THE CANADIAN ENTOMOLOGIST 41 



tures of the egg and first-stage maggot are, on the whole, of prime 

 taxonomic rank in the Muscoidae. They are therefore available 

 for family definition in the case of large groups or pronounced 

 types where other characters fail us. We may also justly con- 

 clude that the reproductive s^^stem and general egg and maggot 

 structures furnish characters of inferior rank but of great service 

 in the definition of such taxonomic categories as genera, group- 

 units, subtribes, tribes and subfamilies, and even at times of 

 families if they are supported by other important characters. 



A comparative study of plant and animal taxonomy suggests 

 (1) that the eggs, embryos, early and adolescent stages of animals 

 will always furnish us the main key to their affinities whether 

 such is present or lacking in the adult; (2) that the characters of 

 the reproductive system, while of less rank, will enable us to fix 

 definitely the limits of the lower taxonomic categories when their 

 definition is obscured in the adult; and finally (3) that the more a 

 structure becomes specialized, the more the taxonomic value of its 

 characters contracts. The first point justifies the erection of the 

 eleven families outlined and recognized in this paper. The last 

 point emphasizes again the extreme taxonomic difficulties that 

 exist in the muscoid flies, which are undoubtedly not only among 

 highly specialized but also among the most recently specialized 

 of all arthropods and hence the most difficult to classify in a con- 

 venient system. However much the values of certain characters 

 may contract, in other words however obscured may become the 

 group relationships in the structures exhibiting these characters, 

 we are nevertheless often compelled, in the absence of others more 

 distinctive, to use them if we wish to define certain of the higher 

 taxonomic categories. Thus, in order to attain the greatest degree 

 of clearness and practicability, we should in actual practice limit 

 our main group-definitions to the fundamental group-categories 

 or lowest groups of genera in these flies, which have been called 

 group-units. Each group-unit consists of the typic genus together 

 with those atypic genera which are found to be more closely related 

 to it than to any other typic genus. For definition of typic and 

 atypic genera, see Tax. Muse. Flies, p. II.; and for many pertinent 

 considerations, pp. 7-13. As an example, Exorista may be taken 

 as a typic genus, and Euphorocera as an atypic genus belonging 



