4.7. Further Implications of Dark Matter

Variations in M / L ratios for individual galaxies, in addition to being relevant in the overall census of cosmologically important material, also effects the determination of the extragalactic distance scale as well as the ability to measure large scale flows. If dark matter dominates the potential wells of galaxies there are two intriguing possibilities that would cause small calibration errors in the derivation of relative (and absolute) distances that would produce a false peculiar velocity signal and an erroneous value for H₀.

1. As remarked earlier, the dark matter constituent of galaxies is related to their chemical evolution if the dark matter content is stellar remnant dominated. The luminosities of stars at different wavelengths are strongly dependent on chemical composition as atmospheric opacity often depends on the free electron density. Metals are the primary donors of free electrons. As metals are synthesized from massive stars that eventually become the dominant remnant source, then metal-rich galaxies may have more remnant mass per unit luminosity than metal-poor galaxies. This would lead to slightly higher M / L values in their luminous parts which affects the calibration of both the TF and D_n- relations. There is weak evidence that the residuals from the D_n- relation correlate with metal line strength (see Gregg 1995). Curvature in the H-band TF relation, first noticed by Aaronson et al. (1982) and more fully discussed by Han et al. (1989), may also result from the stars in low mass metal-poor galaxies having lower H-Band luminosity per unit mass (see Bothun et al. 1984).

2. The more intriguing possibility, first suggested by Silk (1989) and Doroskevich and Klypin (1988), is that there are small amplitude but large scale variations in the dark matter content of the Universe. Lets imagine that whatever the dark matter is, it is laid down in waves of wavelength 50-100 kpc with peak to valley amplitude of 10%. Doroskevich and Klypin show that such a distribution is very consistent with the observed large scale flows. If we can imagine sprinkling galaxies (or the seeds of galaxy formation) down on this network of dark matter waves, then some galaxies will be located at peaks, while other galaxies will be located in valleys. Hence, the amount of dark matter mixed in which galaxies as they formed will have ± 10% rms variations. This means that systematic errors in the calibration of relative distances based on the TF and D_n- relations could occur when sampling along different lines of sight.

Trying to measure small differences in M / L is difficult at best so there is no direct evidence for either of these phenomena. However, even systematic differences of 10% are important. Given that no cosmological structure formation scenario can account for the LP result (e.g., Strauss et al. 1995), one needs to seriously ask if the flows are peculiar or the galaxies are peculiar. Since we do not understand the process of galaxy formation or the distribution of dark matter prior to galaxy formation, there may well be slight differences between how much dark matter is mixed in to Local Group galaxies, compared to galaxies 50 to 100 Mpc away. This would produce a zeropoint shift in the calibration of the distance indicators which in turn produces a false peculiar velocity signal due to systematic errors in the determination of relative distance.