For the standardisation of adrenalin preparations.

(a) The colorimetric method is not reliable with weak or impure solutions.

(b) The effect on the pupil of the frog's eye gives uncertain results.

(c) The determination of the minimum effective dose on the arterioles of the perfused frog is tedious and uncertain; on an average 0·1 per million produces an effect.

(d) The determination of the minimum effective dose in causing a rise on the blood pressure of the atropinised rabbit yields fairly satisfactory results.

(e) Adrenalin, suprarenalin, and hemisine all give a precisely similar result, 0·0003 mg. per kilo of body weight, or 0·006 per million, of the rabbit's blood causing a distinct rise in the blood pressure in the rabbit, and 0·012 mg. per kilo of body weight, or 0·24 per million, of the blood causing a rise in the cat.

(f) The most satisfactory method is the determination of the dose just sufficient to antagonise 0·6 mg. of nitroglycerin (Parke, Davis & Co.). Of adrenalin, 0·0075 mg. is sufficient.

I wish to express my indebtedness to Dr Noël Paton for much help in the preparation of this paper.

The expenses of this research were defrayed from a grant made to the Laboratory by Mr J. Francis Mason for investigations on the physiology of the ductless glands.