lent process in the deeper parts of the cornea than from the purulent infiltration of the edges of the ulcer. The floor of the ulcer is often also purulently infiltrated, and this is especially true of small ulcers; or it may be quite clean and smooth; or finally, by the help of focal illumination and a magnifying lens, it may be seen to be gray and clouded, and to contain numerous gray-white, punctated, purulent infiltrations. This purulent infiltration of the margin of the ulcer is generally only at one part of its circumference, and shows itself in the form of a dirty, yellowish-white line, while the remaining part of the circumference is clean and smooth, and therefore more difficult to see. The ulcer spreads toward the side of the infiltrated margin, and after a time comes to a stand-still at this point, but the process is renewed at some other portion of the circumference, and the tissue in its neighborhood is destroyed.
Since this destructive process gradually creeps over the surface of the cornea, Sämisch[1] has designated this form of disease as ulcus corneæ serpens. These ulcers frequently appear flatter than they really are, since the thinned floor yields to the pressure of the aqueous humor and is pressed forward. Between the corneal abscess and the ulcus serpens there are many intermediate forms.
In the great majority of cases of purulent keratitis pus collects in the anterior chamber. In such cases iritis generally exists at the same time; but cases do undoubtedly occur in which hypopion exists, while the iris still reacts promptly under atropine and there are no adhesions with the capsule, nor any other traces of inflammation.
The belief formerly generally accepted, that the hypopion was due to a participation in the purulent process by the epithelium of the membrane of Descemet, seems no longer plausible, since we have come to regard the pus-cells as identical with the white blood-corpuscles. The process of a direct rupture of the abscess into the anterior chamber, as described by A. Weber,[2] I have never been able satisfactorily to observe. But one very often sees a yellowish-white thread of pus, of greater or less thickness, lying close to the membrane of Descemet, and extending from the ulcer
