... galaxies

Based on results collected at the European Southern Observatory, La Silla, Chile and Onsala Space Observatory, Sweden.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... galaxies

These were chosen on the grounds that they provide a good sample of Seyferts. As well as this, they constitute the same sample of Heckman et al. (1989), thus enabling us to also test their finding that type 2 Seyferts have higher molecular gas abundances than type 1s (see Curran 2000).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... HCN

From now on, unless otherwise stated, HCN or CO refers to the  $J=1\rightarrow0$ transition for the appropriate molecule.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...

This actually corresponds to 7 kpc from the central position but in order to account for beam smearing, we have deconvolved (as explained in Paper I) the HCN beam, giving 5 kpc.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1993

Usually, in the absence of maps, the HCN $1\rightarrow 0$ line is assumed to have the same source size as the CO $2\rightarrow 1$ line, i.e. as Curran et al. (2001). But here we may find the former to be considerably larger, i.e. 30'' cf. 20''.This would have the effect of increasing the central $\frac{{\rm CO}~1\rightarrow0}{{\rm HCN}~1\rightarrow0}$ intensity ratio from $\approx$8 to $\approx$16 in Curran et al. (2001), although it does not significantly change their results as the kinetic temperature and molecular hydrogen density are constrained from the CO isotopomer intensity ratios. If true, it would, however, lower the density of the dense gas component slightly, which supports the authors' hypothesis that Circinus possesses denser cloud cores in comparison to NGC 4945.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... extended

Of $\approx$20 kpc radius according to the NASA/IPAC Extragalactic Database.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1998)

Again this may affect the $\frac{{\rm CO}~1\rightarrow0}{{\rm HCN}~1\rightarrow0}$ intensity ratio (Curran et al. 2001), raising it from $\approx$17 to $\approx$23, and, for the same reason as before (Sect. 2.1.4), this does not alter the kinetic temperature and molecular hydrogen density constrained from the CO isotopomer intensity ratios. The density of the dense gas component could again be lowered but not to the same extent as for NGC 4945.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... sample

Although of the near-by galaxies of Curran et al. (2000), NGCs 1068, 1365 and 2273 appear to have their CO fairly confined, as opposed to NGCs 5033 and 6814 (Sects. 2.1.1 to 2.1.7). However, only NGC 2273 has the majority of the CO within the central beam (Table 3 of Paper I).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1988)

Or that the HCN luminosity of Circinus is relatively weaker, Table 2.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... beam

The maps are not shown in Papadopoulos & Seaquist (1998). When we compare our values of $T_{\rm mb}$ with theirs, however, we find that, after correcting for the different NRAO and SEST beams, while the CO $1\rightarrow 0$values match between our and their observations, the (total mapped?) CO $2\rightarrow 1$ value of Papadopoulos & Seaquist (1998) is only about half of ours (Curran et al. 2000). Our only explanation for this is that their $2\rightarrow 1$antenna temperature has perhaps not been corrected for main-beam efficiency (our uncorrected antenna temperature better matches their quoted value). In any case it appears as though we have indeed captured all of the CO $2\rightarrow 1$ flux in the SEST beam.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... region

In fact the ratio is comparable to that of the global ratio of the distant sample (Sect. 2.2) and of ULIRGs (Solomon et al. 1992; Helfer & Blitz1997).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1994)

In fact Maiolino et al. (1998) suggest that the star-burst luminosity on this scale is comparable to the luminosity of the AGN, while being only $\sim$3% of this within 12 pc of the black hole, thus supporting the notion of an inner limit.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... 2000b)

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... thought

At least in the near-by sample.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1992)

From Fig. 2 of Solomon et al. (1992) perhaps it is not so surprising that we find such differences between the near-by and distant samples, as the FIR/HCN relationship spans all the way from normal spirals (e.g. the ratio in Circinus is similar to that of the Milky Way) all the way to ULIRGs, i.e. a range of $L_{\rm FIR}\sim10^{10}{-}10^{12}{L}_{\odot}$, Figs. 19 and 20.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... engine

Note that in NGC 4194 Aalto & Hüttemeister (2000) find a deficiency of HCN, although the star forming efficiency may be similar to that of ULIRGs. Also Evans et al. (2001) suggest that the high infrared to CO luminosities in ULIRGs and QSOs may be due to an active nucleus heating the circumnuclear dust.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...$\frac{L_{\rm FIR}}{L_{\rm CO}}(\tiny {\rm distant}\normalsize )\mathrel{\mathch... ...>\cr\sim\cr}}}}4\frac{L_{\rm FIR}}{L_{\rm CO}}(\tiny {\rm near-by}\normalsize )$

This and the previous point may also indicate a CO excess in the near-by sample which may constitute a resevoir which fuels star formation in the near-by sample. See Curran Marconi et al. (2000) and the following argument.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... answer

Of course the obvious solution is that, whereas we have sampled most of the HCN (Paper I), in the near-by galaxies, all of the FIR flux may not be sampled. However, comparing the results of the IRAS catalogue (Lonsdale et al. 1985) with those for the large optical galaxies, which are expected to be resolved by IRAS (Rice et al. 1988), we see that for the two galaxies of our sample in common with the catalogues, NGCs 1365 and 5033, that all of the FIR seems to have been sampled, even despite the vast extent of NGC 5033 (Paper I and Thean et al. 1997).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... beam-filling

This could be compounded by the fact that the gas traced by the HCN (and CO) is expected to increase in optical depth (e.g. Krolik & Begelman 1986; Krolik & Begelman 1988) as one approaches the matter responsible for the obscuration of broad lines in (type 2) Seyfert galaxies (e.g. Antonucci 1993; Urry & Padovani 1995). Therefore there would be line emission only from the facing edge of this in the beam, whereas more of the infrared continuum would be visible through scattering (Miley et al. 1985; Edelson et al. 1987). Note also that by compiling HST (NICMOS) near-infrared ($2~\mu$m) observations to produce spectral energy distributions, Quillen et al. (2000) find that the mid-infrared opacities in Seyferts could be much less than previously thought. Further to this, due to the large opening angles (see Ulvestad & Wilson 1984; Wilson & Tsvetanov 1994; Sandqvist et al. 1995; Curran 2000b, (or for a review)) in the obscuring "torus'', the radiation from the accretion disk may be able to heat significant portions of the galactic dust located well away from the nucleus.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... source

For example, Sandqvist et al. (1988); Arsenault (1989); Baum et al. (1993); Taniguchi & Ohyama (1998); Chapelon et al. (1999); Grossan et al. (1999); Oliva et al. (1999); Papadopoulos & Seaquist (1999) and references therein.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...sample

For example, Doherty et al. (1995); Calzetti (1997); Richmond et al. (1998); Hill et al. (1999); Anantharamaiah et al. (2000); Carilli & Taylor (2000); Poggianti & Wu (2000) and references therein.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... holds

Although it is based only on a single Sy1 datum for the near-by sample and two values for the distant sample.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... densities

See Curran et al. (2001) for a detailed analysis of NGC 4945 and Circinus.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...$2\rightarrow 1$

As is found in NGC 4945 (Dahlem et al. 1993; Mauersberger et al. 1996) and the Circinus galaxy (Curran et al. 1998).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... observations

As has already been done for NGC 1068 which appears to confirm our results (Helfer & Blitz 1995).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... ALMA

For example, using the full 10 km baseline, a frequency of $\approx$100 GHz will give a linear resolution of $\approx$40 pc at the distance of Mrk 273. Also, being located in Chile, ALMA will prove invaluable in mapping NGC 4945 and Circinus which are too far south to be observed by current arrays.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.