The solution should be autoclaved before use and the skin at the site of injection painted with iodine. Dudgeon has called attention to the constant production of oedema and necrosis in the area of the injection. This tissue necrosis occurs immediately and persists for a long time. If the injection is made in the neighborhood of an important nerve, neuritis may ensue. Repeated injections should not be given in the same area. Of course in the use of quinine salts through the medium of the hypodermic needle everything must be sterile. _Intravenous Injections._—Bass and many others think that when quinine cannot be administered by mouth it should be given intravenously. Not only is there the objection of inflammatory reactions or necrosis when the drug is given subcutaneously or intramuscularly but the absorption of the drug is so slow that the patient may die before we obtain the desired effect. Ross condemns the subcutaneous method and recognizes the advantages of the intravenous method over the intramuscular one when rapidity of action is desirable. In giving quinine intravenously Bass thinks that 10 grains at one time is sufficient and that a 20-grain dose is not without danger. He does not think it necessary to give more than 30 grains daily in this way. Intravenous quinine seems to be entirely eliminated within twenty-four hours and most of it within twelve hours. When used in cerebral malaria he repeats the 10 grains intravenously in eight hours if the drug cannot then be given by mouth. Bass thinks that theoretically amyl nitrite might relax the cerebral capillaries which are obstructed by parasite-infected red cells and thus enable the quinine in the circulation to reach such cells. The best known method of administering quinine intravenously is that of Bacelli. In this method 1 gram (15 grains) of a soluble salt of quinine is given in 10 cc. of water. MacGilchrist has shown experimentally that such a strength of quinine (1-10) will coagulate blood serum. In my opinion this is a dangerous method if the injection is made rapidly. There is no doubt as to the necessity for using the intravenous channel in cerebral or algid types of perniciousness when intramuscular injections do not give results. The generally accepted method is to use a salvarsan technique with a dilute solution of quinine, giving 1 gram (15 grains) of some soluble salt of quinine in 250 cc. salt solution. Such injections should be given cautiously. Quinine hydrochloride, which is soluble in 40 parts of water, is the salt usually recommended. MacGilchrist considers the very soluble acid salts as haemolytic and prefers to give quinine base—3 pints of a solution of the alkaloid, containing about 12 grains.