In the first paper1of this series Pugh and I reported the results of an investigation of the various magnitudes of the dose in roentgens reported in the literature as being equivalent to the erythema dose and also the variation in this dose in actual clinical use in Australia, the United States and England. In order to obtain a more complete picture of the various clinical aspects of the erythema reaction the extensive experimental program herein reported was carried out. A new criterion of the threshold erythema reaction is proposed and defined, which, it is believed, would greatly increase the degree of accuracy with which different workers could correlate their observations by avoiding the necessity for judging the degree of the reaction. A number of tests were performed before a suitable portion of the body could be chosen as a standard location for the tests, and the question of