Published in "The Extragalactic Distance Scale", eds. M Livio, M. Donahue and N. Panagia, Cambridge University Press 1997.

For a postscript version of the article, click here.


George H. Jacoby

National Optical Astronomy Observatories, P.O. Box 26732, Tucson, AZ 85726

Abstract. The planetary nebula luminosity function (PNLF) technique for determining distances to galaxies now has been applied to 34 galaxies, including 6 in the Virgo cluster and 3 in the Fornax cluster. Of these, 16 galaxies are late-type or spirals and presumably contain Cepheid variables useful for verifying the PNLF method. For 7 of these galaxies, Cepheid distances exist; the PNLF distances agree with the Cepheid distances within the dispersion of 8% and within a zero-point offset of 1%.

In addition, 3 small groups were studied (NGC 1023, Leo I, and Coma I) where both spiral and elliptical distances were obtained to investigate the magnitude of any systematic dependence on spiral versus elliptical Hubble type. None was found. Since the PNLF method agrees well with the Cepheid system, and there is no measurable dependence on Hubble type, it follows that PNLF distances to the ellipticals in Virgo and Fornax also are on the Cepheid scale. This conclusion is strengthened by the Cepheid distances to several Virgo galaxies and the recent determination of a Cepheid distance to Fornax.

Challenges to the PNLF method by Bottinelli et al. (1991) and Tammann (1992) are demonstrated to be incorrect. In particular, the allegation that the Virgo distances suffer from inadequate survey depth is rendered baseless with recent observations that extend the PNLF beyond the power law regime and well into the plateau region. Using the new observations of 320 PN, bootstrap tests show that any sample size effect is smaller than 3%.

Finally, a simple thought experiment is presented whereby M87 is placed at 22 Mpc as argued by those favoring H0 ~ 50. The consequent luminosities for observed planetary nebulae are inconsistent with stellar evolution theory, thereby invalidating the assumption of a distance greater than ~ 17 Mpc; alternatively, a drastic change in stellarevolution theory is required.

Table of Contents