AND THE MORPHOLO&Y OF THE EYE IN INSECTS. 



391 



Boll and Grenadier have, however, more recently adopted a modification of Midler's 

 hypothesis, based upon a view held by Max Schultze. Gottsche described the very 

 remarkable structures, at the inner extremity of the great rods of the lobster, under the 

 term " doppel-jPyramiden "*, better known as the spindles. Max Schultze regarded the 

 spindle as the true nerve terminal; it is the " Hetinula" of Dr. Grenadier t- 



On the other hand Wagner's theory has had numerous supporters, for the weak point in 

 Midler's hypothesis is met by it, as many insects with very few corneal facets have 

 evidently extremely acute vision. The absence of any retina in the position of the 

 image below the cornea is, however, fatal to it. 



Another difficulty militates strongly against this theory, which has been already dwelt 

 upon by Dr. Grenadier : the extreme minuteness of the subcorneal image would necessitate 

 recipient retinal elements far smaller than any known to exist in the animal kingdom. 

 Moreover, as each corneal image corresponds, as a rule, to an angle of about fifteen 

 degrees, and as the mean difference in the axes of adjacent facets is not usually half a 

 degree, the images of adjacent facets are nearly identical $, a fact in itself almost fatal to 

 Wagner's hypothesis. 



Formerly I accepted a modification of Miiller's hypothesis § ; but a further investiga- 

 tion has led me to discard both the theories of arthropod vision, and to substitute one 

 which is, I believe, in complete accord with all the facts, and which, so far as I know, has 

 not been even suggested by any previous observer. 



The compound eye is divided into two parts by a membrane which I shall speak of as 

 the membrana basilaris. The structures between this membrane and the cornea are the 

 crystalline cones and the great rods || ; I regard all these as dioptric structures, and shall 

 speak of the cornea, and all the parts which intervene between it and the membrana 

 basilaris, as the Dioptron. I shall use the term great rod provisionally to designate the 

 complex structure between each cone and the membrana basilaris, and segment of the 

 dioptron for the parts, collectively, beneath each corneal lens. 



Beneath the membrana basilaris I have discovered a layer of bacilla, comparable in 

 most respects with the rod and cone layer of the vertebrate, in the place where Gottsche 

 suggested such a structure might exist. This layer is succeeded by others, similar to 

 the nuclear and molecular layers of the vertebrate eye. I shall speak of all these struc- 

 tures as the Neuron. 



The existence of a continuous basilar membrane, in some arthropods at least, excludes 

 the great rods from any share in the nervous mechanism, a conclusion which also 

 follows from a more complete knowledge of their structure. 



Nearly two years ago I made an examination of the eye of a Plume Moth (Ptero- 

 flwrus pentadactylus), in normal saline solution, immediately after the death of the 

 insect. I was surprised to find a very considerable deviation from any structure previ- 

 ously described by others or observed by myself. Fig. 3 is a representation of the 

 optical section of a portion of the eye : in this condition the great rods presented the 



* Miill. Arch. 1852, p. 434. 

 § Phil. Trans, ibid. p. 577. 



f L.c. % Phil. Trans. 1878, pt. ii. pp. 596, 597. 



I| " Rhabdia" mihi, Phil. Trans. I.e. 



58* 





