The Bio7iomics of South African Insects. 527 



Lytta mocsta and all the species of Epicauta are purely 

 terrestrial in their habits, all being very common insects 

 and evidently protected by their vesicating properties. 

 The blue-winged Pompilids when running on the ground 

 bear a distinct general resemblance to these insects. Of 

 the five species of dark-winged flies, Exoprosopa ^cmhrosa 

 is the only one which visits flowers, and this lias a rapid 

 flight. The two Lapari and Orcctocera {Paraphania) 

 diaboliis are to be found settling conspicuously on lov/ 

 plants, and have a comparatively slow flight. Tabanus 

 biguttatvs is a rapid flier, with the usual habits of the 

 genus. The Zygrenid moth, Tascicc homochroct, is a very 

 common species ; it is occasionally seen on flowers, but 

 is more frequently observed settling in a consiDicuous 

 manner on shrubs and plants. It is very sluggish in 

 habits and emits a strong-smelling liquid when crushed. 



B. Mimicry of Diploptera by Fossores and Insects of other 

 Orders. (E. B. P.) 



The dominant members of the group described above 

 are evidently the Fossores, the Diploptera and Antho- 

 phila being represented by only two species each. 



It is therefore somewhat surprising to note that in one 

 of the Scolias (Figs. 6, 7, 8, Plate XX), evident mimicry of 

 Diploptera has been brought about in a very interesting 

 way. The outer parts of both wings, except the costal 

 area of the fore-wing, have become extremely transparent 

 and invisible, possessing a peculiarly illusive quality pro- 

 bably due to approximation of the refractive index to that 

 of air. The eti'ect is that both in flight (Figs. 6, 8) and 

 at rest (Fig. 7) the resemblance to the much narrower 

 wings of Diploptera is most striking. The resting resem- 

 blance is peculiarly interestmg, for the dark areas of the 

 wings exactly resemble the narrow longitudinally-folded 

 organs of the Diploptera, as may be seen by comparing 

 Fig. 7 on Plate XX with 17 on Plate XXI, or 11 and 13 

 on Plate XXII, while the transparent portions of the Scolia, 

 although in two layers (compare Figs. 1, 4, 10, etc., on 

 Plate XX), are not seen. The transparency is well shown 

 in Fig. 8, Plate XX, where the dark apex of the wing of 

 Soolia fratcrna was made to underlie the transparent part of 

 that of Scolia alaris, when the photograpli was taken for 

 the preparation of the plate. The underlying wing loses 



