562 THE SENSES AND SENSORY ORGANS. 
spherical cornea, and Exner has not even shown that the 
plane of the retina corresponds with the theoretical plane in 
which such super-position could occur. 
Moreover, Exner’s ‘super-position theory’ is admitted by 
him only to be applicable in certain rare instances where he 
assumes the dioptron to be an optically homogeneous vitreous, 
an assumption which I do not think is ever justified. Exner, 
therefore, falls back on what he terms the theory of images by 
‘apposition,’ in which he assumes that each ommateum acts 
independently. 
If Exner’s view is accepted with regard to the condition of 
the light-rays after leaving the lens or refractive cone, I can 
see no difference in this view and Miiller’s theory of ‘ mosaic 
vision.’ If, however, instead of parallel pencils the foci of 
converging pencils fall upon the receptive surface, Exner’s 
apposition theory and mine are identical, and the retina 
receives a second image which is not inverted. 
Exner compares each ommateum to an astronomical tele- 
scope, but he does not apparently see that if the focal lengths 
of the refractive agents are slightly altered it is possible that 
a second real image may fall on the receptive surface of each 
ommateum when his theory and mine are identical. 
Exner chiefly worked with eyes in which the crystalline cone 
is well developed, whilst my observations have been principally 
directed to eyes in which there is no crystalline cone properly 
so called. 
If we admit Exner’s theory of the formation of an image by 
a refractive cylinder, and suppose that the rays would leave 
the cylinder as parallel pencils, when the cylinder is a real 
cylinder and the first image lies midway between its ends, and 
then add a spherical surface to the inner end of this cylinder, 
it is possible to so arrange the curvature of this surface that 
a real image is formed, the convex extremity of the cylinder 
acting as a plano-convex lens (see Pl. XL., Fig. 2, C, D). 
On the other hand, in the Blow-fly imago and in the Dragon- 
flies no such cone exists, and the subcorneal image is produced 
by a corneal lens. Ifa second image is produced at all, it must 
