578 ELECTIVE LOCALIZATION OF BACTERIA 



similar changes might occur in various locaHzed but mild inflammatory processes or 

 foci of infection. This suggested that the tonsils, where these conditions prevail, might 

 be a good place to search for causative bacteria possessing elective localizing power. 

 Clinical evidence of the probable presence in these foci of the causative organism was 

 not lacking, for Billings' had reported good results from removal of such foci in a series 

 of cases. Two other important observations were made (one prior to the work on 

 mutation, the other during that period) which led to the use of proper methods in 

 the newly planned experimental work. The one concerned the importance of injecting 

 freshly isolated strains in order to produce specific lesions, first noted in a series of 

 experiments in subacute bacterial endocarditis/ and later in rheumatic fever ;^ the 

 other concerned the importance of oxygen tension for the successful isolation and 

 maintenance of specific or elective localizing power of streptococci, first noted in my 

 experiments in rheumatic fever. Localization in animals of the freshly isolated strep- 

 tococcus in these diseases was remarkably specific. The streptococcal strains from 

 tonsils or blood in cases of subacute bacterial endocarditis manifested marked affinity 

 for the valves of the heart and produced, soon after intravenous injection, severe 

 hemorrhage and later huge vegetations. Aside from producing lesions of the glomeruli 

 in the kidney, the strains had little tendency to localize and produce lesions in other 

 structures, such as joints, myocardium, and pericardium. The streptococcal strains 

 freshly isolated by me from tonsils, blood, stool, muscles, lymph nodes draining af- 

 fected joints, and especially from the joint fluid in rheumatic fever, manifested simul- 

 taneous affinity for joints, endocardium, pericardium, and myocardium, and those 

 from diseased muscles affinity for muscles as well. The lesions produced experi- 

 mentally resembled, in many respects, those noted in cases of rheumatic fever. The 

 early hemorrhage and later vegetation? on the valves of the heart were relatively 

 small, and the tendency to the formation of scar was marked. Many joints became in- 

 volved without suppuration or destruction of tissue. The lesions in the myocardium 

 often resembled Aschoff bodies. The number of organisms in the vegetations, and in 

 joint fluid of injected animals as of patients, was relatively small, and they were dif- 

 ficult to isolate. Lesions such as iritis, appendicitis, pyelonephritis, and ulcer of the 

 stomach, and symptoms resembling chorea occurred in a small number of injected ani- 

 mals, paralleling roughly the incidence of these affections in patients. Requirements 

 for causal relationship of the streptococcus isolated in cases of rheumatic fever ap- 

 peared fulfilled. After prolonged cultivation on artificial media and successive passage 

 through animals, the characteristic localizing power of both types of strains disap- 

 peared. The isolation from cases of rheumatic fever of streptococci possessing char- 

 acteristic localizing power has been reported since by Clawson^ and more recently by 

 Small^ and by Birkhaug.*^ 



' Billings, Frank: Anii. Int. Med., 9, 484. 1912. 



2 Rosenow, E. C: /. Infecl. Dis., 11, 235. 1912. 



3 Rosenow, 10. C: ibid., 14, 61. 1914. 

 •• Clawson, J.: ibid., 36, 444. 1025. 



5 Small, J. C: Am. J . M. Sc, 173, lor. 1927. 

 '■ Hirkhaug, K. Iv: ./. Iiifed. /)/,v., 40, 549. 1927. 



