— 41 — 



• 



ill most Dipt era be much longer than the ommatidia, whereas 

 in Diopsidae it is, on the contrary, shorter than they are. 

 Consequently the ommatidia of Diopsidae have moved nea- 

 rer to their ideal centre from which they start than those of 

 other D i p t e r a, or, wiiich is less probable, they have elon- 

 gated proximally. The first case would result in a considerable 

 narrowing of the ommatidia, the second — at least in the narrowing 

 of their proximal ends. But here we meet with a special 

 adaptation which allows tliem to preserve their primary width: 

 that is a great diminution of the number of 

 ommatidia in the eye of Diopsidae. In all the 

 other Diptera with strongly convex eyes the number of om- 

 matidia is considerable. Thus, e. g., in D i e t r i c h's figures, on 

 a longitudinal section through the eye of MicrocUrysa there 

 can be counted 97 ommatidia in the male, and 92— in the 

 female, whereas in my figure of an analogous section (fig. 8) 

 only about 53 ommatidia are seen at about the same conve- 

 xity of the eye. Similar numerical relations are also to be 

 found in comparing with many other Diptera. Evidently, 

 such a small number of ommatidia in the eye of Diopsidae 

 allows them to occupy the position named without perceptibly 

 narrowing from that. If in the conditions given the number 

 of ommatidia were increased, their proximal end would gra- 

 dually narrow to such an extent that there would remain no 

 place even for the passage of the nerve-fibres. The existing 

 number of ommatidia gives room only for one ganglion of 

 relatively small size inside the cavity of the eye. Besides, it 

 is quite possible that the small number of ommatidia corres- 

 ponds to the small volume of both optic ganglia. It is also 

 noteworthy that the form of ommatidia in Diopsidae de- 

 viates somewhat; this is caused by the small number of them 

 in connection with the great convexity of the eye. If we com- 

 pare a separate ommatidium with a wedge directed with its 

 pointed end inside the eye, then in Diopsidae it termina- 

 tes more bluntly, i. e. its walls diverge stronger than in the 

 other Diptera. When the number of ommatidia is reduced 

 to a half, each of them — all conditions being equal — posses- 

 ses an interior angle twice-greater than when their number 

 is complete. In connection with the form of separate omma- 

 tidia of Diopsidae evoked by such circumstances, I should 

 like to mention another fact, that is that their retinulae and 



