ABT. 8 REVISION OF THE FLY GENUS BELVOSIA ALDRICH 3 



their height, often much more; penultimate joint of arista distinct 

 and a little elongated, not genicidate; arista usually thickened for 

 most of its length, sometimes flattened nearly to the tip; cheeks uni- 

 formly a little less than half the eye height; palpi and proboscis nor- 

 mal. Thorax with usually four sternopleurals (from three to six or 

 more) ; scutellum with a row of long, depressed laterals, from four to 

 six pairs, and sometimes a smaller, suberect, spiny apical pair, which 

 are highly variable and poorly developed as a rule. The number of 

 laterals is somewhat variable also and of slight importance in sepa- 

 rating the species. Abdomen generally broad, never with discals, the 

 genitaha of the male rather small and showing but feeble specific 

 characters; they are of a common type, consisting of inner and outer 

 pairs of thick, short forceps, the penis elbowed and ending in a vase- 

 like opening. The wings are very uniform, third vein with only a 

 few bristles at base, apical cell open far before the tip. Legs with 

 good characters in the male pulvilli, which are usually enlarged, and 

 with the bristles of the outer side of the hind tibia variously developed 

 but subject to sexual variation in the species. 



The characters which seem most useful are the color of the small 

 hairs below the lowest frontals, and the extent of the pale pollen on 

 the third and fourth abdominal segments; other details of the pollen 

 of the abdomen furnish many minor points, and these seem to be 

 very constant in well-preserved specimens. I have omitted the tho- 

 racic chaetotaxy almost entirely, as it is too uniform to give any spe- 

 cific characters except in rare cases. A delicate character of some 

 importance in males is the shape of the upper curve of the compound 

 eye; where this is more broadly rounded the front does not begin to 

 widen for a short distance from the vertex; but in some species (bifas- 

 ciata, for example) the eye is narrow and almost pointed above, so 

 that the front widens immediately from the vertex. The pollen of 

 the parafrontals gives a number of good specific characters, especially 

 in males. The frontal bristles I view with suspicion, as they seem 

 highly variable. The length of the third antennal joint as compared 

 with the second has a considerable value, although some variation 

 must be expected. The pulvilli of the males offer several degrees of 

 development, and seem constant for the sex and species. 



The nearest related genus is perhaps Atacta Schiner, in which the 

 thoracic and abdominal structure are much like unifasciata; it, how- 

 ever, has a narrow front in the male, very fiat facial ridges, frontal 

 bristles reduced, and the ocellars are present. 



The reproductive habits are the same in all species as far as known; 

 the females lay large numbers of minute eggs on foliage which are swal- 

 lowed by caterpillars along with the leaf which they are eating. This 

 is the second group of Pantel, which includes Gonia, SpaUanzania, etc. 



The genus is known only from the New World. 



