295 



6. Dorsal bristle with aiiieal part rounded F. texanus* 



— Dorsal bristle with apex pointed F. pergandei. 



FVFM 



1. Thoracic respiratory organs shoe-shaped C. fusculus. 



— Thoracic respiratory organs drumstick-like , .2 



2. Abdomen with a pair of spines on segments 2 to 4, those on the lat- 



ter much shorter than the other pairs F. wlieeleri* 



— All the segments which are exposed beyond the larval exuvia with 



spines, which are not limited to one pair on each segment 3 



3. Bristles on abdomen very short and inconspicuous; no spine on 



head { F. hrumalis* 



\ F. cilipes. 



— At least some of the bristles on abdomen long and conspicuous. . . .4 



4. A distinct spine present on either side of head 5 



— Head without a distinct spine F. stenammatus* 



5. Dorsal and dorso-lateral abdominal bristles very unequal in size; 



ventral surface of abdominal segments 3-5 with a short spine 

 at middle F. specularis. 



— Dorsal and dorso-lateral abdominal bristles not noticeably different 



in size; no ventral spine present F. pergandei. 



CuucoiDES Latreille 



Culicoides has generally been distinguished from Ccratopogon, 

 in the restricted sense, by the small size of the empodia as compared 

 with the size of the tarsal claws. In no species which I have retained 

 in Culicoides have I found distinct empodia, and but for the fact that 

 my material is rather scanty I should not hesitate to indicate their ab- 

 sence as a character for distinguishing the genus. I have, however, 

 found a character by which the genus may readily be separated from 

 its allies, which is both easily appreciable and of real phylogenetic 

 value. This character is the presence, on the anterior portion of the 

 thoracic disc, of a pair of distinct cavities or depressed areas (PI. 

 XXIII, Figs. 1-3) which I believe are sensory organs of some func- 

 tional value to the species. These cavities are remarkably conspicuous 

 in all species having the disc of the thorax gray pruinose, and are very 

 easily detected even wath a low-power lens. I have dissected several 

 species in an unsuccessful effort to discover whether they are con- 

 nected with tracheae. I have, however, established the fact that the 

 surface of the oval or slitlike cavity is either finely perforated or has 

 many minute areas of extremely thin membrane, which fact seems to 

 justify the opinion that they are functional. Most of the species of 

 Culicoides whicli I have seen have, like pulicaris Linne, spotted wings, 



