I HAVE been investigating the Raman effect in a number of crystals, using the same method of excitation which proved to be particularly successful with gases.1 The primary source consists of a powerful water-cooled mercury arc, which gives an extremely intense and sharp resonance line λ2537. The advantages of this method are: (1) the ratio of the intensity of the line λ2537 to the other mercury lines in the ultra-violet is so large that in a spectrogram of moderate exposure all the Raman lines are excited by the former radiation ; (2) a short exposure is required (10 minutes to three hours) ; (3) a wide range of frequency shifts is available (about 20,000 cm.–1) ; (4) the exciting radiation can be easily absorbed by a filter of mercury vapour between the substance and the photographic plate, thus allowing the observation of modified frequencies in the immediate neighbourhood of the primary line.