The quantitative study of the interstellar medium (ISM) began with Hartman’s detection in 1904 of the resonance lines of Ca and K, seen in absorption against bright stars by intervening diffuse clouds. Other atoms (Na,Fe,Ti) were subsequently detected, although the overwhelming abundance of H was not recognized until the 1930s. In 1934 came the first discovery that was to suggest the possibility of interstellar molecules—the detection of four diffuse interstellar bands in the visible. Although these bands (now numbering three dozen) have never been identified, Russell in 1935 suggested that they were molecular in nature. In 1937 Swings proposed that the diffuse bands were C02 bands, and the same year Swings and Rosenfeld stressed the possibility of interstellar CH, OH, NH, CN, and C2. In the years 1937 to 1941, the first discoveries of interstellar molecules were made at Mt. Wilson, of CH+, CH, and CN by means of their optical absorption spectra in the 4230-4300-A region (e.g., Adams 1941). However, the possibility of a rich chemistry in the ISM, as suggested by such detections, remained unrecognized at the time. Indeed, despite an extensive study by Adams (1949) of the optical spectra of these molecules towards many stars, there remained doubt as to whether these lines were interstellar or circumstellar. The 1951 work by Bates and Spitzer which concluded that the formation of even simple (diatomic) molecules in space was slow and that destruction processes by UV radiation were rapid further cast doubt on the notion of interstellar molecules, although the only formation mechanism considered was radiative association.