CECIDOMYID^E. 125 



they are, not a few other revolutions of a similar charac- 

 ter would impend, notably that of the genus Milesia, as I 

 have made clear in my Synopsis of the North American 

 Syrphidse. 



In brief, the classification, or rather the nomenclature 

 of this family is still in a more or less chaotic condition, 

 and will be so until much more study has been given to 

 the insects, especially those of America and Asia. As 

 Aldrich has said, there is no more fertile field in all dip- 

 terology for thorough, exhaustive, ethologic, morpho- 

 logic, taxonomic and phylogenic studies than among the 

 Cecidomyidse. But there are many difficulties in their 

 stud} 7 . Doubtless if Kieffer's principles of generic sub- 

 division are accepted, there are a host of new 'genera' to 

 be discovered in North America, for the permutations of 

 antennae, palpi and tarsi are by no means exhausted. 

 My advice is, however, for the non-specialist to go very 

 slowly indeed in adding to what may prove to be an 

 already unnecessary confusion. 



TABLE OF GENERA. 

 (For the most part in their wider sense.) 



Fourth longitudinal vein wanting. . . . Cecidomyinse. 

 Fourth longitudinal vein present (ocelli present) . Lestreminse. 



CECIDOMYIN^. 



1. First tarsal joint shorter than the following one. . . 2 

 First tarsal joint longer than the following one. ... 14 



2. Fifth longitudinal vein not furcate. (8) . . . Spaniocera. 

 Fifth longitudinal vein furcate. . . . . . 3 



3. First and third longitudinal veins approximated to the costa, dis- 



tinguishable \\ith difficulty; costa tomentose. . . . 4 



Third longitudinal vein distinctly separated from the first. 6 



4. Fifth longitudinal vein furcate near the middle. . . 5 

 Fifth longitudinal vein furcate at base, giving an appearance of 



two longitudinal veins. (See DD, p. 127.) 



