A Comparative analysis has been carried out between thernexperimental excess heat capacity of solid solutions of"N2 andrn''N2 in Ar and Kr due to rotation of the nitrogen molecules, andrnthe existing theoretical models in the temperature range 0.5 torn6 Kelvin. It is shown that the well known Devonshire-ManzMirskyrn(DMM) model can not explain the observed heat capacityrnbelow 2K neither on the basis of the assumption of frozen hightemperaturernnor equlibrium concentration of different spinrnnuclear modifications of nitrogen. Splitting of the lowestrnrotational energy levels of the impurinty molecules isrnproposed, so as to provide an agreement between thernexperimental observations and theoretical data. It is alsornconcluded that experimental observations favour the frozenrnhigh-temperature composition of the different spin species ofrnnitrogen down to 0.5K in temperature, and the equilibrium'rnorientation of the molecules is along the directions ofrnthe crystal