Two very strong visible band systems in the orange (6120 Å) and green (5480 Å) regions are emitted when Ca compounds are added to flames and electric arcs. The origin of these bands is not yet understood, and there is considerable doubt especially in the case of the green bands, whether it is due to CaO, or, as suggested by experiments on the isotope shifts when H 2 is replaced by D 2, due to CaOH. Since the bands have mainly been observed in flames and arcs, which contain large amounts of H 2O and H 2, it has not been possible to decide about the nature of the emitting molecule. In the present investigation the green and orange bands were excited through a short-duration high-tension spark between electrodes of pure Ca in an atmosphere of O 2. It was very difficult to remove H 2O completely from the walls of the vessel and the electrodes, and observation of the H a line showed that the residual hydrogen could not be kept below about 270 ppm XXX 2.7 × 10 14cm -3. The particle density of Ca and Ca + in the discharge gap, however, is much greater than the hydrogen impurity, being about 10 16cm -3. Time-resolved experiments show that the intensity ratio of the green and orange bands remains almost unchanged over a wide temperature range and that both these intensities are of the same order of magnitude. Application of the normalized temperature concept leads to the conclusion that hydrogen cannot be a constituent of the emitting species responsible for the bands, because CaO and CaOH should show very different emission functions.