78 BACTERIOLOGICAL DIAGNOSIS. 
the sputum of a person who had presented no sym- 
ptoms of the disease for eight years and was apparently 
cured. Buta person in whom the bacilli are present is 
always in danger of a recrudescence of the disease, and 
may be a source of infection. Absence of the bacilli does 
not disprove the diagnosis of tuberculosis; bacilli do 
not appear in the sputum until the lung-tissue in which 
they occur breaks down, and are therefore absent in the 
early stages of acute tuberculosis. 
In some cases of ordinary chronic phthisis bacilli may 
occur in the sputum in very scanty numbers and may be 
missed unless a very careful search is made. Bacilli 
should not be considered as being absent until well 
stained films have been examined for at least half an 
hour, and the examination repeated on several occa- 
sions. 
The finding of tubercle bacilli in the urine is prac- 
tically absolute proof of tuberculosis of some part of 
the urinary tract, probably the kidneys or bladder. 
Absence of bacilli implies nothing unless the exam- 
ination has been made very thoroughly and repeated 
several times at intervals. Then it affords presumptive 
evidence that the urinary passages are free from the 
disease. 
The same is true of the examination of pus. Tubercle 
bacilli rarely occur in inflammatory exudates except in 
very small numbers and can only be demonstrated by 
animal experiments. If you examine pus from a 
chronic abscess and find no organisms of any kind it is 
almost certain that the process is a tubercular one; and 
the negative evidence obtained by the failure to find 
tubercle bacilli should not be allowed to carry much 
weight. The same is true for the clear exudates. 
