272 TUBERCULOSIS. 
Although containing no bacilli, tuberculin does contain the 
toxins which the bacilli produce, and these toxins, if inoculated 
into an animal in sufficient quantity, would poison it. When 
injected in small quantity, the material has no effect upon the 
healthy individual; but if the individual is already affected with the 
disease, this inoculation produces a marked rise in temperature, 
which soon disappears. 
The fact that the injection is followed by a rise in temperature 
makes it possible for this material to be used among cattle in detect- 
ing tuberculosis. Healthy animals fail to respond to this inocula- 
tion and are wholly uninjured by it. The farmer may, therefore, 
have his herd tested with the confidence that his healthy animals will 
not suffer by the test. On the other hand, the animals that have 
become infected with the disease will show a rise in temperature, 
and the test will thus make it possible to separate the affected 
animals from those that are yet in health. 
The accuracy of the test has been the subject of much dispute. 
It has been found subject to some error. If animals are tested 
under abnormal conditions, as, for example, when in new barns, or if 
taken from a cattle car and tested at once, even healthy animals may 
respond. But when the animals are in normal conditions the 
healthy animals probably never respond, or at all events so rarely 
as not to interfere with the accuracy of the test. Secondly, some 
animals very far advanced in the disease fail to respond. These 
cases are of little importance since they are commonly detected 
readily by clinical symptoms. Thirdly, all animals which are 
moderately attacked, and all of the very incipient cases of tuberculo- 
sis, are detected by the tuberculin. Even a single minute tubercu- 
lous gland is sufficient to cause a positive reaction to the test. 
This last fact forms at once the strength and the weakness of the 
tuberculin test. Tuberculin does pick out with great accuracy all 
mild cases, and clinical symptoms will pick out the rest. But this 
test fails to distinguish between severe and mild forms, putting in 
one class the animal that may have a small tuberculous gland, 
which may heal in a short time, and the animal with a severe case of 
intestinal tuberculosis which is scattering bacilli to the great danger 
