394. ME. B. THOMPSON LOWNE ON THE COMPOUND VISION 



fenestrated membrane *, and it is usually believed to be perforated for the passage of 

 nerve-fibres to the great rods. In many sections it is easy to see that the membrane is 

 continuous, except where it is perforated by the tracheal vessels. In the great Dragon- 

 flies it is also undoubtedly pierced by the fringes from the pigment-cells of the dioptron, 

 which intercommunicate with the fringes of the pigment-cells, situated on the neural 

 surface of the membrane (fig. 69). In some insects the membrane is actually thickened at 

 the inner terminations of the great rods, so as to form small lenticular swellings (fig. 71); 

 and in a specimen of the eye of a lobster, in which the neural elements beneath the great 

 rods are not in the same line with them, the swellings are prismatic (fig. 70). 



In some insects, especially in Notonecta, the basilar membrane is strongly ridged, the 

 ridges corresponding with the attachment of the cuticular sheaths of the segments of the 

 dioptron. It is apparently defective in parts, the openings being closed by the cellular 

 layers which cover the two surfaces of the cuticular membrane. 



In the crane flies f (Tipula), in some Coleoptera (Telephorns), and in other insects 

 in which the component segments of the dioptron resemble distinct ocelli rather 

 than parts of a compound eye, the cuticular sheaths of the great rods extend inwards, 

 and also include the nervous structures. The basilar membrane then appears to be 

 actually perforated, so that the inner extremities of the great rods come into contact with 

 the nerve-terminals. This condition, however, is very exceptionnl, and is capable of 

 explanation from the manner in which the parts are developed. 



I shall recur to the consideration of the question of the passage of nerve-fibres through 

 the membrane when I discuss the function of the great rods and the bacillaof the neuron |. 



The various modifications of the cornea will be more conveniently considered here- 

 after ; I shall only mention in this place the fact that in all those cases in which there 

 are no lenticular facets, and where the cornea is simple and continuous, the outer face of 

 the crystalline cone is strongly curved, being a portion of a prolate spheroid, and there- 

 fore probably capable of producing an image of great sharpness. Such a surface has, as 

 is well known, the property of forming a perfect image, when the eccentricity of the 

 generating ellipse is the reciprocal of the refractive index, a condition which appears to 

 me to be very nearly attained in the curvature of these cones. 



Prolonged investigation has gradually convinced me that the very considerable diffe- 

 rences which the plan of the arthropod eye exhibits in different species and families 

 are due to differences in the consistency and chemical nature of the parts. The highly 

 refractive structures may consist of some modification of chitin, or some allied albumi- 

 noid, or of an oil-like fluid contained in the meshes of a fine stroma and enclosed in 

 elastic capsules. 



In the former case, where the refractive medium is chitinous, it undergoes but little 

 change of form in the preparation of microscopic sections : in the latter case the fluid 

 escapes from its capsule, and the whole appearance of the parts is modified, even when 

 such solvents as clove oil and absolute alcohol have not been used ; and the appearances 

 are still further altered by the use of such fluids in the preparation of sections. 



* Miillcr's Arch. I.e. t Phil. Trans. 1873, l.c. p. 570. t See page 409. 



