In 1916, Herbert Koch1 described what he called initial fever in tuberculosis. He had in mind a brief period of fever at the particular time that allergy develops after tuberculous infection has taken place. Since then, several authors have considered this conception and have either supported2 or been inclined to doubt3 the existence of initial fever. The observations concerning this question are as yet too few to justify a definite opinion. One can rarely carry out systematic tests with tuberculin in children in this stage of tuberculosis and prove that a rise of temperature, as allergy develops, does not depend on causes other than tuberculous infection. Positive cases in which negative tuberculin reactions have been obtained immediately before the fever, in which the cause of the fever is not demonstrable and in which the tuberculin reaction becomes positive during the fever are extremely few. I regard as