4.2. Mass-to-luminosity Ratios and Models of Physical Evolution of Stellar Populations
The method was applied to the Andromeda galaxy (Einasto 1969b, 1970a, Einasto & Rümmel 1970a), and to our Galaxy (Einasto 1970b). In the case of the Andromeda galaxy the mass distribution model found from the rotational data did not agree with the data on physical properties of populations. If we accepted the rotational velocities, based mostly on radio observations (Roberts 1966), then the mass-to-luminosity ratio, M / L, of central stellar populations became very low, of the order of 1 in Solar units. On the other hand, the spectral data (Spinrad 1966) suggested a much higher value, M / L 17.
The next problem was to find internally constitent values of physical parameters of stellar populations of different age and composition. For this purpose I developed a model of physical evolution of stellar populations (Einasto 1971). When I started the modelling of physical evolution of galaxies I was not aware of similar work by Beatrice Tinsley (1968). When my work was almost finished I had the opportunity to read the PhD thesis by Beatrice. Both studies were rather similar, in some aspects my model was a bit more accurate (evolution was calculated as a continuous function of time whereas Beatrice found it for steps of 1 Gyr, also some initial parameters were different). Both models used the evolutionary tracks of stars of various composition (metallicity) and age, and the star formation rate by Salpeter (1955). I accepted a low-mass limit of star formation, M0 0.03 Msun, whereas Beatrice used a much lower mass limit to get higher mass-to-luminosity ratio for elliptical galaxies. My model yields a continuous sequence of population parameters (colour, spectral energy distribution, M / L) as a function of age. The calculated parameters of stellar populations were compared with observational data by Einasto & Kaasik (1973). The available data supported relatively high values (M / L 10 - 30) for old metal-rich stellar populations near centres of galaxies; moderate values (M / L 3 - 10) for disks and bulges; and low values (M / L 1 - 3) for metal-poor halo-type populations. Modern data yield slightly lower values, due to more accurate measurements of velocity dispersions in central regions of galaxies, and more accurate input data for models.
These calculations suggest that the rotation data by Roberts (1966) are biased. To find the reason for this biasing, I analysed the velocity field obtained from the radio observations. My analysis suggested that low rotational velocities in the central regions are due to a low spatial resolution of the radio beam (Einasto & Rümmel 1970b, c). The corrected velocity field was in agreement with a higher value of M / L in the central regions of M31, suggested by direct spectral data and models of physical evolution.