Abstract
A posteriori justification of the statistical theory of nuclear fission is found in the shell effect on mass distribution and kinetic-energy distribution as well as in the constancy of kinetic energy at high-energy fission. A priori justification is found in the calculation showing that the scission time is longer than the characteristic nuclear time and the nuclear relaxation time. It is shown that the condition for the adiabatic approximation to be valid cannot be satisfied in the fission process from the saddle point to scission and therefore the validity of the adiabatic theory of fission is in doubt.