Just as an unidentified _disease_ in an animal may be determined by testing its serum as above described against _known_ kinds of bacteria, so _unknown bacteria_ isolated from an animal, from water, etc., may be identified by testing them against the _blood sera_ of different animals, each of which has been properly inoculated with a different kind of _known bacteria_. If the unknown organism is agglutinated by the blood of one of the animals in high dilution, and not by the others, evidently the bacterium is the same as that with which the animal has been inoculated, or _immunized_, as is usually stated. This method of identifying cultures of bacteria is of wide application, but is used practically only in those cases where other methods of identification are not readily applied, and especially where other methods are _not sufficient_ as in the “intestinal group” of organisms in human practice. The diagnosis of disease in an animal by testing its serum is also a valuable and much used procedure. This is the method of the “Widal” or “Gruber-Widal” test for typhoid fever in man and is used in veterinary practice in testing for glanders, contagious abortion, etc. In some cases a dilution of the serum of from 20 to 50 times is sufficient for diagnosis (Malta fever), in most cases, however, 50 times is the lowest limit. Evidently the greater the dilution, that is, the higher the “titer,” the more specific is the reaction. PRECIPITINS. Since agglutinins act on bacteria, probably through the presence of substances within the bacterial cell, it is reasonable to expect that if these substances be dissolved out of the cell, there would be some reaction between their (colloidal) solution and the same serum. As a matter of fact Kraus (1897) showed that broth cultures freed from bacteria by porcelain filters do show a precipitate when mixed with the serum of an animal immunized against the particular bacterium and that the reaction is specific under proper conditions of dilution. It was not long after Kraus’s work until the experiments were tried of “immunizing” an animal not against a bacterium or its filtered culture, but against (colloidal) solutions of proteins, such as white of egg, casein of milk, proteins of meat and of blood serum, vegetable proteins, etc. It was ascertained that in all these cases the animal’s serum contains a substance which causes a _precipitate_ with solutions of the protein used for immunization. The number of such precipitating serums that have been made experimentally is very large and it appears that protein from any source when properly introduced into the blood or tissues of an animal will cause the formation of a precipitating substance for its solutions. This substance is known, technically as a “_precipitin_.” The protein used as antigen to stimulate its formation, or some part of the protein molecule (haptophore group), which acts as stimulus to the cell is spoken of as a “precipitinogen,” both