To the infectious abortion of mare has hitherto not been drawn much attention here in Japan; on the contrary there has been published a number of reports on the same disease in cow. In 1913, however, an enzoötic abortion of mare developed in a government stud at Kamikita district in Aomori prefecture, one of the main horse habitat in Japan. The disease has been proved to be contagious in all respects. During the winter season from 1914 to 1915 occurred again many, cases of the disease in the neighbouring farms. It is much to be regretted that on that occasion no precise etiological investigations have been carried out. In the begining of the year 1920, however, prevalence of the same disease was reported again, and the authors have made some pathological and bacteriological examinations upon those cases at the requirement of the chief of the Horse Administration Bureau. During about one month's stay at that locality the authors thoroughly examined 10 foetus which were brought there from some distant places; the cases were found to be sporadic, contrary to the authors' expectation, no contagious nature being traced in any of them. In 8 cases, no particular pathological change was found both in the foetus and placenta, even in bacteriological examinations they revealed to be sterile or contaminated with the usual saprophites. Other two materials, however, showed equally pronounced septicaemic changes, giving a bacteria of paratyphus B-enteritis group in pure culture from heart blood, stomachal and intestinal contents, as well as from fluid of pleural, peritoneal and pericardial cavities.It has been comfirmed by authors' investigations in morphological, biological, chemico-bacteriological and pathogenic respects, that the bacillus is identical with what is termed as bacillus abortus equi etc. Serum taken from one of the mother mare was able to agglutinate these 2 strains in a dilution of 1:5.120 or 1:10.240, and to fix complement completely in a dose of 0.025c.c. with extract of one strain. That from another mare, though she has given an infected foetus, gave no positive reaction at all in the same examinations. The normal agglutination titre of the sera from 26 mares, which had nothing to do with abortion, and from a gelding and a stallion was as high as 1:80 (53.6%), the maximum attaining 1:320. Thereupon the authors performed agglutination and complement fixation tests with the bacillus and sera of 32 mares, which were subject to abortion, during the last 100 days in the same district. Out of 7 specimens obtained in a village, 5 were found to agglutinate the bacilli in a dilutions of from 1:640 to 1:10.240, though the complement fixation test has failed in most of them, except two specimens, which have shown high agglutination titres. Hence the authors may safely say that in the same village an infectious abortion due to the bacillus obtained prevailed epizoötically, and those cases in other places had no relation with the bacillus and were probably of no infectious nature.From the above results the authors conclude as follows:1) There is in Japan an infectious abortion of mares due to B. abortus equi.2) It occurs often epizoötically and may also be found sporadically without infecting even the nearest pregnant mares.The bacillus isolated by the authors is not identical not only serologically, but also biologically with Bacillus paratyphosus B in men. The strains obtained by the authors form dry and wrinkled tender colonies on artificial solid media and a distinct film on the surface of fluid media; all these facts agree well with the descriptions of other investigators on this sort of bacillus. Moreover, the authors found that the strains isolated were remarkably inhibited in growth on malachit green agar, almost in the same degree as in the case of B. coli, the medium having no influence on B. paratyphosus B.