At a certain period all nucleated red corpuscles disappear from mammalian blood. Non-nucleated corpuscles take their place. How are the latter formed? For a short stage of embryonic life nucleated cells containing blood-pigment are seen, or are supposed to be seen, in the liver—there is, unfortunately, great difficulty in distinguishing them with certainty from young liver-cells; later they are seen in the spleen; throughout the whole of life they are to be seen in the marrow of bone. The nucleated cells give origin to the non-nucleated corpuscles. It is hardly legitimate to call these cells persistent embryonic corpuscles. Yet the chain which connects the cells which in the embryo are capable of dividing into pairs of nucleated red blood-corpuscles, and the cells which, assuming the rôle of parent cells, do not accumulate hæmoglobin for their own purposes, but for the benefit of the red corpuscles which split off from them, is probably unbroken. In this sense they are persistent embryonic corpuscles which have deserted the blood-stream, and have taken shelter in certain tissues which are particularly favourable for cell division. The situations in which they hide themselves are singularly suggestive. In the liver there is an abundant supply of nutriment, more abundant than in any other part of the body of the embryo. Later, in the spleen, red blood-corpuscles are being destroyed. Materials available for making new ones must therefore be set free. The inside of a hollow bone is a peculiarly sheltered situation. The fat cells of marrow accumulate there after a time; but within some bones the marrow develops very little fat; hence it shows the red colour, which is due to its abundant bloodvessels. This “red marrow” is the most important seat of the manufacture of red blood-corpuscles in adult life. Unfortunately, when we try to answer the question, How are they formed? we are obliged to speak with caution. Some histologists assert that the nucleated cells divide, and that one of the two daughter cells accumulates hæmoglobin, and loses—that is to say, extrudes—its nucleus. Others maintain that the nucleated cells become irregular in form; that hæmoglobin accumulates in the projecting portion of the cell; that this projecting portion breaks off as a non-nucleated corpuscle. It would be indiscreet at the present time to pronounce in favour of either of these reports, although the decision is of theoretical importance. If the former account be true, red blood-corpuscles are nucleated blood-cells which have lost their nuclei. If the latter account be in accordance with fact, it is hardly justifiable to regard them as cells. They are parts of cells which finish their existence independently of the cell body and nucleus to which they belong. As circumstantial evidence, favouring the theory that cell division is normal and the nucleus subsequently lost, may be pleaded the existence in marrow, and also in the embryonic