By inoculating into the tail vein of mice 40 samples of culture positive bacilli obtained from different kinds of the excised pulmonary foci and by the estimation culture of the viscera of these test animals, the author studied the virulency of the tubercle bacilli, and also carried out observations on the relationship between the color reaction and the ability of cord formation. 1. Tubercle bacilli in the excised pulmonary focus generally tend to show a less potent virulency that that of H(37)Rv, human F strain, and ones isolated from human sputum. 2. However, the tubercle bacilli from the cavity expelling becilli up to the time of operation show the virulency almost as potent as the tubercle bacilli isolated from sputum. The longer the time after the cessation of bacillus expulsion the weakeer the virulency tends to be. In the caseous focus the virulency of bacilli tends to be less potent than in the case of the cavity. In addition, the softer the contents of the focus the weaker is the bacilli. 3. Tubercle bacilli in sll kinds of foci show no fixed relationship between the virulency on one hand and the size of focus, the kind and amount of drug administered befoer operation, and collapse therapy or no collapse theapy on the other hand. 4. Of various stainings, neutral red reaction in the case of extremely strong or extremely weak stain coincides well with the degree of virulency in the animal experiment, but in the case staining in an intermediate degree, although coinciding well with the potency of virulency, it does not necessarily agree with the results of animal experiment. It seems that this method is fairly useful for a rough estimatiou of the virulency of tubercle bacilli but it is not adequate enough for a precise analysis. Furthemore, P-K reduction test will yield the degree of virulency to a certain extent for the virulent strain, but for the determination of the degree of virulency in human infected bacilli, irrespective of the potency observed in animal experiment, no maked discoloration can be observed by this thst and it seems to possess no great significance. 5. No fixed relationship between the resistance against SM and PAS and the virulency toward mice, but in the case of INAH-resistant strain that acquired the resistance of over 10 γ, a weakening tendency can be recognized in their virulency. Such a tendancy has also been observed in the N-R reduction test. In addition, no decrease in the coloration can be recongized by the N-R reaction in SM-highly-resistant strain. 6. The cases showing the colony formation even after a long period of culture carried over 8 weeks amount to 12.7 per cent, but the number of colonies is small in all cases, snggesting a fall in the power of subsistence. However, the virulency is not necessarily diminished. Judging from the fact that with an exeption of the bacilli in the cavity expelling bacilli np to the operation, the potencyof the virulency and the power of subsistence in the majority of badillido not coincide with one another and that the virulency grows less potent along with the encapsulation and hardeniog of the foci, it is not difficult to understand that most of tubercle bacilli found in foci undergo physiological and bacteriological changes to a greaert extent than those tubercle bacilli usually found in the sputum of patient. Furthermore, in two caseous foci whose diamater is under 0.5 cm and whose contents are caseously hardened, the bacilli, though weak in viruleucy, have beeu found to be positive to the estimation culture. It is not reasonably impossible to say that such characteristic trait of tubercle bacilli may recover the virulency, and from such a reasoning there remains a problem to be solved before deciding definitively that the bacilli proving to be smear-positive and culture-negative are all dead.