Eyes of Molluscs and Arthropods. 659 



of a primitive optic cup, becomes somevvhat flattened, tlien those om- 

 matidia ou the edgc of the vesicle would be turued inward, retaining 

 ineautime their rods PI. 32. tig. 150); those ominatidia oii the anterior 

 wall, which in such cases are usually short and fimctionless, would, 

 with the foldiug in of the cup, lose their pigmeut, so that light could 

 still penetrate to the posterior wall of the vesicle. Bya contiuued 

 ingro wth of the ommatidia ou the periphery of the outer wall towards 

 the centre of the eye , the inverted ommatidia of the anterior wall would 

 become expressiv adapted to the function of vision, while those on the 

 floor of the optic vesicle would gradually become functionless, or modi- 

 fied in other directions. In the primitive optic cup, the retinophorae 

 were probably surrounded. as often happens, by a double circle of pig- 

 ment cells. an outer and an inner one. After the closing of the cup, this 

 arrangement appears to have been partly retained, the outer row of 

 ganglionic cells being the product of the modified basal row of retinulae, 

 while the inner ganglionic cells are the modified outer ones (PI. 32, 

 fig. 151). A similar condition must have obtained in the floor of 

 the optic cup, which uow forms the inner wall of the optic vesicle, 

 whose layers, if they are homologous with those of the retina, 

 should be in the inverse order. In this case the structureless and 

 non-cellular rete-vitrosum would be homologous with the layer 

 of rods; a continuation of this comparison would make the argentea, 

 composed of flattened and colorless granules, homologous with the re- 

 tinophorae with which it agrees in the secretion of the hyaline 

 layer, aud in the presence of the colorless and refractive globules, so 

 characteristic of the retinophorae in general. Lastly, the tapetum, 

 composed of pigmented cells. would be homologous with the entire 

 ganglionic layer: both therefore represent modifications of pigment 

 cells, or retinulae. After the closure of the optic cup, the modifications 

 of the surrouudiug connective tissue necessary to form the lens aud 

 pseudo-cornea, present no morphological difficulty. The basal mem- 

 brane of the primitive optic cup would, with the formation of the optic 

 vesicle, form the ommateal sac, from the thickening of whose anterior 

 and posterior walls would be formed the septum and sclerotica,, 

 both of which are double layered. ' 



There is still the greatest difficulty to solve , namely : why is it that 

 Pecten and Area are so richly supplied with liighly developed eyes, 

 wheu there seems to be no more necessity for them here than in any 



44* 



