VISION. 343 



ing disposed in seven rows, the number in each regularly 

 diminishing from the base to the apex; an arrangement 

 which is shown in Fig. 421.* 



The compound eyes of insects are formed of a vast num- 

 ber of separate cylinders or elongated cones,f closely packed 

 together on the surface of a central bulb, which may be con- 

 sidered as a part of the optic nerve; while their united bases 

 or outer extremities constitute the surface of a hemispherical 

 convexity, which often occupies a considerable space on 

 each side of the head. The usual shape of each of these 

 bases is that of a hexagon, a form which admits of their uni- 

 form arrangement with the greatest economy of space, like 

 the cells of a honey-comb; and the hexagonal divisions of 

 the surface are very plainly discernible on viewing the sur- 

 face of these eyes with a microscope, especially as there is 

 a thin layer of black pigment intervening between each, like 

 mortar between the layers of brick. The appearance they 

 present in the Melolontha, when highly magnified, is shown 

 in Fig. 422.J The internal structure of these eyes will be 

 best understood from the section of that of the Libellula 

 vulgata, or gray Dragon-fly, shown in Fig. 424, aided by 

 the highly magnified views of smaller portions given in the 

 succeeding figures, in all of which the same letters of refe- 

 rence are used to indicate the same objects.^ The whole 

 outer layer (c c) of the compound eye may be considered 



* Kirby and Spence's Introduction, 8cc., ill. 494. 



f The number of these cones or cylinders which compose the entire organ 

 differs much in different species. In the ant, there are only 50$ in a Scara- 

 baeus, 3180; in the Bombyx mon, 6236; in the house-fly (Musca domestica,) 

 8000; in the Melolontha vulgaris, 8820; in the Phakna cossus, 11,300; in the 

 Libellula, 12,544; in thePa/wVio, 17,325; and in the Mardella, 25,088. 



$ In the Phalense, and other tribes, they are arranged in squares (as shown 

 in Fig. 423,) instead of hexagons, and frequently much less regularly; as 

 must necessarily happen, in many parts, from the curvature of the spherical 

 surface. 



$ These figures, as well as the account of the anatomy of the eye of the 

 Libellula, are taken from the memoir of Duges, in the Annales des Science* 

 Naturelles, xx. 341. 



