warrants further search for evidence of the infection in this country. In the serological diagnosis, however, difficulties may arise through repeated failure even late in the disease to demonstrate agglutinins in the sera of individuals proved to be infected by the isolation of the brucella (Burnet, 1924; Carpenter, Boak and Chapman, 1929), and through the agglutinative effects frequently shown by the sera of individuals who give neither clinical evidence nor history of undulant infection (Sedg-