Our conclusions from the literature and our own experiments may be summarized as follows:
I. Blood for bacteriological examination during life should be taken directly from the veins and in considerable quantity.
II. Resorption of toxines is the most important feature in cases of sepsis; pyogenie bacteria invade the general circulation in a rather small proportion even of severe eases, and, as a rule, late in the course of the disease.
III. A general infection by the pnenmococcus can be demonstrated occasionally in the late stages of acute lobar pneumonia.
IV. The value of blood cultures as a means of diagnosis in obscure cases of sepsis is limited by the fact that invasion of the blood by the specific organism cannot be demonstrated during life in the majority of cases. Positive cultures are very valuable; negative cultures do not exclude local septic infections.
V. The detection of specific bacteria in the blood of cases of sepsis and of pneumonia gives a very unfavorable prognosis in most cases.
VI. General terminal infections with pyogenic cocci occasionally occur as an immediate cause of death in chronic disease. Local infections processes play this part more frequently.
VII. As far as our experiments have shown, invasion of the blood by bacteria during the death agony, with subsequent distribution of the genus to the organs by the circulation, is a rather uncommon occurrence.
VIII. Owing to the relative infrequency of agonal invasion, we believe that in the majority of cases where the autopsy is performed promptly after death, the bacteria which are found in the organs succeeded in reaching these organs previously to the death agony, and are associated with the course of the disease.
IX. The presence of bacteria in the organs of late autopsies is due in many cases to post-mortem extension from one organ to another, and in some cases to the post-mortem growth of small numbers of genus which were distributed to the organs by means of the circulation.