606 HOWARD. [Vov. XIX. 
tion with a deeper differentiation in structure” (Hensen, 
67; Schultze, °724,). This differentiation was observed as 
regular rifts in the inner substance of the rods. But well 
preserved rods, I find, do not show these rifts. In such cases 
with comparatively low power (1,000 diameters) the sheath 
usually appears continuous, though sometimes thinner in the 
bottom of each furrow. 
As I have shown (pp. 579, 590-591), the longitudinal ridges 
of the outer segment in osmic material correspond with the 
stained fibrils demonstrated by other methods of fixation and 
with the fibrils seen in optical sections of fresh rods. Hensen’s 
('67, Fig. 7) figure of the fresh rods of frogs corresponds with 
what I have seen in Necturus. It seems that the longitudinal 
cleavage lines occur between the fibrils and that the vacuoles 
of some material are due to penetration of fluid along these 
grooves, perhaps through a membrane. The appearance then 
of deeper radiating lines is probably due to artifacts determined 
by these peripheral structures. 
Some preparations suggest the presence of a set of fibrils 
or tubes parallel to the stained fibrils and alternating with 
them, but I should put little reliance on such a conclusion with- 
out more evidence. 
The fibrils of fresh and fixed material are usually slightly 
oblique, otherwise I have found no fibrils in the rods to com- 
pare with: the ‘spirals ‘of Hesse: (\!'04),) or of Ritter, (ome 
gt”). The evidence, such as I have obtained (Howard, : 03), 
is all against W. Krause’s (’92) opinion that the transverse 
cleavage is due to a system of closely wound spiral fibrils. 
The fibrils which I have demonstrated in fixed material, 
most satisfactorily in Necturus, but also in the frog and the 
goldfish, are probably the same which Schneider (:02) has 
observed in the frog, though on this point I am not quite cer- 
tain, for Schneider’s description is accompanied by two small 
text-figures only. It is unfortunate that he has not published 
a more complete account of these “neurofibrils,”’ as he terms: 
them. 
