WERNER MARCH AND 175 



to assume an oblique position which leaves a triangular space in 

 the middle of each segment. This alone should indicate that 

 where we have only triangular spots present there must have 

 been originally a whole pattern of the type described. In T. 

 exul (fig. 8) we find white triangular spots in the middle of each 

 segment, while the subdorsal and lateral spots are still traceable 

 in the form of ill-defined shades. Probably there are related 

 species which will show them still present. In Tabanus mel- 

 anocerus (fig. 9) the white triangular spots are well marked on 

 all segments while the remainder of each segment has become 

 infuscated. However, dark spots bordering on the white tri- 

 angles are still plainly visible and even a second row of dark 

 spots is present; between the two a slightly lighter area is in- 

 cluded wdiich corresponds to the subdorsal light band of forms 

 like T. lineola (fig. 5). In Tabanus trimaculatus (fig. 10) no 

 traces of subdorsal or lateral spots are seen, the whole abdomen 

 is uniformly black, but the presence of the typical three white 

 triangular spots on segments three, four, and five requires no 

 further explanation; they are the last remainder of the abdominal 

 pattern of Haematopota and of T. sufis. There are a great many 

 Tabanids in which two or three such white triangles on an other- 

 wise dark abdomen are a conspicuous character (note T. bi- 

 guttatus Wied, a common African cattle-pest). Sometimes the 

 triangles are combined with oblique spots on segments one and 

 two, as already indicated in T. lasiopthalmus (fig. 7). From the 

 black abdomen bearing triangles, it is not a long way to a wholly 

 dark abdomen, as in T. stygius (fig. 12) a form related to T. tri- 

 maculatus. Forms like T. fuscopunctatus (fig. 11), a Florida 

 species nearly related to T. exul, will illustrate the gradual dis- 

 appearance of the white triangles, leading to uniform black. 

 Needless to say that black as a final result may be reached by 

 different processes of pattern reduction, so that in cases where the 

 abdomen is simply black, we cannot always tell exactly in which 

 way the pattern actually has disappeared. Rut in forms like 

 T. stygius for instance, there is a good deal of likelihood that the 

 black final stage was preceded by a stage marked with the tri- 

 angles which are still present in the near relatives, and the same 

 thing may be said of T. giganteus when compared with its rela- 

 tives T. exul and T. fuscopunctatus. 



TRANS. AM. ENT. SOC, XLIV. 



