278 Memoir upon the compo7tnd 



If the mechanism of vision, such as it is in insects^ seemed 

 even to the most eminent anatomists of the present age so dif- 

 ficult to comprehend, it would seem that the way in which they 

 proceeded to the dissection of the compound eyes (lor few inter- 

 fered with the smooth eyes) has been in a great measure the 

 cause. The case would have been different, if, in their dissection* 

 of compound eyes, they had not always proceeded from the in- 

 side to the outside, a method of operating which may lead inta 

 errorj for, however little we disturb the optic nerve, the filaments 

 tv'hich it gives off, drawn a little back by being disturbed, are no 

 longer exhibited on the tunic of the cornea. If, on the contrary,- 

 vit. carefully remove the cornea, we see in an evident manner 

 the numerous nervous filaments which pass through the tunic 

 of this membrane, and spread a little in order to form the peca- 

 liar retina of each facet. From that instant, we need no longer 

 explain how the light is able to act on the retina through an 

 opaque varnish. 



If we set out from the organization of the eyes of insects, it 

 would appear that we might conceive that vision is produced 

 among them in the following manner : 



When the light meets a diaphanous body terminated by a, 

 curved surface, which gives it access into its interior, it undergoes 

 a refraction. If its rays being parallel meet the surface of this 

 body, and its medium be denser than that in which the incident 

 rays move, the broken rays will apjnoaoh the perpendicular by 

 converging upon each other. But if the rays of light fall ob- 

 liquely, malving a very wide angle, the more obliquely will they 

 fall, and the less will they approach the perpendicular. These 

 Are precisely the different effects which light undergoes in falling 

 on the cornea, which unites with transparency a convex sur- 

 face, and a denser medium than that in which tl>e luminous rays 

 move. 



In fact, the hmninous rays, direct or reflected, which issue fronrt 

 a visible body, and reach tlie eye, form ditferent cones, the points 

 ©f which are at the object, and the bases on the cornea. Those 

 wiiich fall on this membrane in an oblique direction, and by 

 forming a very wide angle, are reflected, and do not traverse. 

 Those, on the contrary, which fall under a convenient angle (an 

 angle which in the human eye is estimated at 48"^;, pass through 

 the cornea, or the facets which compose it, and undergo from '\t 

 ft refraction, which ought to bring them near the perpendicular. 

 It results that these nervous filaments, situated immediately un- 

 der, are struck bv the great quantity of rays which reach them, 

 «nd which, as a consequence of their direction, are concentred at 

 rhe most sensible part of the optic nerve, if we may so express 



