8 
In his first paper Rosenberger reports positive results when 2 or 
2.5 c. c. of the sediment were inoculated in two instances into guinea 
pigs. 
Various observers since the time of Yillemin have reported the 
finding of tubercle bacilli in the blood in tuberculosis. 
Yillemin ° in 1868 reported the successful inoculation of a rabbit 
with tuberculosis b} T the blood collected from the femoral vein of a 
tuberculous rabbit; also, of a rabbit- with the blood from a case of 
human tuberculosis. 
I shall not attempt to give a review of the subject of tubercle 
bacilli in the blood, as the literature upon this subject up to 1904 
has been well discussed by Bergeron b in his admirable thesis on 
the subject. In his paper Bergeron gives a fine critical review of 
previous work and in addition experiments by himself. In his con- 
clusion he sums up the status of the subject so well that I shall 
quote the following from him: 
1. The pathological anatomy shows that at a given moment the 
tubercle bacillus is able to pass into the general circulation, but the 
lesions capable of being interpreted as being the result of an infection 
in the blood are only to be encountered in acute tuberculosis. 
2. Guinea pigs experimentally infected never present bacillemia 
either at the beginning of the infection or during the period of 
cachexia. 
3. Rabbits experimentally infected usually present a kind of lesion 
which testifies to the passage of the bacilli into the blood at a given 
moment; but these bacillary discharges should be considered essen- 
tially transitory and ephemeral. In short, in spite of experiments to 
the contrary, it seems impossible to demonstrate the presence of the 
bacilli in the blood by means of the inoculation of guinea pigs as well 
in the beginning of the disease as later, when the infection has been 
made by subcutaneous or intraperitoneal injections. 
In cases where the infection has been made intravenously the 
bacilli are localized more or less rapidly in the tissues and disappear 
from the circulation in a time which may not exceed four hours 
(Xocard), but may be as late as six days (Gaertner) or even later, 
though not much later. 
4. The presence of tubercle bacilli in the blood of persons with 
miliary tuberculosis is a fact proved long ago, both by the patholog- 
ical anatomy and autopsy and by examination of the blood during 
life. Researches, however, tend to show that the bacillemia of acute 
tuberculosis is a relatively rare phenomenon. 
o Yillemin: Etudes sur la tuberculose. Paris, 1868. 
b Bergeron. Andre: Etude critique sur la presence du bacille de Koch dans le sang. 
These, Paris, 1904. 
