Summary and Discussion By cultivating Leishmania brasiliensis, Leishmania tropica, Leishmania donovani, Leishmania infantum, Leishmania canis and Leptomonas ctenocephali in Noguchi's medium for Leptospira, in which anti- Leishmania brasiliensis serum was used instead of normal serum, the author was able to verify that only when the flagellates were heterologous in relation to the serum employed could their growth take place; otherwise, their lysis occurred in the presence of homologous serum. The only explanation for the lysis is the development of lytic specific antibodies in the rabbits used in the preparation of the agglutinating sera. These rabbits received about 7 cc. of a culture of living flagellates, in five inoculations. The complement indispensable to the production of lysis must have proceeded, not from the immune serum, which as a rule was not of recent preparation, but from rabbit the hemoglobin added in the culture medium. The discordant results, obtained whenever the medium was kept for two months before being planted, seem to confirm this opinion. We also obtained lysis with a homologous strain, and absence of lysis with heterologous ones, by using, as an antigen, Leishmania brasiliensis and Leishmania tropica grown in Noguchi's culture medium, washed and suspended in salt solution, anti-Leishmania serum concentrated to 1:10 being added to the opalescent suspension. The only explanation for the lysis in these cases may lie in the existence of complement in the serum used, because of its being of very recent preparation. The phenomenon of lysis as produced by normal sera upon protozoa has already been the subject of some detailed study particularly in regard to certain Trypanosomata. While working with Leishmania brasiliensis Vianna, 1911, we observed that a non-specific lytic action of this parasite may be caused by highly concentrated rabbit normal serum. That this lysis was really due to the combined action of antibodies and complement it seems to have become clear in the light of the following experiment. Three hanging drops were prepared with a strain c of Leishmania brasiliensis of our collection: No. 1, with one drop of culture in Noguchi's media for Leptospira. No. 2, with one drop of the same culture added to one drop of rabbit normal serum inactivated at 55° for half an hour. No. 3, with one drop of the same culture added to one drop of non-inactivated rabbit normal serum. The examination performed fifteen minutes later disclosed: In nos. 1 and 2, aspect of a well developed normal culture. In no. 3, all the flagellates motionless and unable to recover their motility. The flagellates became less motile in the beginning, then suddenly stopping; turgence and vacuolization might be seen, the protoplasma becoming granulous as in degenerated forms. This non-specific lytic action is to be observed only in high concentration. Therefore, it can not be considered a cause of error in the reaction we described in this paper, since we know that the Leishmanias may be readily cultivated in culture media having one-tenth or even more of concentration of normal serum (Noguchi's media for Leptospira, by instance). Noguchi already made some references to the lysis of Leishmania in pure rabbit serum, in a work published in “Proceedings of the International Conference on Health Problems in Tropical America” (9). He, however, stated that the lysis is no more present in a dilution of 1:10. The results herein reported on appear to warrant the distinction of the genus Leishmania Ross, 1903, as represented by an anti-Leishmania brasiliensis Vianna, 1911 immune serum, from the genus Leptomonas Kent, 1880, as represented by Leptomonas ctenocephali (Fantham, 1912). Cross-reactions with more numerous species of Leishmania and Leptomonas must have been performed to confirm the experiments, for no one could say that the fact of anti-brasiliensis serum not being lytic for Leptomonas ctenocephali means that specific sera against other species of Leishmania are not lytic as well; nor can any one affirm a priori that an anti-Leptomonas serum is not able to produce lysis of the Leishmania strains. Considering, however, the results obtained by other workers with different methods, this disagreement is not probable. The results obtained in the lysis experiments seem to warrant the distinction between Leishmania brasiliensis Vianna, 1911, on the one hand, and Leishmania tropica Wright, 1903, Leishmania infantum Nicolle, 1908, Leishmania donovani (Laveran and Mesnil, 1903), and Leishmania canis Nicolle, on the other hand. Therefore, they confirm the specific differences pointed out by both Noguchi, Kliger and others. 1 Another conclusion we may draw from this paper concerns the close antigenic affinity among the 13 species of Leishmania brasiliensis employed, some of which were isolated in Brazil and others in Argentina. Contrary to Noguchi's hypothesis, this observation appears to demonstrate the probability of the Leishmania strains of American leishmaniasis belonging to but one serological species. This species is perfectly distinct from Leishmania tropica (in spite of the affinities shown in the pathogeny of the infections they both determine on humans) because of the differences noticed in the infections they cause experimentally. The results obtained by the use of the above described technic seem to indicate that it deserves a place among the reactions proposed for the specific diagnosis of this important and complex group of flagellates, provided that some technical conditions dealing in short with the principles of bacteriolysis be firmly established.