The relationship between the maximum luminosity (magnitude) of a nova and its rate of decline (MMRD) is the usual starting point for deriving extragalactic distances. That luminous novae decay more rapidly than intrinsically faint novae was first noted by Zwicky (1936). The MMRD correlation for Galactic novae was confirmed and studied in more detail by McLaughlin (1945), Schmidt (1957), and Pfau (1976). The most complete recent study of the MMRD relation for Galactic novae is by Cohen and Rosenthal (1983) and Cohen (1985), who have derived expansion parallaxes for a number of (previously undetected) spatially resolved shells around old Galactic novae. The physical basis for the MMRD relation has been discussed by Hartwick and Hutchings (1978) and Shara (1981a, b), and is reviewed by Shara (1989). These authors show that more massive white dwarfs require less accreted matter to produce a thermonuclear runaway, and these lower mass envelopes can be ejected more rapidly; thus the most luminous novae are also the fastest.
To measure the distance to an external galaxy using the MMRD relation, it is necessary to determine the apparent magnitudes of novae at maximum light, and a mean rate of decline over two magnitudes. At present the calibration of the MMRD relation is in the B or mpg bands (see below), so that observations through these bandpasses are preferred. It is essential that the observations sample the light curves of novae frequently enough near maximum light that mmax can be estimated for the fastest (i.e., brightest) novae. The signal to noise ratio of the photometry should be high enough that novae discovered near maximum light can be followed at least 2 magnitudes below this level.