SUMMARY : Variation in the Actinomyces coelicolor species-group comprises loss of pigment and aerial mycelium, and occasionally of agar liquefaction. Stable variants may arise from degenerate, aged, vegetative mycelium, but do not normally do so when the vegetative mycelium is kept in vigorous condition by frequent subcultivation in suitable media. Single spore isolations from the aerial mycelium of typical and of variant colonies show that there may be inherent differences in the sisterspores of the same chain. Thus, in an agar-liquefying strain 3 of 15 spores had lost the power to produce pigment and to liquefy agar; and an atypical colony of the same strain yielded three viable isolates each with a strong tendency towards sectoring, with the ultimate production of a colourless, non-agar-liquefying variant as well as the typical growth. A non-agar-liquefying strain, which by prolonged cultivation in the vegetative phase had lost its power of producing the red-blue indicator pigment, yielded a variant giving rise to sectored colonies with occasional restoration of the blue pigment. Spontaneous occurrence of variants may be detected in certain spores of the aerial mycelium of a well-grown typical colony, although it is more readily seen in the spores of degenerate colonies which have been rendered atypical by artificial methods of cultivation.Much of the confusion existing in the literature concerning the extreme variability of the actinomycetes is due to lack of recognition, or forgetfulness, of their peculiar habits of growth. The ordinary actinomycete colony (Cohnistreptothriz Group I Orskov, Streptomyces Waksman) is not a colony in the bacterial sense; i.e. it is not an assemblage of disparate cells issuing from one or more similar cells, but the filamentous ramifying extension of the parent cell or cells. Moreover, when mature, it is composed of two phases: the vegetative substratum mycelium, and the aerial sporogenous mycelium. Recent work by von Plotho (1940)~ Klieneberger-Nobel (1947), and Erikson (1947a), has confirmed the existence of constitutional differences ' between these phases, already suggested by the studies of 0rskov (1923). It follows, therefore, that in a discussion of variation, such as the frequent phenomenon of asporogenous sectors produced within a single colony, it is essential to consider the origin of the inoculum in every instance. With very few exceptions, this has not been done by students of this particular problem.Actinomyces coelicolor was selected as the test organism for this study for the following reasons. (1) Like many other saprophytic actinomycetes, it readily produces asporogenous sectors or discrete sporeless colonies on certain media ; (2) it is strikingly pigmented, the range of colours varying with the pH achieved by the growth of the organism (Oxford, 1946; Cochrane & Conn, 1947) and therefore [yielding a partial index of metabolic activity; ( 8 ) some