[Plate 26]The isolation of naturally occurring forms of oestrogenic and male hormones was soon followed by the demonstration that similar activity might be possessed, to a greater or lesser degree, by related compounds made artificially. When progesterone was isolated from the corpus luteum it was assumed that there would be an analogous lack of specificity, and investigation was immediately begun on the effect of slight changes in chemical constitution on the power to evoke progestational prolifera tion. In the course of this work, Butenandt and his co-workers {see the review by Westphal, 1935) examined a large number of compounds closely related to progesterone (pregnene-3: 20-dione), but failed to find progesterone-like activity in any. Among the more interesting of these inactive compounds were " dihydroprogesterone " (pregnen-20-ol-3-one), inactive in a 2 • 7 mg. dose; pregnenolone (3-ol-20-one), inactive in 25 mg.; pregnanedione, inactive in 50 m g.; pregnanediol, inactive in 50 mg.; pregnan-20-ol-3-one, inactive in 9 mg.; and androstenedione, inactive in 30 mg. With the possible exception of dihydroprogesterone, which was tested in only a small dose, this work seemed to establish the in activity of these very closely related compounds, and by implication indicated the absolute specificity of progesterone.A general investigation of the biological properties of compounds of the androsterone-testosterone series carried out in this laboratory during 1935-36 led to the examination of certain of the compounds, especially those methylated or ethylated in position 17, for progesterone-like activity {see Ruzicka, 1936, for discussion of the chemical relationships). To our surprise, in view of Butenandt's results, certain of these compounds proved to be active (Klein and Parkes, 1936), and fuller investigation has shown that seven of them exert the specific activity previously thought to be restricted to progesterone.