In February, 1909, Rosenberger1 published his announcement of the finding of tubercle bacilli in the blood of every one of 125 tuberculous patients examined. Previous to that time the prevailing opinion had been that only occasionally in cases of so-called localized tuberculosis2 did bacilli escape into the blood-stream. It had been supposed that in the miliary form alone was the number of organisms sufficient to give any promise of detection, and that even in these cases the search was apt to be futile. The description by Landouzy3 of typho-bacillose, a stage of tuberculosis which precedes the formation of localized lesions and which is supposed by him to be a septicemia, had added interest to this subject.
After unsuccessful attempts to confirm Rosenberger's findings in a search of the blood of human beings, I undertook, in conjunction with C. M. Haring, D.V.M., Assistant Professor of Veterinary