Infrared (IR) photoacoustic1spectra are usually ‘‘source-compensated,'’ i.e., normalized to account for the variation in the intensity of emission of the source at various wavelengths. Such compensation or normalizing is done by comparing the single-beam spectrum of a sample with the emission spectrum of the source measured with a detector or, more commonly, with the single-beam spectrum of a ‘‘reference'’ carbon. The latter is assumed to be a flat black absorber. We showed, however, that this assumption was not generally valid because different carbons gave different spectra, and concluded that it would be better to avoid carbons and to use direct measurements of the exciting source for the compensation of spectra2. These conclusions were supported by the data of Riseman and Eyring.3They examined different carbons and found not only that the spectra differed but that changes in the interferometer scan speed would bring about further differences, and also concluded that no form of carbon was suitable for source normalization. They suggested using a detector whose response function closely matched that of an ideal photoacoustic black body absorber, i.e., in practice, a DTGS (deuterated triglycine sulfate) detector.3