TAXONOMY of MUSCOID&VN FUKS — TOWNSEND 13 



ogeneous tribe or subfamily. The refractory elements should rather 

 be left to stand alone. 



In such manner as the above will it be possible to work out a 

 serviceable system of classification, which will indicate, so far as may 

 be, the true relationships, and at the same time preserve appr< »x- 

 imately the relative values of taxonomic divisions in the Cyclor- 

 rhapha. 



A very important point remains to be noticed: What is a species 

 in this superfamily? The preceding remarks on intermediate forms 

 apply especially to the higher divisions, but are also largely true of 

 genera and species. The difficulties as to genera can be practically 

 overcome by the erection of a sufficient number to accommodate all 

 the intermediates. But who can tell what is a species in nature, and 

 especially what is a species in the Muscoidea? It is clear that we 

 must have a definition that will answer to the term. In large assem- 

 blages of insects, where intergrades and intergradants have not been 

 lost, there is no such thing as a species in the generally accepted 

 sense. No sharp specific distinctions can be drawn in such cases. 

 The term is a necessary conception in taxonomy, however, and it is 

 to be noted that the only reason for its employment is the necessit} 

 for being able to distinguish between assemblages of individuals that 

 are unlike. Therefore it seems clear that the only safe course to 

 pursue is to give a name to every assemblage that can be distin- 

 guished from other assemblages. 



It is proposed to use the term "'species" in a well-restricted sense. 

 Typic species are already explained. The term atypic species will 

 be used for recognizable assemblages of individuals grouping around 

 typic species. The term "forms" may be used interchangeably as 

 referring to either or both. 



When two atypic species are connected by intergradant individ- 

 uals, the former should be given names and the latter referred to as 

 intergradants between the two atypic species. A few words of de- 

 scriptive matter will serve to fix practically the exact taxonomic 

 position of these intergradants. Such a course will afford students 

 of bionomics an opportunity to attain some degree of definiteness in 

 their investigations. As the names now stand in the Aldrich Cata- 

 logue, this element of definiteness is totally lacking. Many distinct 

 forms are bunched under one name on almost every page. Absolute 

 exactness is impracticable in this phase of nature, where variation 

 through pressure of environment is constantly at work in the evolu- 

 tion of new forms. But a reasonable degree of definiteness is possi- 

 ble of attainment. So long as we can refer by name to recognizable 

 forms, we may be certain that we are not going wrong. Such forms 



