existence of these fine spindle elements in connective tissue, I termed them, in my description of Fig. 6, "longitudinal fibres of elastic network," and in the text I spoke of them as "fine fibrillæ which present oval swellings at short intervals." I also stated that they communicate with a transverse network, which encloses the primary bundles in its meshes. Subsequent investigations have shown me that I was mistaken in regard to this supposed communication between the two sets of fibres. The longitudinal elements seen in Fig. 6 are chains of minute spindle cells, which lie between the primary bundles of muscle, and are distinct both from the muscle fibrilla and from the transverse network.1

When the temperature produced by the solution of caustic potash in an equal weight of distilled water falls to 107°-105° Fahr., a cornea, or part of a cornea, placed in the solution at that temperature becomes completely dissolved, with the exception of masses of cells, which can be seen isolated or in connexion with each other, when a part of the altered corneal mass is examined in a drop of the solution.

For a more complete account of these cells, the reader is referred to the paper in which I described them, in the Proceedings of the Royal Society, No. 155, 1874. In the Plate which accompanies that paper, figures are given, showing the relative position of the cells to each other. Fig. 7 shows the size and form of some of these cells.

FIG. 7.-Flat cells from ox cornea by potash solution.-Hartnack, Obj. 8, Oc. 3.

In sealed aqueous humour preparations in which the cornea has been carefully cut obliquely, the same cells are seen isolated

singly and in small masses, as is shown in Fig. 8. They are then seen as hyaline bodies, with a slightly projecting, faintlytinged nucleus. In a preparation in which they are visible, they can usually be traced from the obliquely cut surfaces a certain distance into the structure. Sometimes, but much more rarely, a preparation is obtained, in which their position relatively to the bundles is distinct. Such a preparation is represented in Fig. 9. It is then seen that the long narrow cells are applied to the surfaces of 1 I take this opportunity of correcting a misapprehension which I find some readers of my previous paper have fallen into, through my neglecting to state that the acetic acid which I used was not glacial acetic acid, but the weaker preparation known in England in commerce as concentrated."

FIG. 8.-Isolated flat cells from obliquely cut sheep cornea two days

sealed in aq. hum.-Hartnack, Obj. 8, Oc. 3.

66

the primary bundles, and that their edges are in contact after the manner of an epithelium, although, as will be seen from the figure, some of the cells become

FIG. 9. From frog's cornea sealed in aq. hum. Long narrow cells seen lying on the primary bundles. a, Cell. b, Part

primary bundle from which the cell

disintegrated, and disappear during the manipulation. In gold preparations, which have been prepared by placing the whole bulb for one to one and a half hour in the solution, removing the surface epithelium and making incisions in the cornea, and finally excising it of and placing it in acidulated water, the polygonal and rounded cells can be seen to form layers in the substance of the cornea. Successful preparations of these cells by gold in the uninflamed cornea are very rarely obtained. In a very large number of corneæ treated in different ways by gold, I have only seen them three or four times. If the cornea has been irritated for a few hours, they are more frequently seen in gold preparations.

has disappeared.-Hartnack, Obj. 8, Oc. 3.

It is to be noted that, although layers of the polygonal or rounded cells of the ordinary epithelial type can be seen of various sizes in the cornea, both in potash and aqueous humour preparations, patches can be isolated which are continuous with the long narrow cells that cover the primary bundles.

It was shown some years ago by Schweigger-Seidel, that there. are cells in the cornea which show the dark lines of epithelium when treated by nitrate of silver. He was for an unusually long time in the minority of one, in which Carlyle says all truth begins. But the demonstration of the fact is not very difficult. Of all the animals on which I have operated, with a view to the demonstration of this epithelium, the cornea of the mouse is the best adapted, and gives most frequent success. The cornea is placed, unwashed, in solution of 1 part nitrate of silver in 200 or 400 water for one to three minutes, dipped for a second in salt solution of the same strength, and then exposed to sunlight in glycerine. The preparations which succeed best are those in which the ground substance is only slightly stained, being seen as faint brown islands. scattered over a white field, instead of the darker and more abundant stained ground substance of ordinary silver-stained corneæ. The dark epithelial lines can be seen traversing the white groundwork, generally appearing to be interrupted when they come in contact with a stained patch of ground substance, and this usually prevents the completion of the visible boundaries of the cell outline. The interruption is, however, only apparent, the black line being lost in the dark mass which surrounds it, as may be seen when a patch is only slightly stained. The line can then be frequently traced over and beyond a dark island, completing the cell-contour.

1875.]

DR THIN ON INFLAMMATION.

Fig. 10 shows part of a successfully stained epithelial surface in the interior of a mouse's cornea, in which this continuation of the

FIG. 10.-Part of a layer of flat (epithelial) cells in the substance of a mouse's cornea.
Silver preparation.-Hartnack, Obj. 7, Oc. 3.

dark lines over the stained islands is in some parts visible. In the preparation there is another similar layer below the one represented, cornea substance intervening.

It is important to note that the cells indicated by the dark lines in Fig. 10 belong to a class of larger size than those of the layers of polygonal cells seen in the mouse's cornea in gold preparations. The ordinary appearance presented by a cornea which has been treated by nitrate of silver, solid or in solution, is represented in Fig. 11.

In such a preparation the silver has completely permeated the substance of the fibrillary bundles. Wherever there is an uncoloured space, we know that the ground substance is wanting to an extent sufficient to be entirely out of focus. In other circumstances these spaces may be found filled with a black albuminate of silver. Sometimes a large unmistakable nucleus of the stellate cell can be seen in each space, establishing the relation between the cell and the space. But the space is larger than the cell, does not correspond to it accurately in outline, and the colourless space-processes are very much broader than the processes of the stellate cell. It is evident that the silver permeating the fibrillary substance with ease, either does not readily find its way into the spaces, or if it does, that the spaces are bounded by a structure that does not stain by nitrate of silver. We have seen that it is possible to show the dark lines of epithelial cells in the walls of the spaces; and it is thus clear that either the substance of the cells does not stain, or that in such preparations the silver does not penetrate them. The fact of the dark lines being present

VOL. XXI.-NO. V.

3 E

would seem to show that there is a comparative difficulty of staining as regards the cells.' But that the epithelial layers in the

cornea form an obstacle to the passage of silver solution is shown by preparations obtained by placing the bulb entire in the solution for a few minutes, and subsequently removing the upper from the under layers of cornea substance. It is then often found that the silver solution has diffused itself equably through the whole of the fibrillary substance of one layer, but has not penetrated at all to the layer below it; from which I infer that a layer of flat cells in the cornea, and the delicate sub-cellular membrane on which such cells always lie, present an obstacle to the easy passage of fluid.

The demonstration of layers of flat cells in the cornea, and of an epithelium covering the spaces, shows that Recklinghausen was mistaken in supposing that the colourless stellate spaces and connecting processes are lymph canals (saft-kanälchen), channelled out of a substance which cements the fibrillary bundles. Each stellate space is formed by the superposition of lamellæ, one over the other, the surfaces being at points corresponding to the situaAnalogous cells can be stained in the subcutaneous tissue by an interstitial injection of silver solution, as can be shown by thus injecting the subcutaneous tissue of a mouse's back immediately after it has been killed.

1

tion of the stellate cells, removed from each other to an extent sufficient to enable the distance to be appreciated under the microscope. In a silver cornea where the staining is pronounced, the distinction of the surfaces where they are in contact is lost to the eye. Thus, a "space" simply shows that two membranes, which are at other points in close contact, have at one point left a gap between them, and it is in these gaps that the stellate cells are to be found.

I have previously shown that the canals in which the nerves lie are lined by an epithelium, and are in reality large lymphatic channels, and, as will be seen by referring to Fig. 11, there are points in the lymphatic where the colourless spaces adjoining communicate with the uncoloured outline of the vessel. At these points it is certain that the fibrillary or ground substance fails, but it does not follow that the uncoloured breach in the outline of the wall of the vessel represents the extent of the actual free communication. That there is, however, a free communication at these points, between the lymph channel and the spaces, can be satisfactorily made out by watching the movements of bacteria in a cornea which has been placed in conditions favourable to their development. It can be seen that they move in the large lymph channel, but that at certain points one oscillating rod after another leaves the vessel and enters an adjoining space, and that always at the same point. When, as we shall afterwards see, white bloodcorpuscles enter the inflamed cornea, they are sometimes found in considerable numbers in these lymph channels between the nerve and the wall of the channel, and it can then be seen that they make their way first into those spaces which are next the vessel. Whatever the nature of the communication may be-and it may not be more than a slight aperture at the angle of junction of the epithelium of the vessel-it is sufficient to permit the passage not only of fluid, but of minute solid particles between the lymph channel and the spaces.

The structure of the cornea, as shown by the above histological methods, can be described shortly as follows:-The fibrillary or ground substance consists of parallel bands, whose breadth is nearly the diameter of a human red blood-corpuscle, and whose thickness is somewhat less than their breadth. These bands, which I have described for the first time, and which I call primary bundles, are covered by narrow, elongated flattened cells, after the manner of an epithelium. Injected masses can be made to pass between the primary bundles, the spaces into which they pass being the corneal tubes of Bowman. These tubes are not, therefore, canals, with a complete unbroken wall, but are the spaces left between the bundles, on account of the somewhat cylindrical or bevelled form of their borders. The walls of a corneal tube are formed by parts of the surface of several contiguous bundles.