Three strains of pneumococcus (types I and III), equally sensitive to penicillin, have been shown to be killed by the antibiotic in vitro when grown either in enriched beef infusion broth or in a thin serous exudate. Killing of the bacteria resulted promptly when the penicillin was added during the logarithmic phase of growth but failed to occur if addition of the antibiotic was delayed until the later "stationary" growth phase. In analogous experiments with thick purulent exudates from established subcutaneous abscesses, the pneumococci failed to grow rapidly, and added penicillin exerted only a relatively slow bactericidal effect.
The relevance of these in vitro observations to the curative action of penicillin was demonstrated in a systematic histologic study of the antimicrobial effect of the drug in experimental (type I) pneumococcal pneumonia. Evidence was obtained that at least two distinct processes are involved. The first, the direct bactericidal effect of the penicillin itself, was shown to operate in the outer edema zone of the spreading pneumonic lesion where the micro-organisms multiply rapidly in the thin serous exudate. The second, which predominates in the older more central portions of the lesion, was demonstrated to depend upon destruction of the pneumococci by phagocytosis. Here the bacteria, having presumably reached a relatively stationary phase of growth in the alveolar exudate, are resistant to the bactericidal action of the penicillin but are readily destroyed by the phagocytes.