It is conceivable that a change from the virulent, non-phagocytable S form of Pneumococcus to the avirulent phagocytable R form may take place in pneumococcus disease, but the experiments here reported do not settle the question whether or not this is an important factor in determining the outcome in natural infection. It has been shown experimentally that the degradation from the S form to the R form actually does take place in cultures of Pneumococcus growing in agar subcutaneously embedded in guinea pigs, in agar enclosed in vials subcutaneously embedded in rabbits, and spontaneously in the blood stream of infected horses. However, it was not possible in any of the experiments here cited to demonstrate the complete change from S to R pneumococci before the bacteria disappeared from the body. When the intermediate or R forms did appear, they were always accompanied and usually exceeded in number by the S forms and all three forms disappeared together. S organisms may disappear entirely without evidence of first going through the intermediate and R stages. On the other hand, contrary to expectations, pure cultures of R forms remained viable in subcutaneous foci for weeks although apparently freely accessible to the action of phagocytes. It seems of some significance that the R forms appeared early in the vials (inoculated with S pneumococci) in immunized and normal rabbits alike, indicating that the presence of demonstrable specific immune bodies was not alone responsible for the variation of the bacteria.
Of some importance also is the fact that R forms were never derived from similarly prepared control cultures growing in vitro at the same temperature and immersed in normal serum, although the S forms remained viable and unaltered for 6 weeks. It is likely that variations of pneumococci do not occur readily when S cultures are exposed to normal serum in vitro, especially when growing in closed vials under a diminished oxygen supply, for it has previously been shown (2) that only slight variation occurs even after prolonged (240) transfers in heterologous serum broth in the test-tube. It is possible, therefore, that the variation which occurred among pneumococci growing in agar vials embedded in normal rabbits was actually provoked by unknown influences in the living tissue fluids.
Although R forms have been shown to occur in vivo, no positive evidence can be derived from these experiments to prove that recovery from pneumococcus infection depends upon the degradation of the virulent S forms of pneumococci to the avirulent R forms and the subsequent destruction of the latter by phagocytes.