Eyes of Molluscs ami Arthropods. G55 



fication of the invaginate eyes? Although I would not entirely exclude 

 the former possibility, I have only foimd evidence in favor of the hitter. 

 As far as we kuow, only Area, Pecümculus, and possibly the aucestral 

 foruis of Pecten, were supplied with evaginated, or faceted eyes. while 

 all other Mollusca possess only the invaginate types. We may mention 

 the foUowing facts in favor of the supposition that the faceted eyes are 

 moditìcations of the invaginated oues. The latter are fouud in all stages 

 of development. large or small, deep or shallow, round, oval, or elong- 

 ated. If now the faceted eyes were formed from the summits and ridges 

 separating the pits and grooves , we should expect to find a scries of 

 intermediate forms, parallel with those of the invaginate ones. But this 

 is not the case, for the faceted eyes are usually of a large and Constant 

 form, sunkeu in pits more or less shallow according to the development 

 of the eye. It is very rarely that oue tinds small faceted eyes composed 

 of two or three ommatidia. Moreover there is a tendency on the poste- 

 rior portious of the mantle to produce intermediate forms between the 

 faceted and invaginated types. Finally the faceted eyes are only fouud 

 in Area, Pectimculus and probably the aucestral forms of Pecten, while 

 all other Molluscs are provided only with the invaginate types. These 

 facts render it probable that the latter are the simplest and ancestral 

 forms, from which the faceted oues have arisen by an evaginati on of the 

 pits, accompanied by the degeneration of the pigment cells into protec- 

 tive ones, while the retinidia of the retinophorae have developed in a 

 correspondiug degree (PI. 32, figs. 132 and 133). The faceted structure 

 of these couvex eyes is, as we have already said, a necessary result 

 of the radial arrangement of the retinidia, upon whose fibrillae it is 

 necessary for the rays of light to act at right angles. 



In the shallow grooves on the mantle edge and at the base of the 

 siphonal tentacle of Lamellibranchiata, as we have already remarked, 

 the cuticula forms a continuous layer, the inner surface of which, the 

 retinidial cuticula, is filled with a thin Stratum of tangential nerve fibres, 

 the retia terminalia. In the deeper grooves, or in the less developed 

 invaginate eyes, the cuticula is considerably thicker, and its inner layer 

 shows a tendency to form, over each cell, a cuticular cylinder, or rod, 

 containing the terminal fibrillae of the nerves surrounding each cell. 

 As the divisions between these cuticular cylinders be- 

 come more and more marked, so also the nerve fibres in 

 them assume amore Constant relation, giving rise to the 

 retinal rods, each of which contaius a specialized portion 

 of the retia terminalia, the retinidium. The corneal cuti- 



