Maximilian Salzmann.

The anatomy and histology of the human eyeball in the normal state, its development and senescence ; online

. (page 4 of 27)
Online LibraryMaximilian SalzmannThe anatomy and histology of the human eyeball in the normal state, its development and senescence ; → online text (page 4 of 27)
Font size
QR-code for this ebook

the cornea the height of the cornea ; this amounts to about 2 . 6 mm.
It increases with the curvature of the cornea and also with the size of the
cornea, for a larger segment of a given sphere is higher than a smaller
one. Large corneae are, therefore, easily held to be more sharply
curved, because they are especially high.

Looked at from within, the cornea seems deepened, just as it appears
set forward when looked at from without. The margin of this con-
cavity shows a very weak rounding which, so to speak, is the negative


of the sulcus sclerae externus and lies along the inner border of the sulcus
sclerae internus. The configuration of the sinus angle (cf. chap, xiv),
so important for pathology, is, therefore, in part conditioned by these

Apart from the indistinctness of the marginal portions, the cornea
everywhere has a uniform transparency, at least in middle life. These
two zones concerning which we now speak, cannot, therefore, be macro-
scopically separated from one another in life and are only characterized
by their microscopic structure. Bowman's membrane, in particular,
does not extend over the entire cornea; it ceases about i mm from the
border (PL I, b) and the remaining marginal portion of the cornea, which,
too, is characterized by further anatomic peculiarities, is known as the
limbus corneae.

The limbus is therefore a zone of about i mm width limited centrally
by the margin of the Bowman's membrane and peripherally by the
corneoscleral border, and, properly speaking, includes only the anterior
layers of the cornea. For the purposes of description, solely, we think
of the limbus as bordered by a plane going through the depths of the
cornea from the border of Bowman's membrane to the border of
Descemet's membrane.

The remainder forms the cornea proper, which is, therefore, char-
acterized anatomically by two basal membranes. I begin with the descrip-
tion of the cornea proper, because it is then easier to describe the limbus
as the transition zone between the sclera and cornea.

a) The Cornea Proper

In this portion one distinguishes 5 layers from before backward (with-
out inward): (i) Epithelium; (2) Bowman's membrane; (3) Stroma-
corneae; (4) Descemet's membrane; (5) Endothelium.

(PL II, 6, Ep)

Over the greater part of the cornea this has a uniform thickness of
37 to 58 mu; a slight increase in thickness is first noted near the edge
of Bowman's membrane. It possesses two wholly smooth border sur-
faces, an anterior, formed by the anterior surface of the cornea, and a
posterior, which is in contact with Bowman's membrane; the cross-
section, therefore, shows two exactly concentric contours, exactly parallel
when small stretches are looked at by high power.

In the great regularity of the anterior surface (the anterior contour
in the microscopic preparation) the epithelium of the cornea probably


surpasses the epithelium of all other parts of the body, and this regu-
larity has a connection with the function of the anterior surface as a
part of the optic system. The great brilliancy of the cornea, the
regular reflex of its anterior surface, is only present when the epithelium
is completely intact.

Unfortunately, one rarely sees this normal structure of the epithelium
in the microscopic preparation, since it is easily subject to mechanical
lesions and post-mortem changes, such as drying, and also because the
hardening fluids act upon the epithelium before they do upon the
deeper portions and to a greater extent.

The corneal epithelium is a stratified pavement epithelium of 5 to 6
layers of cells.

The deepest layer (basal or foot-cells, b) lies directly over Bowman's
membrane, and consists of cylindrical cells of some 18 mu in height and
10 mu in breadth; each cell turns an absolutely flat surface (foot) toward
Bowman's membrane, and a rounded end (head) to the succeeding cell-
layer. The nucleus is slightly oval (5X7 mu) and lies with its long axis
at right angles to the surface of the cornea.

As shown by these figures, the basal cells of the corneal epithelium are
characterized by a considerable size and rich protoplasm. The proto-
plasm is for the most part somewhat lighter than that of the succeeding
cell-layers. Yet places are found in the basal cell-layer in which the
protoplasm stains darker; these cells are longer, as well, and have
concave sides; their nucleus is displaced into the head and is obliquely
oval. These are the cells seen passing up into the next layer (H.
Virchow, 234) (PL II, 6, the cell next to the last on the right).

According to v. Ebner (48), the basal cells are positive, monaxial,
and double refracting, with the optic axis perpendicular to the surface.

The second layer ("wing" cells) consists of polyhedral elements with
convex anterior surfaces and concave posterior ones; the edges between
these concave surfaces are more or less drawn out into the shape of wings.
The long axis of the nucleus is parallel to the surface of the cornea, and
the protoplasm is of a darker color than that of the basal cells. The cells
of the middle layers are not double refracting (v. Ebner).

While the cells of the second layer are of about equal size in all three
dimensions, a beginning flattening is apparent in the third layer, i.e., the
cells are larger parallel to the surface and the nuclei are more oval. The
transition to the very flat cells of the surface layer (0) (i.e., 5th to 6th) is
carried out in this way. These cells are very large in surface expanse (up
to 46 mu) but are very thin, in general, and only measure some 4 mu in
the region of the nucleus. This thickening projects backward toward the


deeper layers, while the anterior surface is wholly flat. The complete regu-
larity of the anterior surface of the epithelium comes about in this way.

The nuclei of the surface cells are likewise much flattened (2.5X12
mu), stain less densely than those of the deeper layers, but show normal
structure and no indications of dissolution, therefore no keratosis, as in
the epidermis. According to v. Ebner (23), the surface cells are negative,
monaxial, double refracting, with the optic axis perpendicular to the

The union of the cells with one another is formed by means of cell-
bridges, as in the epidermis. The cell-outlines acquire a similarity to
ladders, because these bridges shrink easily and become drawn out into
threads. The totality of the space between these bridges forms the
intercellular canal or lymph system; it can be injected with fluid and is
often found widened in pathologic conditions (e.g., in glaucoma).

The intercellular spaces are plainest between the basal cells; toward
the surface they gradually disappear. They do not come out plainly
after all hardening fluids; with Mueller's fluid one usually gets simple
linear cell contours; formalin-alcohol, and other similar reagents, bring out
the cell-bridges better; one sees them best, however, in pathologic cases.

Leucocytes (wandering cells) are quite often found in this system
of spaces. They are characterized by their very heavy staining, con-
stricted, or fragmented nuclei. Under normal conditions they are prob-
ably only to be found between the bases of the foot-cells just in front of
Bowman's membrane (PL II, 6; a little to the left of the center). Under
pathologic conditions they are much more numerous and extend farther
into the epithelial layers. Concerning the nerves of the epithelium see

P- 35-

2. BOWMAN'S MEMBRANE ( Lamina elastica anterior)
(PL II, 6, B]

This membrane has a uniform thickness of 10 to 16 mu throughout
almost its entire expanse, and, with the exception of its pores, is wholly
structureless, showing neither cell-nuclei nor fibrillation. Its anterior
surface is sharply set off from the epithelium and has a curvature con-
centric with the anterior epithelial surface. It is absolutely smooth; I
have not seen the fine serrations described by H. Virchow (234) in any
of my preparations.

When, therefore, a larger expanse of epithelium has been lost, as after a slight
burn, the recognition of such a loss of substance is often very difficult; the curvature
of the cornea appears unchanged but, when one gets the reflection of the image of the
mirror from the margin of the loss of substance, the step caused by the desquamation
of the epithelium is recognized.


On section, the posterior surface of Bowman's membrane does not
show so sharp a contour as the anterior, for it merges with the most
superficial lamellae of the corneal stroma. Furthermore, it is not possible
to detach Bowman's membrane from the stroma.

In its staining reaction it agrees fully with the stroma lamellae, and
only in its homogeneity does it differ from the fibrillar appearing lamellae.
Bowman's membrane is therefore to be looked upon as a specially
modified superficial stroma layer.

The sole details of structure which one occasionally but by no means
constantly recognizes in cross-section of Bowman's membrane are the
pores for the rami perforantes of the corneal nerves. These are not
characterized by a different staining but solely by their extremely deli-
cate contours. Corresponding to each pore there are two contours,
visible only by very sharp focusing; they are usually parallel to one
another, although not coursing exactly at right angles to the surface.

In most eyes the periphery of Bowman's membrane is sharpened
from behind forward.

3. THE Substantia propria OF THE CORNEA (Stroma corneae)

(Substantia propria corneae)

(PL II, 6, 9, C)

This forms the main mass of the cornea, comprising about 90 per cent
of the entire thickness.

No tissue of the eye has been more thoroughly and carefully studied
than the corneal stroma, and yet we are far from a full understanding of
its structure. It is not my task to go into all the disputed questions and
small details; those who wish to familiarize themselves with these things
should consult H. Virchow's (234) exhaustive and critical presentation.
The following description will be limited to the most important struc-
tural details, especially to those which one can see in sections after ordi-
nary stains, and these alone can serve as the basis for consideration of
pathologic conditions.

The circumstance that the spaces in the tissue in which the nuclei
lie do not unite with one another and course, throughout, parallel to the
surface is the most striking thing seen in a section perpendicular to
the surface; this is characteristic of the corneal stroma and differentiates
it from the sclera, as well as from pathologic scar-tissue. This picture
comes out in any section, independent of its direction, provided it is per-
pendicular to the surface. From this alone one can draw the conclusion
that the stroma corneae is made up of lamellae parallel to the surface.


That which lies between two spaces is not, however, a simple lamella;
this strip of tissue is not a unit of structure but a still larger number of
finer outlines parallel to the surface can be made out (PL II, 9), i.e., it
is made up of a large number of very narrow strips and the structure is
uniform only in these narrow strips only such a strip is to be looked
upon as the elementary stroma lamella.

One cannot, by any means, see the elementary lamellae in all places on section.
The corneal stroma is, unfortunately, very subject to swelling, and wave-like undula-
tions or nickings of the lamellae are frequently present. So it is clear that the out-
lines of the elementary lamellae- can be seen only where the section goes exactly at
right angles to the surface; if the cut goes the least bit obliquely to the border line, it
vanishes completely. The elementary lamellae are, as a rule, better and easier seen
in the posterior than in the anterior layers of the stroma.

When no cells lie between them, the surfaces of the elementary lamellae
are quite firmly and flatly united to one another; but when cells are
found, the neighboring lamellae are separated from one another and from
the cells in hardening, and in this way the spaces already spoken of arise.

Neighboring spaces do not, however, lie in the same level; for
if one follows out the whole surface expanse of a lamellar-complex
inclosed between two superimposed spaces, one sees forthwith that it
divides up and its component parts stick to other lamellar-complexes,
etc. In a surface preparation of the cornea each cell requires a separate

The individual elementary lamella consists of fine, straight connect-
ive tissue fibrillae, strictly parallel to one another. However, the direc-
tion of the fibrillae varies from lamella to lamella. The appearance of the
lamellae upon cross-section (PL II, 7) is, therefore, different: if the
fibrillae run perpendicular to the cut surface (q), the lamellae appear
portioned off in a peculiar way as if the fibrillae had been divided into
small bundles. On the other hand, when the fibrillae course parallel
or nearly parallel to the cut surface (/), the lamellae are finely striated or
homogeneous. (High-power magnification and a narrow aperture are
necessary to make out this striation.) One recognizes this change of
direction as well as the parallelism of the fibrillae in each individual
lamella still better in surface-sections.

According to Tartuferi (220), the corneal stroma is permeated by
numerous elastic fibers, coursing mainly in the direction of the corneal
lamellae ; they lie between the lamellae and form nets with beaded meshes
(perifascicular nets) ; at times there are membranous expansions at the
nodal points. These fibers arise from the corneal cells, according to
Lieto Vollaro (141) and Seefelder (202). An especially thick layer of


such elastic fibers lies just in front of Descemet's membrane and has
been called the lamina elastica corneae by Seefelder.

According to my observations, these elastic fibers course mainly in
the direction of those collagenous fibrillae in the particular lamellae which
they support, as in the sclera. The fine undulations described and
depicted by Tartuferi are apparently a result of his method of prepara-
tion. Seefelder's drawings and my preparations show them absolutely
straight (PL II, 8,/).

The elastic fibers of the cornea are objects difficult of demonstration
by the microscope; the ordinary staining methods do not show them at
all, as a rule. They can be demonstrated most easily by the molybdic-
acid-hematoxylin method of Held after fixation in Zenker's fluid, as See-
felder has shown. This method is, in general, a very simple one and, too,
is indispensable for the demonstration of the corneal cells.

The reason that I have deferred discussion of the dimensions of the
elementary lamellae is that we have had to rely very much upon conjec-
ture in this matter. The thickness is easiest to measure: in places such
as those shown in PL II, 7, where the individual lamellae have separated
from one another, I estimate the thickness of the cross-section of the
lamella to be 1.3 mu. In other places I have found the thickness to be
as great as 2.5 mu. Pes (170) gives the thickness of an elementary
lamella as i mu.

The breadth of a lamella, i.e., its expanse parallel to the surface
and at right angles to the direction of the fibrillae, cannot be directly
measured with exactness, because one cannot follow the individual
lamella for so great a distance with certainty. But it is certain that the
figure of 10 to 20 mu, which Pes gives as the breadth, falls far short of
the actual breadth.

The length of the lamella, i.e., its expanse in the direction of its
fibrillae, attains the width of the whole cornea, according to H. Virchow.
We must, therefore, look upon the lamellae as broad, thin bands, which
cross each other at wide angles, even of 90, and lace in and out among
each other at very narrow angles, so that they vary only slightly from a
course parallel to the surface.

In the anterior layers of the cornea the spaces are shorter and the
variation from a course parallel to the surface is greater than in the
posterior layers. One must, therefore, conceive that the lamellae in
front are narrower and thread in and out among each other more than
do those behind.

This oblique fiber course is especially plain in many places just behind
Bowman's membrane. Such trains were called the fibrae arcuatae by the


older authors. They appear to me to have a certain relation to the
nerve pores of Bowman's membrane, and possibly they are only trains
of fibers which accompany the corneal nerves, in case, indeed, they are not
the nerves themselves.

The greater thickness of the marginal parts of the cornea is due to the
greater number of lamellae, found only in the posterior layers. If one
follow these layers from the periphery toward the center, for example,
one sees that from place to place the most posterior lamellae (those
nearest Descemet's membrane) become thinner and finally disappear
(end in wedges).

The cells of the cornea are fixed and wandering cells. The fixed
corneal cells lie in the spaces of the stroma and for this reason are flat
elements placed parallel to the surface. The cell-body consists of a deli-
cate membrane, which only increases to a notable thickness in the
neighborhood of the nucleus, as shown on cross-section (PL II, 6, 9).
On surface-section the cell-body is not visible after ordinary staining and
only Held's stain shows it to be an extremely weak staining, finely
granular or reticular structure of irregular form with a few broad pro-
cesses (PL II, 8). One gets a similar picture in gold preparations, as
drawn by Druault (46) and Fuchs (69). The cells build a closed network,
or better a trabeculum or syncytium, by the help of these processes.

The nuclei of the corneal cells are markedly flattened, their thickness
reaching only 2 mu or less. Therefore, in cross-section or when on edge,
they seem to be very narrow, elongated, and dense staining; on surface
view, on the other hand, they stain very weakly and are of a great variety
of form, with rounded angles, lobulated, constricted like a kidney, or
elongated. In general, they are larger than the nuclei of the sclera;
rounded forms have a diameter of some 12 mu; elongated forms may
attain a length of 27 mu. The nucleus has a very delicate and fine-
meshed chromatin net and i to 3 very fine nucleoli lying in a district
poor in chromatin.

According to Ballowitz (14), each cell contains a microcentrum in the
neighborhood of the nucleus in the form of two central bodies united
by a bridge ; yet no radiating arrangement of the protoplasm can be made
out about this microcentrum.

According to the view of v. Recklinghausen (177), the cells and their
processes lie in an extensively subdivided system of canals and canaliculi
(lymph canal system) , which provides for the circulation of the fluid and
the nutrition of the cornea. Since then it has been conclusively shown
by Leber (138) that this view is untenable. The union between the cells
and the substantia propria is only more easily broken up than is the union



between the respective lamellae of the substantia propria, and for this
reason one can inject the canal system about the cells, and for the same
reason the wandering cells are principally found lying between the fixed
cells and the substantia propria.

The wandering cells are migratory leucocytes with clear or weakly
granular protoplasmic bodies and heavy staining, lobulated or frag-
mented nuclei (PL II, 8, w). Their form is determined by the amount
of room which the cell has; in the clefts they are pressed down flat;
within the lamellae they appear to be drawn out into long spindles con-
forming to the direction of the fibers.

Blood-vessels are entirely absent in the stroma of the cornea proper;
on the other hand, a richly developed nerve-plexus is present; this is
distributed over Bowman's membrane and to the epithelium, as well as
to the stroma.

The Nerves of the Cornea

Unfortunately ordinary stains do not bring out the corneal nerves at
all. To demonstrate them one uses gold chlorid, or better the Dogiel
methyl-blue method, which makes the finest branches and the nerve-
endings visible.

According to Hoyer (107), 60, and according to Dogiel (43), 60 to 80,
small branches enter the periphery of the cornea; the finer ones lie for-
ward, the coarser ones behind. They contain medullated as well as
non-medullated fibers; the former lose their medullary sheaths 0.3 to
o . 5 mm from the margin of the cornea (Hoyer) ; they give off non-
medullated fibers (which break up into finer fibrillae, Dogiel) at the nodes
of Ranvier, as well as at the ends of the medullated fibers.

In this way an extensive branching of the small trunks comes about,
and they and their branches press toward the center of the cornea and
toward the surface. The totality of this branching and anastomosing
forms the plexus proprius of the cornea (H. Virchow) ; the peripheral parts
of the cornea are supplied by the anterior branches, the middle by the
posterior. The plexus proprius is not found at all in the posterior layers of
the cornea, is coarser and looser in the middle layers, and becomes finer
and finer and more richly subdivided as it approaches the surface. It
ends near Bowman's membrane in a reticular formation (the terminal net
of H. Virchow).

From this the perforating bundles and fibers (the rami perforantes
of the older authors) are given off; these go obliquely toward Bowman's
membrane and come out through its pores. Thus they reach a position
beneath the epithelium, break up then into fibrillae and broaden out
between the epithelium and Bowman's membrane as the basal expansion


(H. Virchow) (subepithelial plexus, of the older authors). Part of the
fibrillae or branches from the basal expansion extend directly forward
in the intercellular spaces of the epithelium (intraepithelial expansion).

The nerves in the stroma end by little plates, according to Dogiel;
these are irregular, quadrilateral or spade-form, flat structures with
dentate or serrated edges. These endings are found only along the
border of the cornea. According to Dogiel, the nerves do not unite with
the corneal cells.

Hooked and looped endings as well as end-skeins (identical with W.
Krause's end-bulbs) are found immediately beneath the epithelium at the
limbus and also beneath the marginal portions of Bowman's membrane.
The nerves in the epithelium itself end by round or pear-shaped end-bulbs.

4. DESCEMET'S MEMBRANE (Lamina elastica posterior]

(PL II, 9, D)

This is a typical "glass membrane."

By "glass membranes," in general, one means highly refractile, structureless
(homogeneous) membranes of great firmness and elasticity; at times they have an
indistinct lamellation.

Their homogeneity is shown by the fact that no striation or reticulation of their
structure comes out even under the highest magnification of their surface; moreover,
it is shown by the form of broken and torn edges, for these show an irregularly rounded
or slightly angular line. Sometimes the torn edge has a terrace form, indicating a
certain degree of lamellation; at times this same formation comes out in different
layers as a variation of the staining.

The firmness of " glass membranes" is shown by their resistance to chemical agents
and by the lack of histolytic effect of pus in pathologic cases. With respect to their
elasticity it should be said that they, the glass membranes, are not elastically dis-
tensible in the sense that rubber is; they are only elastic in so far as they have the
tendency to take on a certain form of their own, somewhat as a watch-spring does.
This certain form is a curve just the opposite of that which the membrane has in situ,
i.e., the membrane shows a tendency to roll up toward the side opposite. Moreover,
the membrane in situ rests in a certain state of tension ; so wounds of a glass membrane
gape plainly, although not widely.

With respect to their formation, the glass membranes are to be looked upon as
cuticulae, i.e., as surface secretion products of a layer of cells along their base, i.e.,
along the side (of the cells) turned toward the connective tissue.

Descemet's membrane can be detached from the corneal stroma with
relative ease, and so shows two equally plain and sharp contours upon

Online LibraryMaximilian SalzmannThe anatomy and histology of the human eyeball in the normal state, its development and senescence ; → online text (page 4 of 27)