226 Journal New York Entomological Society, f^oi. xxi. 



auxiliary vein, except in Acarthrophthahnns as noted above: where 

 the costa is twice broken the first fracture is near the humeral cross- 

 vein and the second at the end of the auxiliary vein. This tabulation 

 is not infallible, but gives the characters usually and typically found. 

 It is often a difficult matter to know to what subfamily of the smaller 

 Muscidse a specimen should be assigned. The tabulation may there- 

 fore be found useful in verifying an identification. 



In glancing over the enumerated species it will be noted that Agro- 

 niyza is by far the dominant genus, followed by Phytomyza, which is 

 a close relative. The remaining genera have but one or a few species. 

 The characters by which many of the smaller genera are distin- 

 guished, such as the number of fronto-orbitals or dorsocentrals, the 

 shape of the cheeks and the size of the eye, have a range of varia- 

 bility in Agromyza equal to that found in a series of the small genera, 

 but can not be utilized for the segregation of this complex genus. In 

 Agromyza, especially, the profile of the face can not be relied on too 

 strictly. Apparently the drying of the head vaults the mouth-opening 

 so that the epistome at times is projected more than at others and 

 thus in some specimens it becomes visible in profile. The shape of 

 the lunule may likewise change, according to the age of the specimen. 

 In this genus Agromyza it is often impossible to determine what are 

 varietal and what are specific limits. For instance, couplet 59, sepa- 

 rating melainpyga and scutcllata is unsatisfactory, containing char- 

 acters most trivial than some of the varietal characters given under 

 those species. Extreme variations, like the black orbona and the 

 yellow melampyga, would appear obviously distinct, but the other 

 forms exhibit gradations connecting these extremes in every particu- 

 lar. The discal cell varies from small to medium, the width and 

 shape of the front is inconstant, the maculation of the body and legs 

 ranges over almost the entire gamut of coloration in this genus, so 

 there is little that is tangible to use in limiting the species. Under 

 such circumstances my determinations can not be considered infallible, 

 but with determined European material before me for comparison the 

 identifications given may be utilized until the type material is com- 

 pared with larger collections of specimens. It is not unlikely that the 

 less ornate species are similarly variable and that the number of 

 described species will be materially reduced when much larger collec- 

 tions are studied. In several instances in the dichotomv I have 



