170 TUBERCULOSIS 
invariably, in the morning and the highest in the afternoon or 
evening. 
In view of these normal temperature variations, which often exceed 
the thermal tuberculin reaction, it is obvious that before applying the 
test the normal temperature of the animals should be approximately 
determined and that when they are being subjected to the test they 
should be cautiously protected, otherwise the comparatively slight 
elevation necessary to detect the disease may be disguised. 
As the reaction seems to be the result of an affinity existing between 
the tuberculin and the products of the tissues stimulated by tubercle 
bacteria, it is natural to suppose that when the two are brought 
together in the same animal the reaction would invariably take place. 
Experience has shown that it almost always does. It is important to 
understand, however, that under certain conditions a reaction will not 
occur. Tuberculin fails to give a reaction or to indicate the pressure 
‘of infection under the following conditions: 
“Where the tuberculin itself has not been properly made. 
“Where the temperatures are not taken long enough after the injec- 
tion of tuberculin to detect the late reactions. 
“When infected animals are tested during the period of incubation 
before lesions have developed. 
“When the lesions are arrested, encapsulated or healed or when the 
lesions are very extensive, a reaction may not follow the injection of 
tuberculin.” 
Ophthalmic use of tuberculin. Wolff-Eisner and Calmette applied 
tuberculin to the conjunctiva for the purpose of detecting tuberculosis. 
The method consists in applying a few drops of tuberculin to the 
conjunctival sac. The reaction usually begins in cattle in from 5 to 
10 hours. The first evidence is watering of the eyes, reddening of the 
conjunctiva and edematous swellings. This is followed by a 
purulent exudate which tends to accumulate, usually in the inner 
canthus, and finally to drop from the eye. It may dry forming a 
crust which eventually drops off. The reaction continues for some 
time (1 to 2 days or longer). The severity of the conjunctivitis is no 
indication of the extent of the disease. It is reported that the sub- 
cutaneous injection of tuberculin either previously or at the same time 
does not affect the eye reaction. The eye test does not give rise to 
symptoms other than those on the conjunctiva. 
This method has not given uniformly satisfactory results. It was 
found by Foth to give a reaction in tuberculous animals that had 
