A serious challenge to conventional theory on which fever therapy is based comes from the demonstration by Ellingson and Clark1 of the University of Wisconsin that hyperpyrexia causes a significant lowering of specific antibody titer in experimental animals. In 1908 Roily and Meltzer2 tested the influence of artificial fever on the resistance of rabbits to experimental bacterial infections. Groups of 2 or more rabbits were given repeated sublethal inoculations of pneumococci, staphylococci, Escherichia coli or Pseudomonas aeruginosa. One animal of each group was kept at room temperature to serve as a control and the others were confined to a heated chamber. All the control animals died of the repeated sublethal inoculations. Of the 11 febrile animals (rectal temperatures were not recorded) but 4 died. This was taken as proof that the febrile state increases antimicrobic resistance.
According to the orthodox side chain theory of the time, this increase