The facility with which the constituents of blood pass out to the lymph, and the constituents of lymph pass into the blood, depends upon the condition of the walls of the capillary vessels. Water and substances dissolved in water might pass through the wall of a capillary vessel in either of three ways—by filtration, by osmosis, or by secretion. A filter is a porous barrier, which allows water and all substances dissolved in water to traverse it. The solution passes through unchanged in composition. Only solid particles are kept back. The rapidity with which fluid passes through a filter varies as the difference between the pressure on the one side and the pressure on the other. A membrane does not allow of filtration. Water and things dissolved in water pass through it by osmosis. Some things it will not allow to pass; such, for example, as gum, mucin, white of egg. To others it offers resistance in varying degrees. Most of the things that can diffuse through a membrane are capable of crystallization; but the membrane exercises some control over the passage of even crystallizable substances when in solution. If a membranous tube containing water in which proteins, sugar, and various salts are dissolved is hung in a basin of pure water, the proteins remain in the tube; the sugar and the salts pass through its wall into the surrounding water. But they pass at different rates. Those of small molecular weight pass more quickly than those whose molecule is heavy. After a time a condition of equilibrium is established. No more salts pass out of the tube. If now the contents of the tube and the contents of the basin are analysed, it will be found that the tube contains all the proteins, some of the sugar, and some of each of the salts, although not in the proportions in which they were present at the commencement of the experiment. The water in the basin contains some sugar and some of each of the salts, but not in the same proportions in which they are found in the tube. As a matter of fact, the same number of molecules would be present, per unit volume, on each side of the membrane—in the tube and in the basin. In this respect the percentage composition of the two solutions would be the same. But some of the molecules being heavy, others light, the weight of salts which unit volume of the solution in the tube would contain would not be the same as the weight of salts in unit volume of the solution in the basin. A membrane exerts a discriminating action on the substances which pass through it. Secretion is osmosis in disguise. It may be even filtration in disguise. A gland-cell (like an amœba) takes things up and passes them out without regard to their osmotic equivalent. It seems to exercise a choice. It seems to act in disregard of the laws both of filtration and of osmosis. So, at least, it appears to us when we are looking at the result in ignorance of what has happened inside the living cell. The passage from blood to