SUMMARYRepeated selection of plants with unusually severe symptoms after their inoculation by insect vectors which had been injected with dilute inoculum from crude extracts of a stock culture (O strain) of rice dwarf virus (RDV) resulted in the emergence of a severe isolate (S strain). Of the 12 segments of RDV RNA, the fourth largest RNA of the S strain had an apparent Mr about 20000 larger than that of the corresponding segment of the O strain. The Mr of the protein corresponding to the M r 43000 protein of the O strain, which is located on the outside of the outer capsid, was 44000 in the S strain. The implication of the differences in the RNA and protein components between the S and O strains is discussed.Mutants and variants have provided some of the most promising materials for comparative studies of the mechanisms of biological functions of viruses, such as the pathological disorders they induce and their transmissibilities (Nishiguchi et al., 1985;Ohno et al., 1983). Since none of the viruses belonging to the plant reovirus group can be cloned by single local lesion selection, the isolation of mutants of these viruses has generally been considered to be difficult. However, mutants that lost transmissibility by the insect vector were obtained with clover wound tumour virus (Black et al., 1958) and rice dwarf virus (RDV) (Kimura, 1976a;Minobe et al., 1983).During the course of infectivity assays of purified preparations of RDV from infected rice leaves, a few rice plants showed unusually severe stunting which suggested that a severe strain of RDV had been selected. All the plants had been inoculated by injected insects. This severe (S) strain was selected from the ordinary (O) strain in the present investigation in order to study any deviation in the RNAs and component proteins from those of the O strain.RDV was maintained and propagated in rice plants (Oryza sativa L., cultivar Norin No. 29) which were inoculated by the viruliferous leafhopper, Nephotettix cincticeps.Severe symptoms, as judged by exceptionally short plant height, appeared with very low frequency when rice seedlings were inoculated by insects injected with purified preparations (Kimura, 1976b) of the O strain. Leaves showing the severest symptoms were macerated in four times their weight of 0.1 M-sodium phosphate buffer (pH 7.3), and the slurry was clarified in a Hitachi 20PR centrifuge at 1470 g for 15 min. The supernatant fluid was diluted 103-fold and used as the inoculum for the selection test. Second instar nymphs were each injected in the abdomen 2 weeks after hatching with about 0.2 ~tl of inoculum using a glass capillary with a 50 Ixm tip diameter. Injected insects were confined individually in glass test tubes each of which contained three healthy rice seedlings and 5 ml tap water. The exposed rice plants were replaced with new seedlings at intervals of a week and the exposed plants were transplanted to a greenhouse where they were observed for expression of symptoms for 60 days. The selection process was repeated 10 more times.S...