6o BARKER : 



pyris the crystalline cones in outline resemble short test tubes, 

 and the space between the convex inner ends and the sup- 

 posed retinal surface is filled by transparent substance which 

 does not materially interfere with the progress of the light 

 rays which form the image. 



We accept Exner's photograph as a fact, of course, but 

 may well question some of his inferences. Forced to admit 

 the great difficulty of getting such an image in the numerous 

 eyes whose cones are encumbered with long rods (Plate 2, 

 Fig. 8), threadlike prolongations, spindles, etc., he supposes 

 in such cases another kind of image which he calls an " appo- 

 sition " image. It is formed, he says, in those eyes whose 

 retinal elements lie close to the crystalline cones. "The 

 upright retinal image of the eye of Limuhis is formed as a 

 consequence of the beams of light pertaining to the indivi- 

 dual members of the compound eye falling upon the retinal 

 plane adjacent to one another." This is obviously the mosaic 

 image under another name. The image obtained with the 

 eye of Lampy^is, which he calls a "superposition " image, 

 was formed upon a retinal surface, in this case far from the 

 ends of the cones, and he believes that every point of the 

 image received light from at least thirty of the optical units 

 or individual members. He observes also that probably some 

 eyes may yield both apposition images and superposition 

 images. 



Dr. Dallinger seems to have misunderstood Prof. Exner, 

 from whom he had a private communication with a copy of 

 the Lanipyris photograph, for we find in the latest edition of 

 Carpenter's "The Microscope and its Revelations," 1901, 

 page 983, the following statement : — 



" After traversing the pyramids the rays reach the extremities of the 

 fibers of the optic nerve, which are surrounded like the pyramid by pig- 

 mentary substance. Thus the rays vrhich have passed through the sev- 

 eral ' corneules ' are prevented from mixing with each other, and no 

 rays save those which pass in the axes of the pyramids can reach the 

 fibers of the optic nerve. Hence it is evident that as no two ocelli on 



