The question of the relationship of streptococci to the etiology of infectious arthritis and of rheumatic fever is of the utmost importance. If a streptococcus or group of streptococci could be shown to be associated See PDF for Structure with either disease, some form of specific treatment might be available. The possibility of primary streptococcic infection as the cause of rheumatic fever, and, to a less extent, of acute infectious arthritis would seem to be a reasonable conjecture because of the frequency of associated throat, sinus or other focal infection. To consider that these same streptococci remain in or about the affected joint and to such an extent that they are found in the blood stream in cases of chronic infectious arthritis of years' duration demands a rather unique conception. Recent investigative work has certainly tended to confirm the importance of streptococci in these diseases, but, if all the published reports are considered as a group, one can not help being impressed with the inconsistency and peculiarities of the findings. In blood cultures from cases of rheumatic fever Clawson (7) recovered Streptococcus viridans, Small (8) and Birkhaug (9) non-hemolytic (gamma type) streptococci, and Cecil et al. (3) Streptococcus viridans, rarely hemolytic and non-hemolytic streptococci. In blood cultures from cases of infectious arthritis Cecil et al. (2) recovered attenuated hemolytic streptococci and occasionally Streptococcus viridans, non-hemolytic streptococci and diphtheroids and Margolis and Dorsey (10) green-producing and indifferent streptococci and diphtheroids, whereas from synovial fluids and regional lymph nodes Forkner, Shands and Poston (4) and Poston (5) obtained Streptococcus viridans and Margolis and Dorsey (11) recovered green-producing and indifferent streptococci and diphtheroids from epiphyseal marrows, bones and synovial membranes. On the other hand, Jordan (12) and Nye and Seegal (13) and the work reported in this paper have failed to confirm these findings. If a streptococcus is the infective agent, it is difficult to explain why the organisms recovered so consistently by certain groups of investigators are so different and why the findings of other groups are entirely negative. The chances of contamination, even with the most careful manipulation, are extremely favorable when using a cultural technique which demands subculturing a fluid medium every 3 to 5 days for a period of 4 to 6 weeks. The question arises as to what organisms should be considered as contaminants. Cultures found to contain Staphylococcus albus have never been judged significant. Margolis and Dorsey (11) have excluded Gram-positive bacilli from their series, but have included diphtheroids. Cecil et al. (2) reported the recovery of diphtheroids and Micrococcus zymogenes from blood cultures which they did not consider contaminated. Jordan (12) recovered short Gram-negative bacilli and Gram-positive bacilli and questioned their importance. The occurrence of such bacteria would appear to be not unlike that of the non-hemolytic (gamma type) streptococci reported by Small (7), Birkhaug (9) and, rarely, by Cecil et al. (2) and Margolis and Dorsey (10,11). In the work reported in this paper and in previously published work (13) with blood cultures from cases of rheumatic fever, staphylococci, Gram-positive bacilli and diplococci and diphtheroids have been isolated from a certain number of the cultures. Since these were found, as a rule, in only 1 of 2 or more cultures from the same blood and since such organisms occurred about as frequently in cultures from control cases, they have been considered merely contaminants. The findings in the 9 cases of arthritis with positive blood cultures reported by Margolis and Dorsey (10) are certainly far from convincing. These authors found green-producing streptococci in 5 of the 9 cases, but duplicate cultures on the same blood, with one exception, were always negative and the two repeat cultures were negative. Furthermore, in the 3 cases yielding indifferent streptococci in one subculture all subsequent subcultures were negative. In the work reported in this paper a green-producing and a non-hemolytic streptococcus were isolated from blood cultures, but a Gram-positive diplococcus and a Gram-positive bacillus, respectively, were recovered from the same cultures. A consideration of the above would seem to point to the probability, even certainty, of streptococci occurring in some cultures as contaminants; and the work of Olitsky and Long (14) and Long, Olitsky and Stewart (15) has clearly demonstrated that the air contamination of cultures of ground material with non-hemolytic green-producing streptococci can occur just as easily as with diphtheroids and staphylococci. It is obvious, in such cases, that the types of organisms recovered are dependent on the flora of the air of the laboratory or of the throat of the laboratory worker, and this point may well explain the differences in the cultural findings under consideration. A number of years ago numerous articles (16-19) appeared relative to the bacteriologic flora of lymph nodes, particularly those from cases of Hodgkin's disease, and it is interesting to note that the organisms recovered were quite similar to those which have been recovered from the blood and tissues of cases of arthritis and rheumatic fever, with the exception that the diphtheroids were, at that time, much more in prominence. Regardless of elaborate serological studies and animal experiments, streptococci must be recovered consistently by several groups of laboratory workers before their etiologic rôle in chronic arthritis and rheumatic fever can be accepted. Duplicate cultures and repeat cultures should yield the same organism in a generous percentage of cases and cultures from cases of other diseases or from normal persons should be negative. Positive cultures from duplicate cultures opened only at the time when the first culture showed growth would make the findings more significant than if the whole series were subcultured every 3 to 5 days. The work of Cecil, Nicholls and Stainsby on the bacteriology of the blood and joints in chronic arthritis and rheumatic fever has apparently been carried out most carefully and thoroughly and their results are very consistent and convincing. In spite of attempts to follow their methods in the selection of patients and in cultural technique, the results on the relatively small series of cases which are reported here fail to confirm their findings.