28 PROCEEDINGS ENTOMOLOGICAL SOCIETY 



It seems possible that the studies of the internal anatomy of 

 these flies upon which Mr. C. H. T. Townsend is at present work- 

 ing may eventually prove useful as an index to group relations. 

 But the mass of undigested facts, near facts and conjecture with 

 which he is at present deluging the devoted heads of his con- 

 fre" res will require an immense amount of elucidation, rearrangement 

 and generous elimination before becoming available for use. 



To conclude, there is great need of careful rearings of species 

 belonging to homogeneous groups, from known parents, for the 

 purpose of studying variation of structure, color and size within 

 the species and, failing which our knowledge of the true relations 

 of the Muscoidean flies will never extend much beyond its present 

 meager limits. 



In discussion, Mr. Malloch said that the tendency towards 

 variation in the number of macrochsetse in Diptera is more pro- 

 nounced in the higher groups, such as the Anthomyidse and 

 Tachinidse, where the number of those macrochsetse is much larger 

 than in the other groups such as the Tetanoceridse, Ortalidse and 

 Phoridffi. In those groups with a few thoracic, or leg macro- 

 chsetse there is but little tendency to variation, and as a rule their 

 number and situation is remarkably constant. When any dupli- 

 cation such as in the frontal bristles, occurs in the Acalyptrate 

 Muscidse, it is the almost invariable rule that the normal bristle 

 is reduced in size and moved from its usual position. When 

 variation, either duplication or reduction, in the number of macro- 

 chsetffi occurs it is almost impossible to place species of Tachinidse 

 by using the tables of genera in the published works on the family. 

 Too much weight has been placed on the length of the third an- 

 tennal joint as compared with the length of the second. This 

 also is a character that is prone to vary, as is also the length of 

 the hairs on the arista. The hairs on the eyes, while under normal 

 conditions easily discernible, in many species in the males, are 

 most difficult to detect in the females of the same species. This 

 applies not only to the Tachinidse, but also to the Anthomyidse, 

 in which family far too much use has been made of this character 

 in generic tables. In the latter family the generic tables in Willis- 

 ton's Manual are of little use for the identification of some groups. 

 About one-half of the species in Fannia (Homalomyia), for exam- 



