We show that using mole fractions x(s) instead of activities a(s) in the Gibbs equation for analyzing surface tension curves for onset of amphiphilic association may be open for discussion. First, we determined the activity coefficients for the simple binary water-ethanol and water-n-propanol systems from vapor pressure measurements. The activity coefficients are neither unity nor are they constant in the range where the surface tensions sigma of the binary systems decrease from the water value down to the value of the pure alcohols. Second, a break in the sigma vs In x(s) curves disappears when proper sigma vs In a(s) curves are constructed. The implication for determining critical micelle concentrations and headgroup areas of surfactants from surface tension curves is addressed.