patially resolved reflectance spectra of Mars in the 2.2- to 2.4-μm spectral region were obtained in August 1988 using the NASA 3-m Infrared Telescope Facility. The spectra show weak absorption features due to Martian atmospheric carbon monoxide and a surface mineral. Both CO and the mineral absorptions are composed of overlapping narrow features, but in many locations, such as Hellas, Chryse, Eden, and Moab, the mineral absorptions are quite strong, at least 3 times stronger than at the most absorbing wavelengths of CO near 2.33 μm. Therefore CO complicates the analysis of the surface mineral but does not always overwhelm its signature. Model removal of the Martian atmospheric CO has been performed, and the remaining absorption bands are identified as scapolite. Relatively strong absorptions that match bands in the spectrum of scapolite and have little or no CO absorption interference are seen near 2.41, 2.39, and 2.29 μm. Absorption also occurs at the scapolite bands at 2.36 and 2.33 μm, but the analysis is complicated by uncertainty in the atmospheric CO removal at these wavelengths. Weaker scapolite bands are seen at 2.44 and 2.23 μm where there is virtually no atmospheric interference. The scapolite bands observed on Mars are due to HCO3− and HSO4− ions in the scapolite structure. The bicarbonate and bisulfate contents appear to vary with location: the scapolite in Hellas is more bisulfate-rich relative to that in the Chryse/Moab/Eden area. Other locations contain little (Arabia, Syrtis Major, Hellespontica, and Isidis) or no scapolite (e.g., Margaritifer, Ausonia, and Erythraeum). The calculated abundances are unconstrained because the amounts of HCO3− and HSO4− in the Martian scapolites as well as their grain sizes are not known. If the scapolites contain about 3 wt % of each, near the maximum possible, the scapolite abundances probably range from about 5 wt % scapolite at Eden and Hellas; 3–5% at Chryse, Moab, and Oxia Palus; 2–3% at Arabia, Syrtis Major, and Isidis; to less than 2% at Hellespontica, Syrtis Minor, and Margaritifer, assuming a relatively large grain size of 50–100 μm. If the characteristic grain sizes are smaller or the HCO3− and HSO4− contents are lower, the scapolite abundances required to match the observed band depths would be higher. The mineral bands are apparent in many of the Mars spectra measured, so it appears to be widely but not uniformly distributed. The newly observed fine structure also varies greatly in both depth and spectral detail with location on Mars. Thus there appears to be regional variations in composition. The mineral phases appear to reflect local or regional geology and are not primarily contained in the homogeneous, globally redistributed aeolian dust. Higher spectral resolution Martian spectra in the 2.3-μm region as well as at 3.9 μm are needed to confirm the scapolite identification and to constrain its abundance.