INTRODUCTIONIt has been known for many years that a precipitate occasionally appears when two homologous antisera are mixed. If two rabbits be immunized with horse serum, by a similar series of injections, the antisera so produced will precipitate when mixed with suitable dilutions of horse serum, and a mixture of the two antisera will also occasionally develop a precipitate. Several possible suggestions have been put forward to account for this 'mutual' precipitation of antisera. Some workers think that one or both of the rabbit anti-horse sera contain remnants of horse serum, and that for some reason homologous antigen and antibody may co-exist side by side in the same antiserum without undergoing precipitation, whereas others think that horse serum is a mixture of antigens and that an antihorse serum may contain some of the antigenic components of the injected horse serum together with antibodies for other antigens of horse serum, but not homologous antigen and antibody. Consequently such an anti-horse serum will give a precipitate on mixing with horse serum, and a precipitate will also result when this antiserum containing antigenic components of horse serum is mixed with another anti-horse serum containing antibodies for these particular antigens.The experiments which help to elucidate this problem may be grouped into two categories. In the first category there is a series of experiments designed to investigate the equilibrium of precipitation reactions. In these experiments precipitinogen and precipitin have been mixed in various proportions and, after the formation of a precipitate, the supernatant fluid has been examined for the presence of antigen by adding antiserum and for the presence of antibody by adding antigen. The second category of experiments is concerned with searches for antigen in the blood of animals which have received injections of an antigen. In this case an animal, which has previously been immunized with an antigen, receives an inj ection of thesame antigen and, at varying intervals following the injection, samples of blood are withdrawn and the sera tested for precipitins by mixing with antigen, and for antigen by mixing with the serum of the same animal obtained before the test injection or with the serum of another animal similarly immunized.Linossier & Lemoine (1902) investigated the equilibrium of precipitation reactions using various animal sera (horse, human and bovine) as antigens, and concluded that when antigen and antiserum were mixed in various proportions there were three zones. At one end of the scale the supernatant fluid contained antigen, at the other end antibody and, in a broad intermediate zone, both antigen and antibody could be demonstrated in the supernatant fluid. They exDlained their results on a chemical basis, postulating that for complete preThe formation of precipitates on mixing anti-horse sera cipitation of one reagent, the other must be added in great excess. Eisenberg (1902) obtained similar results and also explained the phenomenon on the basis of an ...