234 NUNNELEY, ON THE RETINA. 
pig, the ox, and the sheep, all just dead. (Pl. X, fig. 7.) They 
are, probably, in composition, similar to the granular cells, and 
like them break up and disappear. There is, certainly, no 
pigment matter in them; and whether they are the cells 
described by Bowman, Kélliker, and Hassall as the caudate 
ganglionic, I am uncertain, but believe them to be. In 
which case, either I am wrong in being unable to see the 
prolongations, or the cells described by them are modifica- 
tions produced by the reagent employed, which I am in- 
clined to suspect to be the case, seeing that I have often 
found the angles much prolonged when reagents, particularly 
chromic acid, have been employed. 
5. The fibrous layer is composed of the filaments of the 
optic nerve. This nerve, which consists of nerve-tubes and 
cerebral cells, enters the ball of the eye through the eribri- 
form plate of the sclerotic coat, with which the fibrous sheath 
of the nerve- fibres becomes confounded, and then through a 
narrow single aperture in the choroid coat, which is closely, 
but not organically, connected to the nerve. This entrance 
varies in its relative situation in different animals; in man, 
being to the inner side, and below the axis of the ball. The 
nerve here forms a slight mammillary projection, in the 
middle of which the central artery of the retina is seen. 
From this spot the nerve-fibres expand in every direction, 
forming a complete layer upon the outer surface of the 
hyaloid membrane. 
The fibres pass as far forwards as the ora serrata, but as 
yet the exact mode in which they terminate is unknown ; 
some observers have asserted that they form loops and return 
upon themselves, others that they are lost in the other ele- 
ments; but I believe no one has demonstrated their termi- 
nation. My own impression is, that these fibres are of 
different lengths, and successively terminate as they pass 
forwards, by being lost or continuated into the true retinal 
elements, the granules being the connecting medium between 
the nerve-fibres and the rods: and that the number of the 
fibres at the anterior part of the retina is much less than at 
the posterior, not merely because in forming a continuous 
expansion over a larger area necessarily there must be fewer 
fibres in any given space than there is in the smaller, but 
because the fibres are continually terminating, so that they 
are really fewer in number in the anterior than in the pos- 
terior part of the retina. * 
The fibres lie to the inner side of the granular layer, which 
* Instead of speaking of the fibres as terminating in any of the retinal 
elements, it would doubtless be more correct to speak of them as arising 
from or being continuous with them, the optic being an afferent nerve. 
