Evidence is presented to show that there is a complete variation in composition from augite to pigeonite passing through subcalcic augite. Under ordinary conditions of crystallization of magmas, subcalcic augite does not form; an immiscibility gap exists between augite and pigeonite. However, subcalcic augite forms only when part of Si4+ in Si—O chain of pyroxene is replaced by Fe3+, rendering it possible for Ca2+, Mg2+, and Fe2+ to enter the pyroxene structure in all proportions. The Si4+—Fe3+ substitution is probably favored by rapid cooling at comparatively high temperature. The course of crystallization of clinopyroxenes in the Izu-Hakone province, Japan, is from salite close to diopside, passing through augite, to ferropigeonite. Throughout the greater part of this course, Ca2+ of the salite is successively replaced by Fe2+, while the proportion of Mg2+ remains nearly constant.