A newly designed and constructed Knudsen cell has been tested by measuring the reaction probability for gaseous N2O5 on 65% H2SO4/H2O at 210-230 K to be 0.075 +/- 0.047 (2 sigma), which is in excellent agreement with the literature value. This cell has been applied to the study of gaseous HNO3 reactions with NaCl small crystals and ground powders at 298 K. A rapid initial uptake of HNO3 and production of gaseous HCl are observed when the crystals and powders are pumped but not heated prior to reaction. After this rapid initial reaction, a constant uptake of HNO3 and formation of HCl is observed from which a reaction probability of (1.4 +/- 0.6) x 10(-2) (2 sigma) is calculated. When possible systematic errors (including uncertainties in the effective surface area available for reaction) are taken into account, the overall uncertainty is estimated to be about a factor of 2. The measured reaction probability is independent of the size or preparation of the salt crystals as well as the number of layers of salt in the sample holder. This reaction probability is in excellent agreement with results from the previous work of Rossi and co-workers(21,23) and Leu and co-workers(15) using powders but significantly larger than that measured by Laux et al.(1)2 using single crystals and an ultrahigh vacuum system. Prior heating of the salts while pumping decreased the extent of the initial rapid reaction but did not affect the subsequent constant reaction uptake probability. Experiments on the reaction of HNO3 with NaCl crystals that had been previously exposed to D2O to replace any surface-adsorbed H2O and on the reactions of DNO3 with NaCl show that under all experimental conditions studied here, some water remains on the surface and plays a key role in the uptake of HNO3. We propose a new model for the reaction of HNO3 with NaCl powders in which HNO3 is taken up into strongly adsorbed water (SAW) on the salt. This SAW, for which there is prior evidence in the literature,(34) appears likely to be held at defect sites on the powders. Acidification of this SAW leads to degassing of HCl due to dissolution of NaCl into the SAW from the underlying salt. As gaseous HNO3 continues to be taken up, HCl degasses and nitrate precipitates out as NaNO3. This model represents a fundamental change in the description of the heterogeneous reactions of salt powders. The lower reaction probability for single crystals observed by Laux et al.(12) is consistent with the lack of surface-adsorbed water on relatively defect-free single crystals. No uptake of the gases NO2, NO, HCl, ClNO, ClNO2, or H2O was observed on the finely ground NaCl powder from which an upper limit to the reaction probabilities for these gases with NaCl of similar to 10(-5) was derived. The atmospheric implications of this model are discussed.