The appearances above enumerated would indicate that we have to do with an inflammation of the optic nerve, which had been induced by chemotaxic substances brought thither along Stilling’s canal. Such an idea is not a new one. It was suggested by Straub in order to explain the optic neuritis he found in his two cases of experimental inoculations of the ciliary body, and it also enjoys provisionally the support of Fuchs’s[3] authority. The idea that part of the lymph of the eye passes backward along a passage corresponding to the central hyaloid canal is not generally accepted, and rests largely on inference from the observation of pathological specimens. It would be difficult, however, for anyone who has carefully studied this series to doubt that such a flow exists; it is, of course, not suggested that any large percentage of the lymph travels in this direction.

A confirmation of these views is obtained if we refer to the three eyeballs in which the organisation of the cone of exudate into the vitreous has attained the highest development. We notice in these how extremely far this process of organisation has been carried in the apex of the cone, where it is represented as a well-defined opaque cord (Pl. VI., Fig. [39]; and Pl. III., Fig. [17]). In one of the three a fibrous band, presumably the remains of the canal of Stilling, is clearly seen, whilst in the two others the existence of this structure is at least indicated. It is possible that in the first-named a congenital peculiarity exaggerated the distinctness of the appearance.

The Retina.

Detachment of the retina occurred in 38 of the 54 globes (70·38 per cent.), and was absent in 16 (29·62 per cent.). It was partial and ill-marked in 5 (9·25 per cent.), extensive in 9 (16·68 per cent.), and complete in 24 (44·44 per cent.). The ocular tension was above normal in 11 out of the 16, which presented no detachment, but in only 6 out of the remaining 38; it was below the normal in 13 of the 24 globes with complete detachment, and above it in 3 of them. In the very great majority of the eyes the detachment of the retina was undoubtedly due to traction from within. The sequence of events is plain from a study of the whole series.

At the operation there was an infection of the coats of the eye, and also of the vitreous chamber from without; this led to the formation of inflammatory material within the vitreous chamber; adhesions took place between these new-formed membranes and the retina; finally the shrinkage of the organising inflammatory material tore the retina from its bed. Every step of the process can be traced either in microscopic sections or in the naked-eye specimens. The earliest possible stage is seen under the microscope in sections of an eyeball (No. 37), where in the neighbourhood of the ora serrata the shrinkage of the exudate within the vitreous chamber has just begun to lift the retina from its bed. The individual points of attachment between the inflammatory membrane and the retina are beautifully illustrated. The ultimate stage of the process is to be found in those cases in which the retina is not merely totally detached, but has shrunk posteriorly into a stick-like form (Pl. III., Fig. [19], and Pl. IV., Fig. [25]), whilst it opens out anteriorly into a mass in which the iris, the ciliary body, the lens, the remains of the vitreous, and the retina are inextricably matted and tangled. When sections of such specimens are examined under the microscope, their leading feature is the evidence of severe plastic iridocyclitis, with the formation of abundant cicatricial tissue, which mats all the parts together and severely distorts the normal anatomical arrangement. The retina is dragged forward from the neighbourhood of the ora serrata over the ciliary body, whilst elsewhere it is thrown into abundant folds and completely separated from its normal attachments. A pseudo-cystic condition is thus produced, the apparent cysts being formed by the elaborate folding of the membrane (Pl. III., Fig. [19], and Pl. V., Fig. [31]). These may be small and slit-like, or may be large and round, so simulating the appearance of true cysts. In front of the retinal mass, lens fragments and capsule are seen in a number of the specimens entangled in the scar-tissue. As has already been said, all grades can be traced, from the slightest detachments up to those we have just been describing. The greater or less degree of separation of the retina met with in the different globes is doubtless in part a question of time, but it is also, and probably to a much larger extent, one of the character and grade of the inflammatory process excited in the eyeball. The more plastic the type of inflammation and the more intense the process is, the greater will be the measure of ultimate cicatrisation, always provided that the inflammation is not intense enough to result in suppuration.

There are several different ways in which the exudate which forms within the vitreous chamber may be placed in a favourable position for the production of retinal detachment.

1. The first of these is illustrated by each of those globes (Nos. 44 and 72) in which the site of a wound of the retina forms the point of connection between that membrane and the inflammatory exudate lying in the vitreous cavity (Pl. V., Fig. [30]). The traumatic infection of the retina served to attach the vitreous exudate to its walls, and thus paved the way for the separation of the membrane. In one of these cases (No. 44) a longitudinal fold was detached, whilst in the other (No. 72) the detachment was broad and shallow.

2. In the second method also, it is necessary to postulate an infection of the retina before that membrane could have contracted adhesions, either localised or widespread, to the neighbouring vitreous exudate. Once, however, the virus was planted within the hyaloid chamber, it probably diffused itself widely, and by means of chemotaxis set up an inflammation of the retina; evidences of such a retinitis abound in many of the specimens. Attachments between the vitreous exudate and the retina having been thus formed, the contraction of the former would naturally lead to the separation of the latter from its choroidal bed.

3. In a few of the globes the contracting membrane is merely an infiltration and thickening of the anterior layer of the hyaloid. It is well known that the vitreous body is, under normal conditions, more firmly attached to the retina in the neighbourhood of the ora serrata than it is elsewhere; it is therefore obvious that an inflammatory contracting membrane in the anterior part of the vitreous will pull throughout its whole circumference on the retina in its neighbourhood, effecting a detachment over a very wide area (Pl. III., Fig. [20]). This is just what we see happening in the globes we are now discussing.

4. In a number of the specimens it can be clearly seen that the bands, which drag upon the retina, radiate from the remains of lens masses, which are themselves encased in inflammatory tissue, and are bound thereby to the iris and ciliary body in their neighbourhood. Such bands appear in some cases to lie in the substance of the retina itself (Pl. IV., Fig. [24]); in others they are situate in the vitreous and present the form of membranous sheets, separated from the subjacent retina only by narrow spaces, and finding attachment to it in the neighbourhood of the equator (Nos. 117 and 170). The characteristic of these cases would appear to be that the dislocated lens is in them the principal focus of sepsis within the eye. The point is of interest, since some of them, at least, represent ruptured Morgagnian cataracts; for there is reason to believe, on clinical grounds, that the liberation of Morgagnian fluid within the eye is, sometimes at least, productive of considerable irritation to the surrounding parts.