In agreement with previous work (6) it was found that in normal rabbits after a primary inoculation of 0.01 mg. of human tubercle bacilli the organism shows a preliminary growth, which commences at once in all the organs, and reaches its height in the second week in the liver, spleen, and bone marrow; thereafter destruction begins, and is practically complete by the sixth week in the liver and bone marrow and less complete in the spleen. In the lung and kidney the largest number are isolated in the fourth week and thereafter there is a tendency to destruction, but to destruction far less complete than in the other organs, for even two months after infection large numbers of tubercle bacilli persist. Associated with this multiplication of the human tubercle bacilli were found extensive or moderate tuberculous lesions in the lung and kidney with moderate or slight changes in the liver, spleen and bone marrow. In contrast to these observations, it was found that in rabbits reinfected with the same quantity of human tubercle bacilli, the organism was destroyed immediately without any preliminary multiplication in the liver, spleen and bone marrow, though a few bacilli persisted in these organs even two months after reinfection. Neither could definite evidence of multiplication of the tubercle bacilli of reinfection in the lung and kidney of these rabbits be obtained; although variable and at times very large numbers of tubercle bacilli persisted in these organs, probably due to the primary infection. Nor were tuberculous changes found in any of the organs due to reinfection. Again in rabbits that received a primary inoculation of 0.01 mg. of bovine tubercle bacilli the organism multiplied in all the organs. This multiplication reached its height in the liver, spleen and bone marrow between the fourth and sixth week, instead of in the second week as with human bacilli. Thereafter destruction was practically complete in the liver and much less complete in the spleen and bone marrow; in the lung and kidney multiplication continued unabated to the second month. Associated with these bacteriological data, there developed a massive pulmonary tuberculosis with extensive or moderate disease in the kidney, spleen, bone marrow and liver. In rabbits reinfected with the same quantity of bovine tubercle bacilli after a primary infection with human tubercle bacilli, the organism, in all but a few instances, was destroyed immediately in the liver, spleen and bone marrow. Here again, with the bovine type as with the human type, after reinfection as after primary infection, the destruction was not quite complete and a few isolated bacilli persisted even two months after reinfection. As to the lung and kidney, evidence was obtained that the variable and at times very large numbers of tubercle bacilli that were found in these organs were of human type in their cultural characteristics, and that in these organs also, the bovine tubercle bacilli of reinfection failed to grow. There was also as a rule complete absence of any tuberculous lesions in the liver, spleen and bone marrow. Nor were there any tuberculous changes found in the lung and kidney due to the reinfection, but those variable residual lesions that were found were present before reinfection, as shown in the lung by x-ray photographs. There were a few instances in which restricted multiplication of the virus took place and slight tuberculous lesions developed after reinfection in rabbits in which the primary lesions had all but disappeared. Thus in rabbits having considerable residual lesions from a primary human infection the tubercle bacilli of reinfection, whether human or bovine, are destroyed immediately though incompletely without any preliminary multiplication. Yet these rabbits, which so efficiently destroy the more virulent bovine bacilli of reinfection introduced from without, at times harbor tremendous numbers of human tubercle bacilli in the old lesions of the lung and kidney. Parallel with the bacteriological and pathological studies of these rabbits, the serum of some of them was studied for the presence of circulating antibodies by McCutcheon, Strumia, Mudd (S), Mudd (E. B. H.) and Lucké (7). These investigators found that on primary infection there was only a slight and slowly developing production of agglutinating and phagocytosis-promoting antibodies whereas, in the reinfected rabbits, which showed an immediate destruction of the bacilli, these antibodies rose promptly and in relatively large amount. They suggest that the prompt rise of antibodies in reinfected animals may play a rôle in the immunity to tuberculosis. In the previous study, which was preformed in 1927, emphasis was placed upon the comparatively slower original rate of growth of the bovine bacillus in the rabbit as compared with that of the human type, and in this delay was seen a partial explanation of the greater virulence of the bovine organism for this species. The more rapid the original growth the more rapid was the following destruction. In this series of experiments, which were performed early in 1929, although the delayed destruction of the bovine type in the rabbit was amply confirmed, the original rate of growth is not any slower than the human type. Coincident with this increasingly more rapid rate of growth of the bovine bacillus in the body is a gradual decrease in the virulence of the organism. In 1925 and 1926 0.01 mg. of this bovine strain regularly killed rabbits in about 30 days after intravenous inoculation (8), but in this experiment none of the 3 rabbits died within 65 days after a similar infection and 2 rabbits were still living 110 days after injection.