

AND THE MOEPHOLOG-Y OP THE EYE IN INSECTS. 397 



from the posterior end of the cone as approximately parallel, a conclusion which I 

 arrived at formerly *. 



In a specimen of the cornea and cones of the moth {Smerinthus populi) mounted in 

 balsam, the micrometer screw shows the distauce of the image behind the posterior 

 surface of the corneal lens to be about forty micromillimetres, estimating the refractive 



index of the lens at l - 68 and that of the balsam at 1/5 .\ -r-r- = ri2, and 40 : t-ttt '■ '■ f : 



1"5 1'12 J 



■pgg .•./=26 micromillimetres approximately. The radius of curvature of the anterior 



surface of the corneal lens is about forty micromillimetres, and the distance between its 

 anterior and posterior surfaces is 10 micromillimetres, which would give a focal distance 

 of 26 micromillimetres measured from the posterior surface as before, with the same 

 refractive index. 



The diameter of the image is approximately 20 micromillimetres, in the same balsam- 

 mounted specimen ; it corresponds to an angular aperture of 24 degrees. The distance 

 is 10 X cot 12°=47 in micromillimetres behind the optic centre ; which does not differ 

 widely from the above. The angle was determined by using two lighted tapers as the 

 object. The error arises from the difficulty of determining the exact size of the image, 

 that is the distance between the focal points. Of course the plane mirror of the micro- 

 scope was used in making these measurements. 



I formerly t gave the approximate length of the focus of the corneal facets of the 

 hornet as -»\q of an inch ; it is easy to see that it should have been -^oo, with a 

 refractive index which I estimated at 1'53, which is too low, as the refractive index is 

 nearly 2 % during life — a result which is sufficiently near those given above. 



If we assume the outer ends of the great rods to have a spheroid curvature convex 

 towards the cone, the posterior focus of which corresponds to the bacillar layer, all the 

 mechanism is present for the formation of an erect magnified image of the central 

 portion of the subcorneal inverted image upon the sentient structures. 



In many insects I have observed such a conformation of the outer ends of the great 

 rods as this theory requires, and I believe that when observation fails to show such an 

 arrangement, this is due to the very profound modifications which these structures 

 undergo when removed from the eye, or during the preparation of the eye for investi- 

 gation. 



J. Muller regarded the great rods as nerve-terminals, a conclusion which was justified 

 by his want of knowledge of their structure, but which is no longer tenable. Gottsche 

 first discovered the compound nature of the great rods, and described, as I have already 

 mentioned, the inner extremities of those of the lobster, as " double pyramids." The i 

 structure to which he gave this name is now better known as the spindle; Dr. Grenadier/ 

 speaks of it as a " Retmula." 



Max Schultze examined the structure of the great rods and their spindles in the 

 lobster ; he came to the conclusion that the rods are not the receptive elements, but 

 ascribed tbis function to the spindles. Max Schultze founded this opinion chiefly on 



* Phil. Trans. /. c. p. 581. t Phil. Trans. /. c. p. 581. 



SECOND SERIES. — ZOOLOGY, VOL. II. 59 



