Some comments on high precision study of neutrino oscillations

I discuss here some problems connected with the high precision study of neutrino oscillations. In the general case of n -neutrino mixing I derive a convenient expression for transition probability in which only independent terms (and mass-squared differences) enter. For three-neutrino mixing I discuss a problem of a definition of a large (atmospheric) neutrino mass-squared difference. I comment also possibilities to reveal the character of neutrino mass spectrum in future reactor neutrino experiments.