Nutritive material is carried to the tissues by the blood; but the material is accepted by the tissues only in accordance with their requirements for the moment, and is conveyed to the individual districts in suitable quantities. The muscular elements of the arteries have the most important influence upon the quantity of the blood distributed, and their elastic elements ensure an equable stream; but it is chiefly the simple homogeneous membrane of the capillaries that influences the permeation of the fluids. Not all the peculiarities, however, in the interchange of nutritive material are to be attributed to the capillary wall, for no doubt there are chemical affinities which enable certain parts specially to attract certain substances from the blood. We know, for example, that a number of substances are introduced into the body which have special affinities for the nerve tissues, and that certain materials are excreted by certain organs. We are therefore compelled to consider the individual elements as active agents of the attraction. If the living element be altered by disease, then it loses its power of specific attraction. I do not regard the blood as the cause of chronic dyscrasiae; for I do not regard the blood as a permanent tissue independently regenerating and propagating itself, but as a fluid in a state of constant dependence upon other parts. I consider that every dyscrasia _is dependent upon a permanent supply of noxious ingredients from certain sources_. As a continual ingestion of injurious food is capable of vitiating the blood, in like manner persistent disease in a definite organ is able to furnish the blood with a continual supply of morbid materials. The essential point, therefore, is to search for the _local sources_ of the different dyscrasiae which cause disorders of the blood, for every permanent change which takes place in the condition of the circulating juices must be derived from definite organs or tissues. The blood contains certain morphological elements. It contains a substance, _fibrine_, which appears as fibrillac when the blood clots, and red and colourless blood corpuscles. The red blood corpuscles contain no nuclei except at certain periods of the development of the embyro. They are lighter or darker red according to the oxygen they contain. When treated with concentrated fluids they shrivel; when treated with diluted fluids they swell. They are rather coin-shaped, and when a drop of blood is quiet they are usually found aggregated in rows, like rouleaux of money. The colourless corpuscles are much less numerous than the red corpuscles--only one to 300--but they are larger, and contain nuclei. When blood coagulates the white corpuscles sink more slowly and appear as a lighter coloured layer on the top of the clot.