1.4.2 Early Debates
Measurements with single dish telescopes with (for that time) very sensitive receivers provided the first hints that rotation curves stay flat at large radii. In particular, there was a debate about the rotation curve of M31 at large radii (cf. Roberts 1975, Emerson & Baldwin 1973, and Baldwin 1975), but also about possible sidelobe effects of Arecibo data (see discussion after Salpeter 1978). Interferometer data free from this effect for 5 Scd galaxies (Rogstad & Shostak 1972) showed that rotation curves did not decline. Newer data for a number of galaxies observed with the Westerbork telescope settled these issues : a compilation of 25 rotation curves of spiral galaxies of various morphological types showed that all of them are roughly flat, or rising (cf. Bosma 1978, 1981a, b).
Numerical simulations of spiral galaxies also started in earnest in the early 1970s. Hohl (1971) found consistently that flat disks are very prone to the bar instability. A cure was devised by Ostriker & Peebles (1973), i.e. to embed a disk in a dynamically hot dark halo, which was presumed to be roughly spherical. The required halo masses interior to the disks are rather large, so that the total mass of a large galaxy at large radii can be easily 1012 M. The absence of a decline in a rotation curve implies that mass increases linearly with radius. In this way a mass radius relationship can be established, e.g. for our Galaxy, using various tracers of the mass like outlying globular clusters, satellites, Local Group timing, etc. (Ostriker et al. 1974).
The data in the late seventies from HI observations (Bosma 1978), show that extended flat rotation curves are ubiquitous for large spirals, and that only for small Sc galaxies the rotation curves are still rising. In a series of papers by Rubin et al. (1978, 1980, 1982, 1985) on H rotation curves a nice systematic behaviour as function of type and luminosity of the rotation curves for Sa, Sb and Sc spirals was established as well.