The correct approach to build up the extragalactic distance scale is to use all available primary (novae, cepheids, RR Lyrae) and secondary indicators (brightest stars, globular clusters, largest HII rings) to calibrate without arbitrary extrapolation all reliable tertiary indicators (magnitudes and diameters of galaxies), precisely corrected for all known effects of type, luminosity class, orientation, internal and galactic extinction and redshift. Such data are now available for over 1000 galaxies in the Second Reference Catalogue.
Revised distances to members of the Local Group from primary indicators and new estimates of distances to the nearest groups from primary and secondary indicators are used to calibrate the tertiary indicators via a new, composite luminosity index.
The distances derived from globular clusters for 3 galaxy clusters dominated by ellipticals (Vir I, For I, Hya I) with mean corrected velocities 1000 ≤ Vo ≤ 3650 km s-1 give a mean Hubble ratio <H> = 88 (1 ± 0.15) km s-1 Mpc-1. This value rests entirely on the calibration of the globular clusters luminosity function in the Galaxy as a gaussian of dispersion σ = 1.1 mag. and mean <MB>(⊕) = -6.55, with <(B-V)o > = 0.75, based on the adopted RR Lyrae zero point <Mv> (RR) = + 0.86 ± 0.15.
The distances derived from tertiary indicators to 19 field spirals and 28 nearby groups dominated by spirals with corrected velocities Vo < 1800 km s-1 give <H> = 82 (1 ± 0.15) km s-1 Mpc-1. This value rests on the relations between luminosity index and fully corrected absolute magnitudes or linear diameters of galaxies derived without extrapolation and calibrated in ~20 nearby galaxies by seven secondary indicators. The zero points were derived from the 3 primary indicators, including 15 galactic novae with <M15> = -5.5 ± 0.15, and 13 cepheids in 8 galactic clusters with <MB> (logPo = 0.8) = -2.92 ± 0.15 (for an adopted Hyades modulus of 3.16 ± 0.05).
The low values of Ho≃50 to 55 can be explained by an accumulation of complex systematic errors arising from a multiplicity of sources. These errors, all but one acting in the same sense, have been identified and evaluated; a full report will appear elsewhere.