We present probability distribution functions (PDFs) of the surface densities of ionized and neutral gas in the nearby spiral galaxies M31 and M51, as well as of dust emission and extinction Av in M31. The PDFs are close to lognormal and those for HI
and Av in M31 are nearly identical. However, the PDFs for H2 are wider than the HI PDFs and the M51 PDFs have larger dispersions than those for M31. We use a simple model to determine how the PDFs are changed by variations in the line-of-sight (LOS) pathlength L through the gas, telescope resolution and the volume filling factor of the gas, f_v. In each of these cases the dispersion sigma of the lognormal PDF depends on the variable with a negative power law. We also derive PDFs of mean LOS volume densities of gas components in M31 and M51. Combining these with the volume density PDFs for different components of the ISM in the Milky Way (MW), we find that sigma decreases with increasing length L with an exponent of -0.76 +/- 0.06, which is steeper than expected. We show that the difference is due to variations in f_v. As f_v is similar in M31, M51 and the MW, the density structure in the gas in these galaxies must be similar. Finally, we demonstrate that an increase in f_v with increasing distance to the Galactic plane explains the decrease in sigma with latitude of the PDFs of emission measure and FUV emission observed for the MW.
A framework is introduced for coupling the evolution of galactic magnetic fields sustained by the mean-field dynamo with the formation and evolution of galaxies in cold dark matter cosmology. Estimates of the steady-state strength of the large-scale
and turbulent magnetic fields from mean-field and fluctuation dynamo models are used together with galaxy properties predicted by semi-analytic models of galaxy formation for a population of spiral galaxies. We find that the field strength is mostly controlled by the evolving gas content of the galaxies. Thus, because of the differences in the implementation of the star formation law, feedback from supernovae and ram-pressure stripping, each of the galaxy formation models considered predicts a distribution of field strengths with unique features. The most prominent of them is the difference in typical magnetic fields strengths obtained for the satellite and central galaxy populations as well as the typical strength of the large-scale magnetic field in galaxies of different mass.
We derive the magnitude of fluctuations in total synchrotron intensity in the Milky Way and M33, from both observations and theory under various assumption about the relation between cosmic rays and interstellar magnetic fields. Given the relative ma
gnitude of the fluctuations in the Galactic magnetic field (the ratio of the rms fluctuations to the mean magnetic field strength) suggested by Faraday rotation and synchrotron polarization, the observations are inconsistent with local energy equipartition between cosmic rays and magnetic fields. Our analysis of relative synchrotron intensity fluctuations indicates that the distribution of cosmic rays is nearly uniform at the scales of the order of and exceeding $100p$, in contrast to strong fluctuations in the interstellar magnetic field at those scales. A conservative upper limit on the ratio of the the fluctuation magnitude in the cosmic ray number density to its mean value is 0.2--0.4 at scales of order 100,pc. Our results are consistent with a mild anticorrelation between cosmic-ray and magnetic energy densities at these scales, in both the Milky Way and M33. Energy equipartition between cosmic rays and magnetic fields may still hold, but at scales exceeding 1,kpc. Therefore, we suggest that equipartition estimates be applied to the observed synchrotron intensity smoothed to a linear scale of kiloparsec order (in spiral galaxies) to obtain the cosmic ray distribution and a large-scale magnetic field. Then the resulting cosmic ray distribution can be used to derive the fluctuating magnetic field strength from the data at the original resolution. The resulting random magnetic field is likely to be significantly stronger than existing estimates.
