lymph and _vice versa_ through the wall of a capillary vessel is in certain situations or at certain times a mere process of filtration; at others a process of restricted filtration. If the wall is behaving as a perfect membrane, it is a process of diffusion, or osmosis. It seems unnecessary to regard it, in any case, as a process of secretion. The more widely the capillaries are dilated, the less resistance do they offer to exudation. The narrower their calibre, the greater is the restraint which they place on the escape or entrance of fluid. When the skin of the palm of the hand is not sufficiently thick to protect the soft tissues beneath it from the injurious effects of the prolonged pressure of an oar or an axe, the capillary vessels of the under-skin dilate; more lymph transudes; the skin is raised up as a blister. The same thing happens when the capillaries are dilated and paralyzed by scalding water. The fluid of a blister has much the same constitution as blood-plasm, except that it contains less proteid substance. These results might be regarded as purely mechanical—the direct effects of pressure or heat upon the membranous capillary wall. But the “vital” element is more important. The capacity of endothelium to act as a barrier depends upon its nutritive condition—its vital integrity, as it might be termed; which no doubt in the last resort means its chemical relation to the fluids which bathe it. Now and again blebs, like blisters, are formed on the skin—the herpes which appears about the mouth; urticaria, which is more generally distributed; and various other cutaneous disorders. Frequently a connection can be traced between these eruptions and the consumption of a particular food. An attack of urticaria results not uncommonly from eating lobster, mussels, rook-pie, or some few other articles of diet. Various things—bad fish, for example—may produce the same effect; but shell-fish have an especially evil reputation. If extract of lobster or of mussels be injected into the blood of an animal, the amount of lymph which leaves the blood is markedly increased. The extract acts as a poison upon the endothelium of the capillary walls. It increases its permeability in all conditions in which lymph escapes in undue quantity from the blood-stream, or escapes more rapidly than it is absorbed; the nutritive condition of the endothelium is disturbed. Its unusual permeability is due in part, no doubt, to the dilatation of the capillary tube, the stretching of its membranous wall; but it is due also to the diminished vigour of the endothelial cells. They have lost to a certain extent their capacity for holding their edges in perfect apposition.