When recently examining a series of igneous rocks for the Geological Survey of the United Kingdom, I required a ready method for the determination of potassium in the felspars, whether they occurred as porphyritic crystals or as microlites in the groundmass. The ordinary flame-reaction has always been recognized as unsatisfactory in the presence of sodium, and the use of blue glass has been long recommended, of a sufficient thickness to cut off a sodium flame, the potassium flame then coming through alone.