in asexual 

 ones. There is more justification in this procedure than a mere 

 pragmatic one. The growing elucidation of the relations between 

 genotype and phenotype also justify this approach. Reproductive 

 isolation is effected by physiological properties which have a 

 genetic basis. Morphological characters are the product of the 

 same gene complex. Once this is clearly understood, a new role 

 can be assigned to morphological differences associated with re- 

 productive isolation, namely that of indicators of specific dis- 

 tinctness. This permits the assumption that the amount of genetic 

 difference which, in a given taxonomic group, results in reproduc- 

 tive isolation will be correlated with a certain amount of mor- 

 phological difference. If this is true, it is permissible to conclude 

 from the degree of morphological difference on the probable 

 degree of reproductive isolation. To base this inference on genetic 

 reasoning is new; the method itself, however, of determining 

 empirically with the help of morphological criteria whether or 

 not a population has reached species status goes back to classical 

 taxonomy. This inference method is by no means a return to a 

 morphological species concept since reproductive isolation always 

 remains the primary criterion and degree of morphological differ- 

 ence only a secondary indicator, which will be set aside whenever 

 it comes in conflict with the biological evidence. 



It is possible to use the same kind of inference to classify 

 asexual organisms into species. Those asexual individuals are in- 

 cluded in a single species that display no more morpholog- 

 ical difference from each other than do conspecific individuals 

 in related sexual species. Criteria must be adjusted to individual 



