AND THE MORPHOLOGY OF THE EYE IN INSECTS. 405 



pseudoconic type is, according to him, confined to the Diptera with short antennse. In 

 this, again, I cannot agree with him, as the Dragon-flies exhibit precisely the same con- 

 ditions, and in the imago of the Cockroach I find a similar semi-fluid cone. In all these 

 osmic acid coagulates the fluid, and even the eyes of flies exhibit a well-marked cone 

 when the preparation has been so preserved (fig. 28). 



The elongated cones which have been observed in the eyes of many Arthropods, 

 extending from the cornea to the membrana basilaris (as, for example, in Hyperia *, 

 Typhis t , and some other genera]:), appeared to me for a long time to present great 

 difficulties with regard to the views which I now hold. Further investigation has con- 

 vinced me that these cones are artificially produced by the coagulation and drying of 

 the albuminous tissues of the cone and spindle, with their cellular sheaths. The evidence 

 of this is seen in numerous specimens of the eye of the Cockroach. In some of these 

 there are distinct elongated cones, whilst in others more successfully prepared the spindle 

 is well seen, and has the same form and structure as it exhibits in Notonecta and Tipula. 



The crystalline cones of the Nocturnal Lepidoptera and of the higher Crustacea are 

 probably morphologically distinct from the structures already described. The highly 

 refractive cone in Nocturnal Lepidoptera is surrounded by a softer, or even fluid, 

 sheathing cone, which extends from the apex of the crystalline cone to the spindle (figs. 3, 

 26, & 27, sh). I regard it (the sheathing cone) as the representative of the cone in the 

 Diptera and Dragonflies. 



Both the hard cone and its soft investing substance are divided into four longitudinal 

 segments, indicating their origin from the four cells of the primitive cone ; but I cannot 

 at present decide whether the hard cone is formed from the inner portions of these cells 

 or between them, as Claparede § thought not improbable. 



In both Noctuid and Crepuscularian Moths, the optical transverse section of the cone 

 often appears to contain four groups of deep purple granules ; these, in optical longi- 

 tudinal sections, are seen to lie on the surface of the cone, and are probably deposited by 

 the decomposition of the fluid which surrounds the cone. In dried as well as in balsam- 

 mounted specimens the sheath is so closely applied to the cone that it appears to form 

 its outer portion, and the coloured granules therefore look as if they were imbedded 

 in the substance of the cone itself || . Unicellular organisms are not infrequently found 

 in the fluid contents of the sheath of the cone : these are undoubtedly parasitic 



(ng. 27, z). 



It is by no means easy to understand the contents of the chamber in the Diurnal 

 Lepidoptera in such genera as Colias and Vanessa, although in Pieris brassicce no diffi- 

 culty exists, the cone being similar to that of a Noctuid Moth. I have already figured 

 the remarkable modifications seen in the two former genera^]". I suspect that the 

 cells (c. c.) in my figure are the representatives of the cone ; and that the tetraphore is 

 the outer extremity of the spindle, which retains its ovoid form after the escape of the 

 fluid contents of the tube, and that the tube forms a narrow stalk supporting the chitinized 



* Clapaxede, I. c. t id. $ Leydig. § L, c. 



|| Perhaps the coloured beads which cover the cone in Notonecta have a similar origin. 

 If Phil. Trans. 1. c. fig. 35. 

 SECOND SERIES. — ZOOLOGY, VOL. II. 60 



