Anomalous Redshifts Reviewed

Abstract

IN a survey of the agreement between reliable optical and 21 cm systemic velocities of 130 galaxies, Roberts¹ finds that the regression line has a slope of one and passes within 1 km s⁻¹ of the origin. There is therefore exact statistical agreement between the two rays of determining the systemic velocity, confirming the form of the Doppler relation for wavelengths differing by a factor of 3×10⁵ over a velocity range of 5,600 km s⁻¹. But because the optical velocities are usually measured from emission lines in the nuclei, whereas the 21 cm velocities come from the mean velocity of gas spread over the whole plane, Roberts's result also shows that in general the nuclear systemic velocity agrees with that of the disk. It is therefore unlikely that any component of a non-Doppler redshift is present in the nuclear systemic velocities. It is difficult to reconcile Arp's² suggested anomalous redshift of small galaxies with Roberts's result. In view of the importance of a substantial non-Doppler redshift, were it to be confirmed, Arp's test for a redshift anomaly is repeated here using a larger sample of data.