No abstract

In the preceding study of the relationship of the natural pneumococcidalpromoting activity of the serum to bacteremia in dogs with experimental pneumococcal pneumonia (1), it was found that the blood of non-bacteremic dogs retains its natural humoral immunity with a remarkable degree of constancy. The occurrence of bacteremia was followed by wide individual variations in the concentration of these antipneumococcal properties, ranging from complete disappearance in the presence of relatively few circulating pneumococci to their persistence in blood containing great numbers of microorganisms. These fluctuations were interpreted to signify differences in the body's capacity to elaborate new antibodies in response to the continued escape of pneumococci from the pulmonary lesion. That the observed variation in pneumococcal immune substances represented corresponding degrees of pneumococcidal power of the blood seemed evident from our tests performed with the shed blood; but to what degrees this action of the blood in vitro reflects the ability of the body to deal with pneumococci in the circulation, our observations failed to elucidate, nor did they provide any data as to whether the pneumococcusclearing action of the blood is more or less effective in the diseased animal than in the normal one.It was felt that a profitable approach to these questions could be made by observing the effects of artificially induced pneumococcemia in normal and diseased (pneumonic) animals and that such information would contribute to the solution of certain other fundamental problems presented by the bacteremic state, such as the rate of escape of pneumococci from pulmonary lesions, the capacity of the blood stream to dispose of them, and the mechanism by which the body controls the degree of blood invasion in the absence of demonstrable humoral immunity.

Studies on bacteremia in pneumococcal lobar pneumonia have brought forward accumulating evidence to indicate that this form of pneumococcus infection in the human being is characteristically a localized disease. The older view that lobar pneumonia in its beginning stages is accompanied by a transient invasion of the blood stream is not supported by the results of blood cultures in early cases. While bacteremia may occur within the first 24 hours of the disease, the great majority of cases exhibit a sterile blood at this time (1, 2). x The marked alteration in prognosis occasioned by finding pneumococci in the blood stream at any period in the disease is so generally recognized that it is mentioned here only to emphasize the importance of this phenomenon about which so little is known. The conception that bacteremia indicates a breaking down of the body's defence mechanism is probably correct, since it has been observed that the more extensive the pulmonary involvement and the lower the white blood count the greater the likelihood of bacteremia (3), and that following the onset of baeteremia an increasing number of pneumococci in the blood is associated with a rapidly rising mortality. However, we have little information as to how or why pneumococci escape from the pulmonary lesion.Of the various factors in bodily resistance, which might operate to control the liberation of pneumococci from the lung into the blood stream, that of * Present address:

No abstract