Those who have made separations of heavy minerals from sands or crushed rocks will doubtless have realized that concentrates obtained by panning are not identical with those obtained when heavy liquids (usually bromoform) are used, for certain of the heavy minerals, either because of their shape or somewhat lower densities, tend to be washed away to a greater extent than zircon, rutile, etc. Yet, on account of the high cost of bromoform, panning is frequently adopted; and a compromise method of panning off most of the light material and completing the separation by means of bromoform is also used. As there is no record in geological literature (so far as I can trace) showing precisely to what extent the composition of the residue depends upon the method adopted, the following account may be of interest.