Comparing the pneumonic lesions produced by this very virulent streptococcus with those caused by a mildly virulent strain of the same organism, we find that with one exception the differences between them are only of a quantitative nature and not striking. There are the same intense leucocytic exudation into the alveoli and the leucocytic infiltration of the framework of the lungs. Macroscopically the pneumonia produced by the virulent organism was, as a rule, lobular in character. Occasionally, however, especially when large doses of the culture had been given, the gross appearance of the lesion gave the impression of a lobar pneumonia. But even in these cases the cut surface showed that the solid areas were separated by softer and lighter areas of aerated, congested, or edematous lung substance. Pleurisy was practically absent from the lungs in this series of experiments; only one case showed a narrow strip of fibrinous exudate on the pleura. Although in some instances large doses of the cultures were given and some dogs were permitted to live six or seven days, there was no mortality among the dogs in these experiments, just as there had been none among the animals experimented upon with the streptococcus of very low virulence. The course of the pneumonia caused by the virulent organism seemed to be somewhat longer than that produced by the less virulent one; areas of resolving pneumonia were found to persist on the sixth and seventh days after the injection, and in one such instance the solid focus contained viable organisms. There was some difference in the amount of fibrin present in the exudate caused by the two strains of streptococci. Although even in these experiments the amount of fibrin present was less than that found in the lesion produced by the virulent pneumococcus, it was perceptibly larger than the very small amount found in the lesions caused by the less virulent strain of streptococcus. It should be mentioned that there was no difference in the degree of phagocytosis observed in these two series of experiments. It was not marked in either case. One point of difference in the two series was noted: during the first forty-eight hours after insufflation of the virulent strain of streptococcus the blood of the animals obtained from the jugular vein during life and from the heart after death contained living organisms, while in the experiments with the slightly virulent streptococcus no living cocci were recovered from the blood. The virulence of the strain of streptococcus employed in the present investigation was, as has been pointed out, similar to that of the pneumococcus employed in the investigation of Lamar and Meltzer and in many of our own experiments, as far as mice were concerned, since both killed mice weighing fifteen grams in doses of 0.000,001 of a cubic centimeter in twenty-four to thirty-six hours. The findings in the lesions caused by these two organisms are comparable and therefore some deductions may be drawn regarding the similarity or dissimilarity of their actions. One of the first things we wish to bring out is the correction of our former statement regarding the presence of the cocci in the blood. As our new investigation shows, there is evidently no difference in this respect between the streptococcus and the pneumococcus. When both organisms are highly virulent they can be found in the circulating blood of the living dog or in the heart's blood of the dead dog during the first twenty-four or forty-eight hours after an intrabronchial insufflation. On the other hand, the present investigation rather confirms in a general way our first contention that the pneumonic lesions produced by the two different organisms differ distinctly in many ways. There is the difference in the mortality; even with fairly large quantities of the culture of the virulent streptococcus the mortality was nil in this series, although some of the dogs were kept alive for six and seven days. The mortality of the pneumococcus infection depends upon the quantity insufflated and may amount, as Lamar and Meltzer pointed out, to 16 per cent. The gross appearance of the lesion produced by the pneumococcus is, as frequently stated, that of a lobar pneumonia, frequently accompanied by a fibrinous exudate on the pleura. The gross appearance of the lesion produced even by a virulent streptococcus is, as a rule, that of a lobular pneumonia and is practically never accompanied by pleurisy. In the few instances in which large quantities of the streptococcus were given and the outward appearance of the lungs approached that of a lobar pneumonia, it was found that on section of the consolidated lung the solid foci proved to be separated by lighter areas of aerated, congested lung. In the lesions produced by the virulent streptococci the walls of the finer bronchi and the framework of the lung were markedly infiltrated with leucocytes, while in the lobar pneumonia produced by the pneumococcus the framework remained free from such infiltration. Finally there is the unmistakable difference in the amount of fibrin present in the alveolar exudate. While in the exudate of the lesion produced by the virulent streptococcus the amount was perceptibly larger than the insignificant amount present in the lesion caused by a slightly virulent strain, it is not to be compared with the large amount of fibrin which exists in the exudate of pneumococcus pneumonia. The several investigations which we have carried out seem to show conclusively that in general the streptococcus causes a lobular pneumonia which, besides the leucocytic intra-alveolar exudation, is characterized by a leucocytic infiltration of the lung framework, and that the pneumococcus causes a lobar pneumonia, which is practically free from leucocytic infiltration of the interstitial tissue of the lung. Furthermore, a virulent pneumococcus causes a lesion in which fibrin is a prominent element in the exudate and that element distinguishes the exudate sharply from the exudate of the lesion caused by a virulent streptococcus in which fibrin is present only in moderate amount. It distinguishes it in a still more striking manner from the exudates of the lesions caused by non-virulent pneumococci or streptococci, in which fibrin is present only in very small amounts. It seems that the formation of fibrin is connected in some specific way with the pneumococcus on the one hand, and with the virulence of the organism on the other. For even with the relatively small amounts of fibrin present in the exudates of lesions caused by the streptococcus there is a perceptible difference in the quantity according to the virulence of the organism. Whether fibrin is a means which enhances virulence, or whether it is a reaction product against it, our experiments so far do not entitle us to discuss.