A pale creamy yellow substance, apparently similar to material observed by von Eckermann (1936) in cordierite-bearing rocks in Sweden and by Mathias (1952) in similar rocks of South Africa, has been observed in drill core from the Geco mine, Manitouwadge, Ontario.Detailed electron microprobe analysis shows the substance to contain 12 to 15 percent total H20 but otherwise to have a composition reminiscent of cordierite. Limited X-ray diffraction results, together with the compositional data, suggest that the material is a close approximation of an iron-rich sudoite, an aluminous, Mg-bearing, substantially dioctahedral chlorite known since 1964 as a common component of Kuroko ore environments.Contrary to the assumption by the earlier investigators that the yellow material was an alteration product of cordierite, it is suggested here that, at least in the case of Geco, it represents an earlier formed substance from which cordierite has arisen. It is proposed that a sudoite-like material is formed by hydrothermal sedimentation-alteration-diagenesis during and closely following exhalative ore formation. As such, it is an accompaniment and result of the ore-forming process and is localized in the ore environment. With the onset and progress of metamorphism the aluminous chlorite is converted to cordierite, which in consequence is also localized in the ore environment as observed at Noranda and many other exhalative ore localities.Hence it is proposed that at least some cordierite may be derived directly and essentially isochemically from a sudoitic, substantially dioctahedral, chlorite precursor. As such, this cordierite is a manifestation and, later, an indicator, of paleohydrothermal activity, and hence is a potential guide to exhalative ore.