Some observations on the necessity for serological testing of rodent sera for Pasteurella pestis antibody in a plague control programme.