On the possibility of deciding between Pauling's and Griffith's models of oxyhemoglobin by means of Mössbauer spectroscopy