98 ENTOMOLOGICAL NEWS. [March, 



combining two such images into one, the stereoscopic pictures which 

 bring out the convexity of object provided we use each eye for the pictures 

 on its own side only. 



The mobility of the eyeball and the changes of the form of the lens 

 by the ciliary muscles are other factors assisting the perception of rotundity. 



Now, in the eyes of the insects all these accommodations to vision are 

 wanting. The eyes are immobile, either smgle isolated eyelets (ocelli) or 

 arranged in semi-globular clusters (compound eyes). The axes of vision 

 of the eyelets in one cluster are divergent, nearing parallelism only with 

 its next neighbor — eyelet. The form of the eyelet is not a globe, but a 

 cylindrical cone, or, a better idea would be furnished by expunging from 

 the globular eye a conical piece of the diameter of the cornea through 

 the centre to the retina. The sensitive retinal portion is for that reason 

 not so extensive as the corneal portion. The length of the cone is re- 

 versely proportional to the convexity of the cornea; the more convex the 

 cornea the shorter the focus, the shorter the cone, the more divergent the 

 axes of vision of two neighbor-eyelets, the more convex the cluster of 

 th;: eyelet and the smaller the number of eyelets in one cluster. The re- 

 verse holds good in the same manner; the less convex the cornea the 

 more numerous the eyelets and the less divergent are ihe axes of vision. 



The isolated ocelli seem to be more resembling the form of the verte- 

 brate eye. In the spiders they are arranged by fours in two transverse 

 curves on the front and on the vertex of the head, which is much more 

 movable than in the hexapodal insects and the visual axes by twos are 

 supposed to be almost parallel, consequently have a greater range of 

 vision. 



As we do not know the physiological action of the parts of the retina 

 in our own eye, much less we should speculate on the anatomy of the 

 insect eye. We do not know even the situation of the sensitive parts, 

 but we certainly know that the vision is very good, and the optical law 

 are as applicant as with our own eye, and that the sensitive part must be 

 situated at the end of the transparent part. 



The brainal mass is transverse, connecting the ocelli and eye-clusters. 

 One may suspect that there is a compensation of at least two neighboring 

 ocelli for the formation of a perspective connected image of an object. 



The spherical arrangement of the eyelets of a cluster as they are rep- 

 resented by the eyes of hexapod orders of insects, and in the fossil trilo- 

 bites necessitate the more conical form of the single eyelet with a very 

 limited range of vision and a divergence of the optical axes. If there is 

 no compensation of neighboring eyelets, or of eyelets of the two clusters 

 with parallel or convergent axes of vision as they really exist in large, 

 prominent clusters, or where the cluster occupy almost the whole head 

 the single eyelet would see only a small part of the object, and the several 

 partial images could not form a truly connected image of the whole object. 



The idea of unconnected vision was held by Johannes Miiller, and is 

 known as the theory of the mosaic vision, which is, I think, generally re- 



