ivith Special Reference to the Vision of Insects. 325 



any other. But the position of the images, whether erect 

 inverted, or any other, is obviously immaterial. It is the 

 ultimate effect within the occipital lobe of the brain that is 

 alone essential. 



Section II. — Of Vision with Compound Eyes. 



After these preliminary remarks on vision in general, we 

 seem to be in a position to deal intelligently with the in- 

 quiry — How is the retinal image formed in insects ? and what 

 kind of vision do they enjoy through the instrumentality of 

 the compound eyes with which they are furnished ? These 

 questions may be most conveniently dealt with by describing 

 a rough model of an insect's eye. Imagine a hemispherical 

 shell of some transparent material, e. g., half of a sixteen-inch 

 globe of glass, that is, a globe of which the diameter is six- 

 teen inches. Place your eye at its centre, and look through 

 it at the objects of nature around you. Next, let an accurate 

 picture of these objects be painted on the outside of the globe, 

 so that when you place your eye at the centre you still see 

 the same scene as before. Now let a network of scratches 

 be made all over the painting, dividing it into patches, each 

 of which is the size of a square quarter of an inch. This is 

 about the size of the cross section of a lead-pencil. There 

 will be about 6400 of these patches on the hemisphere. Next, 

 let the paint of each patch be removed, and a sing'e dab of 

 paint substituted, of a tint and brightness which is the re- 

 sultant of the part of the picture which fell within the patch. 

 In this way, a somewhat coarse mosaic is substituted for the 

 more perfect picture of the external world previously drawn. 

 This coarse mosaic gives a rough imperfect representation of 

 the external world, and represents correctly the vision which 

 an insect has of it. The compound eyes of some insects, 

 especially insects that attack other insects, have more nume- 

 rous facets than what correspond to 6400 over a hemisphere ; 

 and in such cases the mosaic is less coarse, and the vision is 

 proportionately better. Thus the eye of a dragonfly is better 

 represented by substituting smaller patches, each the size of 

 a square eighth of an inch. This increases the number over 

 the hemisphere to 25,600. But there is a somewhat narrow 

 limit to improvement in this direction, owing to its necessi- 

 tating a diminution of the aperture of the lenses. The way 

 that nature deals with this difficulty is by increasing inordi- 

 nately the size of the compound eye of the insect out of pro- 

 portion to its other features. In this way the number of the 

 patches, one of which is formed by each facet, can be 



Phil. Mag. S. 5. Vol. 37. No. 226. March 1894. Z 



