ZOOLOGY AND IBOTANY, MICROSCOPY, ETC. 83 



by a netwrork of fine fibres. After noting the disposition of the fibres, 

 Patten describes the arrangement of the ommateal cells on the membrane, 

 and the bundles of nerve-fibres by which it is penetrated. The investi- 

 gation is extended to Galaihea, Pulsemon, Pagurus, BrancMpus, Orchestia, 

 and to Mantis religiosa. Instead of following these, however, it will be 

 more useful to summarize the author's general remarks upon Arthropods. 



General, — The compound Arthropod eye consists of a double layer of 

 cells, the ommateum and the corneal hypodermis. The latter always gives 

 rise to the corneal facets. Just as in MoUusea, the evaginate, convex 

 arrangement of the ommatidia has resulted in the expansion of the outer 

 ends of the latter, and a reduction of the retinulsB to protective purposes, 

 as in Arthropods. The retinophoras of each ommatidium form four 

 equivalent cells, " The terminal, cuticular secretions, or rods, are trans- 

 ferred to the axial faces of the outer ends of the retinophorse ; they there 

 unite to form the crystalline cones, to accommodate which the outer ends 

 of the retinophorse are enlarged into a cup-like expansion— the calyx, while 

 their inner ends are reduced to a slender tube, or style, serving at once 

 as a support for the calyx, and as a protective canal for the axial nerve. 

 Reasons are given for regarding the pedicel as a reflector. The convexity 

 of the eye is a solution of the problem of arranging the layer in the most 

 economic manner. The greater the number of ommatidia, the greater the 

 curvature of the surface and depth of the layer. The colourless cells are 

 the essential elements, the retinulae have not, as Grenacher supposed, 

 anything to do with the formation of the " rhabdom," which is formed by 

 the continuations of the crystalline ones. The crystalline cone-cells are 

 the essential elements, both morphologically and physiologically. 



After a discussion of relative observations, Patten comes to the follow- 

 ing general conclusions; — (1) " That the ancestral forms of all Arthropods 

 were probably provided with a small number of eyes placed on each side 

 of the head ; (2) these eyes consisted of closed optic vesicles formed by 

 invaginations lying close beneath the hypodermis, which formed a con- 

 tinuous layer over them ; (3) the deep wall of the vesicle formed a retineum, 

 similar to that of worms and certain molluscs, composed of colourless 

 double retinophoras, bearing terminal rods and containing an axial nerve- 

 fibre ; each retinophora was surrounded by circles of rodless pigment-cells ; 

 (4) the outer wall of the optic vesicle secreted a cuticular vitreous body, 

 similar to that found in the optic vesicle of worms (Alciopidae) and molluscs 

 (Fissurella, &c.) ; (5) the hypodermis overlying the optic vesicle (corneal 

 hypodermis) gave rise to a lenticular thickening of the cuticula, the lens." 

 Modification has been in two directions : (1) an increase in the number of 

 ocelli, with a decrease in the number of their ommatidia, or a decrease in 

 the number of ocelli with an increase in the number and complexity of 

 the ommatidia. 



The eyes of Enphausia, &c., are briefly described, and the development 

 of Arthropod eyes is briefly discussed. Miiller's theory of mosaic vision, as 

 advocated by Grenacher, is subjected to searching criticism and rejected. 

 Morphologically, the seat of vision ought to be in the crystalline cones, 

 the necessary nerves are only to be found in the crystalline cones, and 

 finally the most perfect optical conditions are obtained in the crystalline 

 cones, therefore the cones are the percipient elements. In eyes with 

 lenticulate facets, an inverted image of those objects lying within the axis 

 of the ommatidia will be formed upon the crystalline cone. In Musca or 

 in Mantis, for instance, " there is absolutely nothing to prevent the forma- 

 tion of a perfect image, not upon one or two nerve-fibres whose surface is 

 in no wise proportional to the size of the image, but upon a complete and 



G 2 



