424 C. R. OSTEN SACKEN'S PRODROME 



often spoils the specimens (and this circumstance, in many instances where I had but a few 

 sf)ecimens of a species, has prevented me from ascertaining the coloring of their eyes). 



The coloring of the antennae, from red to brown and black, is, in most cases, character- 

 istic, but cannot be relied upon in every single specimen ; when, for instance, the atten- 

 uated portion of the third joint, before the annulate portion, is infuscated, the extent of 

 the brown is apt to be very variable ; the two basal joints also often vary from reddish to 

 dark brown or black in diflerent species. The same may be said of the extent of the dark 

 color on pale colored legs ; even the femora undergo, in some cases, the most unexpected 

 variations. Besides, the bristles and hairs covering the legs often assume different hues, 

 according to the direction of the light fiilling in upon them. 



The pollinose, or denuded condition of the sithcaUus (the part of the front between the 

 callosity and the root of the antennje) is, according to Loew, not an altogether trustworthy 

 character, as, in some European species, for instance, T. hiridus, it occurs adventitiously ; 

 while in others it is of a pretty constant occurrence. In the American T. lasio^ihthalmus 

 and rhombicus I always found the subcallus denuded. 



The coloration of the eyes of Tabani consists of crossbands, usually green (or bluish) 

 on purplish ground, or purplish on green ground ; sometimes all but one crossband disap- 

 pear ; occasionally the whole eye is unicolorous. I have no doubt that a more attentive 

 study of the coloration would, in some cases, supply useful hints as to the distribution of 

 the species in groups. The following instances of coloi-ation have partly been seen on the 

 living insect, partly on alcoholic specimens ; in many cases the eyes have been artificially 

 revived. 



1. T. turhldus, tr'tjunctus, catenatus, abdominal is [\ar.), tener, tectiis, variegahis (prob- 

 ably also T. fronto), have their eyes colored on the same pattern, that is, there are two 

 subparallel green bands on a purple ground, the lower one, before its outer end, bending 

 towards the upper one. T. molestiis and irimacidatiis show nearly the same pattern, only 

 the crossbands are more nari'ow, more parallel, and with a broader interval. All these 

 species seem to form one natural group. 



2. T. fuscojyiinctatus differs from all these species in having but a single narrow purple 

 stripe on green ground, the lower portion of the eye being again purple. 



3. T. psammo2)hilus has, on blue ground, two broad green stripes, the broad interval 

 between which contains a third much narrower green stripe. 



4. T. costalis and nigrooittatus have light green eyes, with a single narrow puij^le 

 crossband. T. fidvidus and sagax, evidently related to them, have besides this single cross- 

 band in the middle, two others, one above and one below, but less distinctly marked. 



5. The eyes of T. Uneola have a pattern somewhat akin to the preceding, but decidedl}'- 

 original. The eye is green, with a purple central band, and another above it, which is 

 abruptly interrupted about the middle of the eye ; the whole upper, outer and lower orbits 

 of the eye have a dark purple border. 



6. T. nigrescens and stijgvus have dark colored, greenish eyes, with a central, rather 

 dark purplish crossband, and usually less distinct bands above and below ; the bands are 

 not half as broad as their intervals. 



