136 MILK AND THE PUBLIC HEALTH chap. 



Carried out properly the tuberculin test is extremely 

 reliable, the error not being more than about 3 per cent. 

 The test, however, requires to be carried out by a veterinary 

 surgeon,^ while there are several sources of error. 



Sources of Error. — 1. Cattle may be tested in an unsuit- 

 able condition. The cows, for example, should not be tested 

 directly after such general disturbances as removal to fresh 

 sheds, travelling, exposure to hot sun, or marked changes of 

 food. The test should not be applied within a few days 

 before or after calving, nor while in "heat." 



2. Defective tuberculin may be used. 



3. Animals suffering from advanced tuberculosis may fail 

 to react. This error is readily obviated by clinical examina- 

 tion. 



4. The animal may have been previously treated with 

 tuberculin. This is usually done fraudulently. The first 

 injection must be recent to be a source of error, as after 

 a month from the first injection it is unusual to find the 

 animal not reacting. 



5. The animal may be infected with tubercle bacilli, but 

 owing to insufficient time no lesions may have developed. 

 Such an animal will fail to give a reaction, but will sub- 

 sequently do so. 



6. The lesions may be very slight and difficult to find, so 

 that a reacting animal post-mortem may, on insufficient ex- 

 amination, fail to show signs of tuberculosis. 



1 A feature of the Wisconsin, U.S.A., system for the control of bovine 

 tuberculosis is that many of the tuberculin tests are performed by the farmers 

 themselves. The farmers must, however, have undergone training and in- 

 struction in the work at the Agi-icultural College. Also they are not allowed 

 to make the diagnosis, only to carry out the inoculations and record the tempera- 

 tures on a chart. The actual diagnosis is made by the authorities at the 

 Wisconsin Experiment Station from the records received from the farmers. 



