AND THE MOEPHOLOGY OF THE EYE IX INSECTS. L01 



iricles are often a very long time before they dilate, but I have never failed to sec the 

 reflex after the animal has been kept in the dark for half an hour, although it disappears 

 almost instantly even in diffused daylight. 



In the flies (Muscidse) the cone is surrounded by four large flat cells, which extend 

 from the margin of the cornea to the iris. All trace of ciliary rods and threads is 

 wanting, although the irides are well developed. 



The inner layer of pigment-cells is in immediate contact with the basilar membrane; 

 the cells form rosettes around the inner extremities of the spindles, and send pigmented 

 thread-like processes over their surface. In the Dragon-flies these cells give oil' a number 

 of pigmented fringes which perforate the basilar membrane, and communicate with 

 similar fringes from the pigment-cells of the neuron. 



In the flies and in some Hymenoptera an intermediate set of stellate pigment-cells 

 exists {pg-) between the inner and outer pigment-cells of the dioptron. These cells inter- 

 communicate with each other, and in some preparations this gives the great rods the 

 appearance of branching and uniting with each other. I think that this appearance is 

 usually deceptive: the spindles are certainly always distinct, hut appearances in sections 

 of the eye of the Lobster, and in some osmic-acid preparations of the eye of a Moth 

 (Hemerophila perfumaria), certainly indicate that the sheaths of the spindles do inter- 

 communicate in some parts of the eye. 



I shall conclude my description of the dioptron with an account of some of the modi- 

 fications in its several parts which I have observed in different insects, and I shall add 

 what I am able concerning their development. 



I feel that in many particulars this account will he somewhat incomplete, especially 

 with reference to the development of the several structures ; but I am unwilling to 

 delay this communication for a fuller investigation, which will certainly occupy several 

 summers at least. 



A. Modifications of (he Cuntea mid Lens. 

 The compound cornea exhibits four distinct types in insects, three of which arc seen 

 in the various stages of development in the eye of the Common Cockroach. All the 

 are also found in the perfect condition in other Arthropods. In the earliest stage of 

 development in the larval Cockroach the cornea is simple and continuous, without a 

 ' trace of facetted structure. In this condition it consists of several transparent cuticular 

 layers (figs. 75 and 76. c). The entire cuticular cornea is shed with each ecdysis, and is 

 renewed from a layer of cells which, prior to the ecdysis, is situated between the old 

 cornea and the cone. I believe that the new cuticular cornea is formed heneath, and o 

 on, the surface of this cellular layer. In a larva of a Dragon-fly {Agrion), half an inch 

 long, I found a layer of large flat hexagonal cells covering the whole surface of the 

 cuticular cornea, one cell corresponding to each segment of the eye. These cells were 

 firmly united to each other by their edges, and each contained a large lenticular nucleus. 

 In a section of this cornea the cuticular layers are seen to he very numerous, and they 

 lie beneath the cellular layer. The nuclei of the cells stain very readily with eosin, 

 whilst the remainder of the cornea remains unstained. The outer cellular layer is usually 



