MEASURING THE HUBBLE CONSTANT AND OUR VIRGO-INFALL VELOCITY INDEPENDENTLY

A sample of spiral galaxies with B(T) < 14.5 located in two local volu mes, one in the direction of, but behind, the Virgo Cluster (behind-Vi rgo volume [BV]) and the other in the opposite direction (anti-Virgo v olume [AV]), were used via a Tully-Fisher (TF) relation to derive the following two parameters: H(AB), the mean Hubble ratio between AV and BV, and DELTAupsilon perpendicular-to, the peculiar velocity of the Lo cal Group in the direction of the Virgo Cluster (VC) with respect to a uniformly expanding reference system defined by our AV and BV subsamp les. The two sampled volumes, separated by a velocity interval of 5600 km s-1, form an antipodal pair. This particular geometry not only all ows us to derive the two parameters independently but also reduces the dynamical effect of the Local Supercluster on H(AB) Without increasin g the Malmquist bias. By limiting our sample to spiral galaxies having large velocity widths W(R), we effectively reduce the TF scatter and Malmquist bias in our sample. The TF zero point and dispersion were th en determined by further correcting for the small residual Malmquist b ias. An additional sample of fainter galaxies was used to test for a n on-Gaussian tail to the TF dispersion. We found no evidence for such a tail and formally give an upper limit of about 18% for the fractional contribution of an unseen tail. The average intrinsic TF dispersion f or the dominant Gaussian component is sigma(TF)0 approximately 0.33 ma g for W(R) greater than or similar to 180 km s-1. Our numerical result s are DELTAupsilon perpendicular-to almost-equal-to 414 +/- 82 km s-1 and H(AB) almost-equal-to (84.0 +/- 2.4)(1 + epsilon) km s-1 Mpc-1, wh ere (1 + epsilon) accounts for any systematic error between the calibr ators and the sample galaxies. Various dynamical models were tested to explore the effect on H(AB) of the uncertainties in the local velocit y field. Constrained by our observed DELTAupsilon perpendicular-to as well as other observational quantities, we found that the rms deviatio n from unity of H(AB)/H0 (where H0 is the Hubble constant for each mod el) is 5%, making H(AB) a good indicator for H0. Taking this variation as an additional error, our formal estimate for the Hubble constant i s H0 almost-equal-to (84 +/- 5)(1 + epsilon) km s-1 Mpc-1.