In order to construct an intermediate redshift TullyFisher relation, we used galaxies satisfying the following two selection criteria: (1) a difference between the position angle of the slitlet and the position angle of the major axis of the galaxy less than 45, and (2) consistent [OII] and broad-band scale lengths. Criterion (1) was applied only to the MOS data, since a misalignment between the major axis of a galaxy and the position angle of the slitlet artificially decreases the magnitude of V sin . Figure shows V sin versus kcorrected absolute B magnitude for the 12 remaining kinematically ``normal'' galaxies. V sin 's and B magnitudes were taken from Tables and respectively.
The solid line is the expected H rotation velocity as derived from the absolute calibration of the B-band inclinationcorrected TF relation for local galaxies [\protect\astronciteJacoby et al.1992][\protect\astroncitePierce and Tully1992] and a linear fit to the HI velocity widthH rotation velocity data of 204 galaxies [\protect\astronciteMathewson et al.1992]. See section for details. Also shown is the H TF relation expected for randomly oriented disks (dashed line) based on the fact that <V> = 0.7854 <V> for randomly oriented disks.
Since we selected relatively elongated galaxies on CNOC images, and seeing tends to make galaxy images rounder, it is likely that the observed galaxies were on average much more inclined than a randomly inclined sample. The solid line should therefore be closer to the local TF locus of our sample. The implications of Figure for galaxy evolution are discussed in section .