The tmsubtracted dispersion relation is not compatible with the charge-exchange data, if a power law is assumed for the total cross section difference at high energies and the exponent is taken from the fit to the Serpuchov data. However, there is no difficulty for the subtracted dispersion relation. Recent estimates of a finite high energy limit of the total cross section difference and a possible substantial curvature of the p-trajectory near t= 0 are discussed.