145 SEROLOGICAL STUDIES OF THE ROOT-NODULE BACTERIA. I. STRAINS OF EHIZOBIUM MELILOTI. By J. M. Vincent, School of Agriculture, University of Sydney.-[Read 25th June, 1941.] Introduction. Although a number of studies of the serological relationships of the root-nodule bacteria {RhizoMum spp.) have been undertaken (cit. Fred, Baldwin and McCoy, 1932), papers on this subject are comparatively rare in the more recent literature. Yet the earlier work gave some promising results both in confirming species diffei'entiation and, more especially, in distinguishing strains of the one species, e.g., Stevens (1925) working with strains of Rliizohium ineliloti. Whilst several of the early workers included a variety of serological techniques, precipitation, complement fixation and agglutination, the latter was most used and is probably still the most convenient since the reaction is readily observed and a serum of adequately high titre can be obtained with little difficulty. Little has been done, however, in bringing to the study of Rhizobium, techniques which have given so much information in respect of pathogens, as, for example, the serological differentiation obtained within the Salmonella group by White and Kauffmann (Savage and White, 1925; White, 1926, and citations Topley and Wilson, 1936). Bushnell and Sarles (1939) point out the need for improved techniques and use three types of antigen: whole-cell, heated and saline extracted. Still, their results do not give a very clear picture of the antigenic constitution of the organisms with which they were working (from soybean and lupine), and they have made no differential flagellar and somatic analysis. It seems likely that the serological reactions of the rhizobia will find more appli-cation as a means of differentiation between strains when we understand better the detail of the antigenic constitution of the cells, and attempt the correlation of this with other features of importance in the organisms' behaviour. Two simple points of technique merit further attention, viz.: (i) distinction between flagellar (H) and somatic (0) agglutination, and (ii) the application of serum absorption tests. So far as one is aware these refiiiements have not yet been reported with respect to RhizoMum although the keys to some of the tabulated results (Vogel and Zipfel, 1921; Wright, 1925) indicate that, at times, floccules characteristic of flagellar agglutination have been observed even though their significance has not been appreciated. The fact that readings were usually recorded after 24 hours would militate against distinction between H and since, after that time, in the presence of both types, the agglutinated mass would be compacted. This paper reports results when these techniques are applied to a number of strains of Rh. meliloti obtained from widespread areas and from various species of Medicago. Detailed attention has first been directed towards six organisms, for which antisera have been developed, in order to determine for these a minimal antigenic constitution — flagellar distinguished from somatic. These sera have then been used to study the serological relationships of 42 other strains. METHODS. Organisms used for the Development of Antisera. — These were selected to provide a variety of host species and to be representative of widely separated localities. Details are: Collection No. Host Plant. Locality. 47 Medicago sativa Bathurst, N.S.W. 74 M. sativa Roseworthy, S. Aust. 27 M. hispida var. denticulata Merrylands, N.S.W. 62 M. hispida var. denticulata Rcseworthy, S. Aust. 102 M. arabica Dandenong, Vict. 66 M. minima Tailem Bend, S. Aust. B
BioStor
