1 Introduction

Activity in the centres of galaxies ranges from extremely compact active galactic nuclei (AGN) powered by mass accretion onto a black hole (Begelman et al. 1984; Rees 1984) to extended star-bursts of more modest power. In some extreme ultra-luminous infrared galaxies (ULIRGs), e.g. in Arp 220 the star-powered luminosity may rival that of a compact AGN. Interest in the connection between the star-burst phenomenon and the central AGN has been stimulated by the fact that in galaxies with dominant AGNs, there may be significant amounts of circumnuclear gas, e.g. NGC 1068 (e.g. Tacconi et al. 1994) and Centaurus A (e.g. Rydbeck et al. 1993). Gas fuels the star-forming activity and responds both physically and chemically to the ultraviolet radiation and supernovae produced in a young population of massive stars. Dynamically, gas clouds are highly dissipative, relative to the stars and thus sink readily towards the centres of galactic gravitational potentials, possibly fuelling the black hole. Indeed, recent VLBI observations of molecular masers (e.g. NGC 4258) have revealed a very close connection between molecules and accretion discs surrounding galactic black holes Miyoshi et al. 1995. Also, recent interferometric observations of the type 2 Seyfert galaxy NGC 1068 have shown that the HCN gas is much more confined than the CO (e.g. Tacconi et al. 1996), and all interferometric studies (e.g. Helfer & Blitz 1997) show the HCN to be mainly concentrated towards the centre. We are therefore confident that the HCN emission will trace gas in the nuclear regions. In addition to this, Solomon et al. (1992) (hereafter SDR92) suggest, from their HCN observations, that (galactic) IR luminosities are a consequence of star formation, rather than of an AGN. In order to compare how Seyferts differ from non-Seyferts in their gas content, we investigate the CO, HCN and infrared luminosities of the 18 galaxies detected in CO  $1\rightarrow 0$ by Heckman et al. (1989) with the NRAO 12 m. We chose to observe these galaxies because:

1.

They form a good sample of Seyfert galaxies;

2.

To verify the results of Heckman et al. (1989), who find higher molecular gas abundances in type 2 than in type 1 Seyferts. This issue is addressed in Curran (2000a).

In addition to these results, in this paper we also present the CO  $2\rightarrow 1$  and CS  $3\rightarrow2$ results from the Southern part of this sample.