Depth profiles of chemical impurities found in snow and ice have usually been interpreted simply in terms of temporal changes in atmospheric concentrations of the species of interest. Measurements of pH and conductivity from individual snow stratigraphic layers sampled above 2,600 metres indicate local spatial variation in these parameters is sufficiently large to potentially mask temporal changes inferred from depth profiles. Inferences concerning past climatic variation based on assumed temporal chemical variations for a given depth profile are suspect, prior to adequate measurements of a given analyte's spatial variability in the immediate area of sampling. A method is presented for calculating the number of samples of a given stratigraphic layer necessary to achieve the required precision about the mean pH in a vertical ice profile.