Wide variations in the quantity of the drug used and also in the frequency of administration to prevent withdrawal symptoms are encountered among different human addicts as well as in the same addict at various periods during addiction. This variation is found not only in dosage and frequency of administration, but also in the methods of administration. The choice as to the latter in 95 per cent of our cases was the hypodermic method, but sniffing of heroin, oral administration, pipe smoking of opium and intravenous administration were also encountered. With these variations in mind we administered additional amounts, both intramuscularly and intravenously, in order to determine, first, what physiologic effects are brought about in human addicts by quantities greater than those required to prevent withdrawal symptoms; second, whether abrupt withdrawal for a period of forty-eight hours decreases the intravenous tolerance of the addict, and third whether these additional amounts