TERMINATION OF NERVES OF MAMMALIAN CORNEA. 469 
the most superficial scales) cau be traced above the most superficial nerve - 
fibrils. Oil account of the great thinness of the superficial scales even a 
moderately high power, such as Zeiss’s D and E (-|- and ^ inch), or Hart- 
nack’s 7 and 8, does not suffice. Amongst the great many specimens which 
I have examined — they amount to hundreds — only in very rare instances 
have I seen one or the other fibril at the same level as the most super- 
ficial nuclei. 
Before coming to their ends I must say a few words as 
regards their relation to one another. Cohnheim (1. c.j pp. 27 
and 28) does not speak of any anastomosis of the intraepi- 
thelial nerve-fibres, while Kolliker (1. c., p. 4) observed such 
anastomoses, although, as it appears, not frequently.” 
Anastomoses of the intraepithelial nerve-fibrils, although, on 
the whole, rarer than I thought them to be in my first 
memoir, do no doubt occur. My mistake of assuming them 
to be of very frequent occurrence arose from my not having 
examined with sufficiently high powers. 
It must be borne in mind that these fibrils in many places, 
when crossing, are in very close contact, and if they happen 
to cross at the point of a varicosity the appearance of an anas- 
tomosis is produced. When examining such places with an 
oil immersion -pV of Zeiss I can ascertain that in four out of 
six there is really no anastomosis between the fibrils ; in the 
other two it is impossible to say that there is not an anasto- 
mosis. But there exist real anastomoses, about which there 
can be no doubt whatever ; these refer to such places where 
two of the long fibrils are joined with one another by a 
shorter or longer side branch. In fig. 10 are represented 
such undoubted anastomoses. 
Now, as to the larger or smaller knob-like structures which 
are met with on the apparent extremities of the shorter as 
xvell as the longer fibrils, as mentioned above, and as observed 
by all who have worked at this subject, there is no doubt 
about their presence, and the question is whether they are 
the real ends or not. Kolliker (1. c., p. 4) would not say 
that these knobs are natural or artificial, Hoyer (1. c., 
234) considers them as artificial products, and Izquierdo 
(1. c., p. 28) asserts that some fibrils, at any rate, possess 
them. Cohnheim, Tolotschinow, Krause, and others, con- 
sider them as natural end-knobs. 
It cannot be denied that they possess the same appear- 
ance, whether at the end of a fibril or in its course ; but the 
great differences in their number, size, distribution, and 
shape, seems to point to their being artificial products. That 
some of them are not “end-knobs” this I am going to prove 
presently ; that others are such is doubtful, since it cannot 
with any show of reason he asserted, as Izquierdo does, and 
