It has been previously reported that a filterable microorganism belonging to the genus Leptospira has been recovered from the blood or organs of human beings suffering from the disease known as yellow fever in Guayaquil, and that the organism, which has been termed Leptospira icteroides, induces in certain experimental animals the characteristic symptoms and lesions observed in the patients from whom it was isolated. It has also been previously shown that the serum from patients recovering from an attack of yellow fever in Guayaquil had the power to agglutinate and dissolve the organism when introduced into the peritoneal cavity of a normal guinea pig (Pfeiffer phenomenon). Moreover, the guinea pigs which had once been inoculated with the blood of yellow fever patients without succumbing to the infection, notwithstanding the fact that they had shown a definite febrile reaction after 4 to 5 days, were found to be refractory to a subsequent inoculation of a culture of Leptospira icteroides All these observations pointed to the possible relation of this organism to the disease known as yellow fever in Guayaquil. The demonstration of the filterability of the organism and the transmission of the infection with the same organism by Stegomyia calopus have further strengthened the probable etiological significance of the organism in yellow fever. It was by no means a simple problem to determine the relation existing between Leptospira icteroides and Leptospira icterohæmorrhagiæ. An experiment reported in a previous paper seemed to justify the view that the two leptospiras are closely related but not identical, yet it was necessary to exhaust various other modes of differentiation before the distinction between them was firmly established. The present paper continues this phase of the inquiry in further detail. There have been taken up here the phenomena of agglutination, the reaction of Pfeiffer, complement fixation, the protective properties of various monovalent and polyvalent immune sera, and active immunity. As the result of experiments in connection with these immunity phenomena the following data are presented. Monovalent immune sera prepared by several successive injections in an animal naturally refractory to Leptospira icteroides possess the power to agglutinate in vitro not only the homologous strains, but also all other strains of icteroides tested. On the other hand, a slight effect, or none at all, has been observed when these immune sera have been mixed in vitro with various strains of Leptospira icterohæmorrhagiæ. A similar relation exists between the monovalent anti-icterohæmorrhagiæ sera and the various strains of Leptospira icteroides; that is, there is a slight agglutinating effect in some instances upon the icteroides strains, but it is never so strong as that occurring in tests against the icterohæhagiæ strains. The Pfeiffer reaction gave a sharper differentiation between the two groups, for in most instances the phenomenon was specific for the group. There were occasional doubtful reactions, but not enough to warrant a confusion of the two groups. Polyvalent immune sera, one specific for icteroides, and the other for icterohæmorrhagiæ, showed a high titer of neutralizing power for the cultures of the homologous groups. It was found, however, that the action of the sera is by no means absolutely specific, because the injection of a sufficient amount of the anti-icteroides serum apparently prevented a fatal outcome in a guinea pig inoculated with multiple minimum lethal doses of a culture of Leptospira icterohæmorrhagiæ, and vice versa. The specificity of the serum was demonstrated only when it was used in smaller quantities. More or less specificity was shown by the complement fixation reaction, but it was not absolute. Weak fixation occurred when the anti-icteroides serum was mixed with one or the other of the icterohæmorrhagiæ strains and vice versa, and strong fixation occurred only when the antiserum was mixed with one of the icteroides strains. The question naturally arises whether or not this apparent specificity is due to the homology of the serum and not altogether to a difference in genus of the strains. In other words, it is justifiable to question whether all these variations in the degree of intensity of the reaction are not due to strain variations of the same genus. This question is not finally settled by the present investigation, in which only four icteroides and nine icterohæmorrhagiæ strains have been carefully studied. Nevertheless, on the basis of the findings with these thirteen strains, it seems probable that Leptospira icteroides and Leptospira icterohæmorrhagiæ are closely allied but are nevertheless distinct in their immunological reactions. Perhaps the difference between the two may amount to that between subspecies or races. It has been pointed out earlier that the pathogenicity of the two is also distinct, inasmuch as icteroides produces chiefly icterus and nephritis and icterohæmorrhagiæ hemorrhage and nephritis, the icterus being less and the hemorrhage more prominent in the evolution of the latter infection. In the study of active immunity—exclusive of vaccination—difficulty has been experienced in the evaluation of the results, owing to the existence of natural resistance to infection among guinea pigs. A guinea pig may recover from the inoculation of Leptospira icteroides and then resist a subsequent inoculation with a virulent strain of Leptospira icterohæmorrhagiæ, a condition simulating that brought about by the identity of the two organisms. However, the refractoriness of such an animal to icterohæmorrhagiæ may be due to its natural immunity to it. In the present study, therefore, only those guinea pigs were selected which had reacted typically—though in mild degree —to the icteroides infection, in order to determine whether they were subsequently immune to the inoculation of icterohæmorrhagiæ. Indeed, by this mode of experimentation it was found that the guinea pigs which had once passed through an attack of the icteroides infection were absolutely immune to a second infection with the same organism but reacted severely and sometimes fatally to a later inoculation of icterohæmorrhagiæ. Although there were a number of instances in which a previous infection with icteroides did not confer any perceptible immunity upon the guinea pigs against icterohæmorrhagiæ, another group of guinea pigs showed a considerable resistance to the icterohæmorrhagiæ infection as compared with those which had never been inoculated with icteroides. There is not much doubt, therefore, that an icteroides attack brings about, in some instances at least, a certain degree of resistance to the icterohæmorrhagiæ infection. Hence the study of the phenomena of active immunity strongly indicates that icteroides is closely related immunologically to icterohæmorrhagiæ.