The Eye of the Cod-fish. 207 



the corpuscles was as remarkable as it was unexpected. The 

 cones now displayed a series of curious phenomena ; all of them 

 began to alter in form, the two halves partly separating from 

 one another, whilst each half at the same time gradually 

 assumed a more or less completely oval outline. Usually the 

 upper poles of the twin-corpuscles retained somewhat of their 

 normally truncated figure, and at this end they appeared 

 broader than at the other. Contemporaneously with these 

 changes, the borders of a clear transparent membrane connect- 

 ing the two halves of the cone became visible, and there also 

 appeared two minute globular vesicles, one at either lower pole 

 of each half of the cone. These seemed to be formed by the 

 outward extension of the external investing envelope, and they 

 invariably occupied the position indicated in the accompanying 

 drawing (Fig. 17). These saccular appendages, gradually in- 

 creasing in size, were evidently not the result of mere endos- 

 mosis, inasmuch as there appeared within them distinct evi- 

 dences of another structure, which to all present appeared to 

 be a filament spirally folded upon itself. This coil continued 

 to unroll and extend itself until at length the globular sac 

 assumed the condition of a cylindrical tube, the enclosed fila- 

 ment at the same time losing its essentially spiral aspect 

 (Fig. 18). Many of the cones had by this time separated 

 more or less completely into their characteristic halves, and the 

 delicate outer membranes surrounding the partially uncoiled 

 filaments subsequently disappeared (Fig. 19). No further 

 changes affecting this latter structure, were observed that even- 

 ing, but during my examinations of another fresh eye, made 

 next day with Mr. Murray's " Oberhauser" microscope, I 

 saw one example of the half-cone, in which the filament had 

 unrolled itself to the fullest extent of which it appeared 

 capable (Fig. 20). In addition to the above particulars, I have 

 further to remark that when acetic acid is added to these cones, 

 they immediately lose their normally plastic character, becom- 

 ing brittle, less regular in outline, and refract light more power- 

 fully. Caustic potass, on the other hand, gradually dissolves 

 them. Finally, it remains for. me to state that the twin-cones 

 of the cod, in their unaltered condition, present an average 

 measurement of l-500th of an inch in length by about l-800th 

 of an inch in breadth. On the addition of water they attain a 

 length of 1 -400th, but some normal cones which I have since 

 examined measured only the l-650th of an inch longitudinally. 

 It has occurred to me as possible that some might consider 

 the filament shown in Fig. 20 to be referable to the class of 

 structures known as the radial filaments of Miiller, which are 

 said to be normally connected to the upper end of the cones. 

 Such an interpretation, however, I do not think probable, 



