X4 OHIO BIOLOGICAL SURVEY 



the g&n&ra. Nephrocenis and Barr/ia) and the Conopidae on the other. The 

 relation to the latter tamily is indicated in the porrect and elongate 

 antennae of the genera from Milcsia on, and especially of Ccria in which 

 the antennae end in a terminal style. 



PHYLOGENY 



In this paper the arrangement is that given in Aldr.ich's Catalogue 



of North American Dipt era. 



Williston is probably correct in stating that the characters of the 

 immature stages should not be expected to furnish basis for a classifica- 

 tion to supersede that founded on the more diverse and complex charac- 

 ters of the adults. Nevertheless it seems to me that a careful study of 

 larval structures and habits in many genera .should throw light on funda- 

 mental relationships, and might change the classification to some extent 

 or, at the least, tend to confirm the position of some of the more doubtful 

 genera. Especially does this .seem promising in a group like the Syr- 

 phidae where there is such a dearth of good sub-family characters. 

 For example, according to Williston, the genera Paragus and Pipiza 

 are placed in the tribe Chilosini. These two genera are, in part at least, 

 aphidophagous as larvae and are structurally like aphidophagous larvae. 

 The other genera of this tribe, viz., Chrysogaster, C/ii/osia, Chalcomyia, 

 and Myiolepta live as larvae in decaying matter or on plant tissue, and 

 some of them at least {Chalcomyia) are rat-tailed. Now this would seem 

 to indicate that the former genera naturally belong with the 5;';^///;// where 

 the larvae are predominantly aphidophagous. 



It seems to me almost certain that a complete study of larval char- 

 acteristics and habits, together with what is known about the imagoes, 

 will make possible a satisfactory division of this large family into sub- 

 families, which is conceded to be impossible from a consideration of the 

 imagoes alone. 



As will be shown below the family characteristics are simple and 

 very definite. The species are also for the most part reasonably 

 distinct. But the presence of many vexing, intermediate forms makes it 

 almost impossil)le to break the family up into satisfactory sub-families or 

 in some cases to group the species into genera. 



Three large sub-families are rather definitely noted: (i) The .Sj7-- 

 phiiiac or group with the anterior cro.ss-vein of the wing before 

 the middle of the di.scal cell, the femora rareh' thickened; slender, 

 thinl\- ])ilose, more delicatespecies. (2) The Eristali/iac with the anterior 

 cross-vein near or beyond the middle of the discal cell, femora frequentl^^ 



