No 3.] VISUAL CELDS\\INUVERTEBRADES. 605 
purple. This color was noticeably different from the blue in 
other parts of the rod (PI. 5, Figs. 41, 42) and in the field. 
In most cases this axis showed an irregular outline and dis- 
continuity, as if broken by post-mortem changes, due perhaps 
to the reagents. In some outer segments this stained axis 
projected, at the proximal broken end, beyond the peripheral 
substance of the cylinder (Fig. 42). 
Longitudinal superficial fibrils stain by the above method 
and also with toluidin blue alone. Whether the ammonium 
molybidate was an efficient factor or not, I did not ascertain. 
Hematem.—from a retina immersed for twenty-four 
hours in fluid albumen containing a few crystals of haematein, 
outer segments of rods were obtained which showed alter- 
nating light and dark transverse bands. The bands appeared 
to be the edges of disks or rings and the darker ones pro- 
jected slightly beyond the lighter ones, as do the successive 
segments in a muscle fibril when in a state of contraction. 
Lite DPSCWUSSLON: 
The various appearances presented by protoplasm or “cell 
plasm” after treatment with different fixing fluids is a 
matter of common knowledge. An interesting study of 
such effects on nerve cells is given by Floyd (:03). I have 
described some of the different appearances in the visual 
cells as prepared by various reagents, and I have shown that 
from this standpoint there is abundant opportunity for con- 
tradictory results where workers have depended on single 
preservatives. 
Any ROD MCELLS: 
The outer segments of the rods when prepared in osmic acid, 
have been shown to possess superficial longitudinal lines when 
viewed in toto. The interpretation of these lines from cross sec- 
tions has been (Schultze, Greef, and others) in general that they 
“depend upon a grooving of the surface which has some rela- 
