Recently Tatum and Cooper1reinvestigated a drug (metaaminoparahydroxyphenylarsine oxide) which they named mapharsen. This drug had originally been studied by Ehrlich and Hata,2who expressed the belief that it was too toxic for use in the treatment of human syphilis. Tatum and Cooper, however, found that the "sterilizing therapeutic index" in rabbit syphilis was higher with mapharsen than with neoarsphenamine (1.66 compared with 1.11) and that doses of a given quantity of mapharsen produced results equal to those obtained by from thirty to fifty times as much neoarsphenamine. Raiziss and Severac3later disagreed with Tatum and Cooper, stating that the "therapeutic index" given by the latter authors for neoarsphenamine in rabbit syphilis was too low. Raiziss and Severac pointed out that most investigators found neoarsphenamine to have a therapeutic index of about 5. They studied the effects of mapharsen and neoarsphenamine in rabbit syphilis and determined a