548 THE AMERICAN NATURALIST. [Vol. XXXVII. 



It was, therefore, necessary to use nerve fibres without pro- 

 tective coverings. The naked axis cylinders radiating from the 

 entering optic nerve in the fibre layer of the retina, met this 

 requirement. In order to get a clear demonstration of these, I 

 made tests upon the retina from a perfectly fresh ox eye where 

 the large size of the eye made manipulation comparatively sim- 

 ple. In this case there was little difficulty in identifying the 

 radiating bundles of nerve fibres which were readily distinguish- 

 able from small blood vessels and other structures of a fibrous 

 nature. The bundles of naked axis cylinders proved to be dis- 

 tinctly positive, thus agreeing with the rods and I am conse- 

 quently forced to conclude that in some way Valentin's obser- 

 vations were in this respect erroneous. 



Summary. The outer segments of the rods in the retina 

 of the frog contain each an axial core that differs from the 

 peripheral substance, but the exact nature of this core has not 

 yet been made out. The outer segments, as demonstrated by 

 the use of polarized light, are positively anisotropic and agree in 

 this respect with the axis cylinders of nerves. These outer 

 segments therefore, give evidence of containing longitudinal 

 fibrillae. Since they also show in the fresh state a transverse 

 banding, their structure is in some respects not unlike that of 

 a cross-striped muscle fibre in that in addition to a cross band- 

 ing they also possess a longitudinal fibrillation. 



