STUDIES IN THE RETINA. 345 



system ; the movements of the fluid cannot be so accurately 

 defined. There is evidence, however, to show that it is 

 associated with the nuckn. Indeed, certain phases passed 

 through by the nuclei seem explicable only on this assump- 

 tion. 



The first evidence for the existence of this stream is seen 

 in the early stages of rod-formation. The rods begin as 

 vesicles of fluid protruded from the retina. Some of the 

 processes have been described in Parts I, II, IV. A compact 

 mass of delicately walled vesicles push back the pigmented 

 epithelium, and each eventually becomes a cylindrical rod filled 

 with other substances besides fluid. In every case in which we 

 have analysed the successive stages in rod-formation, whether 

 in young eyes when the rods are short oblong vesicles, or in 

 older eyes when each new rod has to pass through a long 

 slender cone phase, we found that they seemed to pass in 

 turn through certain definite changes of shape. These 

 different form-phases could be best explained by a periodicity 

 in the discharge of fluid into them ; each protrusion not only 

 altered the shape of the particular rod, but, in so doing, 

 compelled a change of shape in those around it.^ 



Further, there is microscopic evidence for such discharges. 

 In the Amphibia, we not only see the large vacuolar basal 

 ends of some cones contrasting strongly with others with thin 

 and more solid-looking bases, but we have a corresponding 

 contrast in the appearances of their nuclei. These, at least 

 in the earlier stages of rod-formation, show two very 

 distinct phases — one large, clear, and vesicular, the other 

 small and full of chromatin (cf. the figures, Part II, fig. 23, 

 contrast the two elements on the right of the figure). 



When the rod has reached its normal size, the large 

 periodic discharges which seem necessary to protrude the 

 rod are apparently no longer required. The rod-nucleus 

 appears to remain fairly uniform in size, but nevertheless 

 keeps up a more regular smaller discharge of fluid into its 



* Except when discharged among fully formed rods which undergo no 

 further changes of shape. 



