4 Discussion

 

4.1 Classification of radio sources

We shall now classify the sources detected in our survey and reported in Table 1 according to morphology, spectral index and coincidence with IRAS sources.

Five complexes of sources have all the specifications for being classified as thermal galactic HII regions. They include all the extended sources plus some additional small diameter source in the same area, more precisely [11 + 12 + 30], [13 + 15 + 16 + 17 + 31], [34], [18 + 33] and [32] (numbers are from Table 1). All these complexes coincide with corresponding sources in the Altenhoff et al. (1978) survey (see Sect. 3.2) and are now resolved in much more detail. Morphologically they show the classical aspect of a cluster of HII regions, of which G9.62+0.19 is a typical example (Testi et al. 1998, 1999), i.e. several sources of different angular sizes (from unresolved to several tens of arcsec) are clustered in the same area. The continuum sources may represent independent UCHII regions born in the same star forming complex but presently observed in different evolutionary phases with the unresolved sources being the youngest and the more extended sources more evolved.

Six of the small diameter sources (2, 3, 8, 9, 19, 22) can be classified as "candidate HII region'' according to their spectral index. No IRAS source is associated with them, but their radio flux is rather low. Non detection at far infrared wavelengths could be either due to the intrinsic weakness of some of these sources or, most probably, due to the incompleteness of the IRAS-PSC in the galactic plane (see also Becker et al. 1994).

The remaining 15 sources (4, 5, 6, 7, 10, 14, 20, 21, 23, 24, 25, 26, 27, 28 and 29) can be classified from the spectral index as non thermal (probably extragalactic) sources. Only five of these have been detected at 20 cm. These have in general greater integrated flux densities at 6 cm than those not detected at 20 cm (the mean 6 cm flux densities of the two groups are 10 and 2 mJy, respectively), so that the effect can be simply explained as due to our higher sensitivity at 6 cm. All 15 sources have been detected at 6 cm and 4 of them at 3.6 cm as well. Given the area observed at 6 cm (0.620 sq. deg.) and that observed at 3.6 cm (0.525 sq. deg.), the number of extragalactic sources above the 1 mJy threshold, which we can assume as a mean detection limit for our survey, can be estimated from deep VLA surveys. Following Fomalont et al. (1991) at 6 cm we expect 15 extragalactic sources above our 1 mJy threshold, while at 3.6 cm the number is reduced to 9 sources for the same threshold (Windhorst et al. 1993). Given the small number statistics, these numbers are in relatively good agreement with the source counts in our surveyed area.

Becker et al. (1994) estimated a total of $\sim 100$ planetary nebulae (PNs) down to a flux limit of $\sim 2.5$ mJy in their 6 cm survey of 50 sq. deg. of the inner galactic plane. This number correspond to less than 2 PNs expected down to the same flux level in our 6 cm survey region. Thus the contamination from PNs in our source lists should be very small.

4.2 IRAS "UCHII-type'' sources

In Sect. 3.4 it was pointed out that all the IRAS sources with a corresponding radio counterpart in our survey (5 out of 43) satisfy the color-color criteria to be classified as UCHII regions (WC89). However, with the possible exception of the double source G045.070+0.132 and G045.072+0.132 (11 and 12), none of the radio sources within 100$^{\prime\prime}$ from the IRAS-PSC position can be classified as bona fide UCHII region using the usual definition (Wood & Churchwell 1989b; Kurtz et al. 1999a). The radio continuum sources are extended (non homogeneous) HII regions, with emission peaks inside them that may appear as UCHII regions when observerved with an extended VLA configuration. A typical example could be G045.455+0.060 which appears as a compact source inside the extended HII region G045.455+0.059 (see Fig. A5), this source has the appearence of an UCHII region in the Wood & Churchwell (1989b) survey (their source G45.45+0.06). The VLA high frequency and high resolution surveys of IRAS selected UCHII candidates are all biased to the detection of only the most compact and dense ionized gas, due to the spatial filtering of the interferometer, and are unable to detect the extended components. Our results agree with those of Kurtz et al. (1999b) and show that, when observed with sufficient sensitivity to extended structures, most, if not all, the IRAS selected UCHII candidates do have extended radio components. This implies that samples of IRAS-PSC sources selected with the WC89 criteria are contaminated by a substantial number of older more extended HII regions (see also Codella et al. 1994; Ramesh & Sridharan 1997; Kurtz et al. 1999b). The number of UCHII regions estimated from the color selected IRAS-PSC counts may be, consequently, overestimated by a large factor. If most IRAS-WC89 sources are indeed associated with extended HII rather than UCHII regions, the lifetime of the IRAS-WC89 color phase of UCHII/HII regions may be much longer than estimated from the early high resolution radio surveys. Consequently, the estimated lifetime of the UCHII phase for O-type young stars is probably much shorter that previously thought (see also Ramesh & Sridharan 1997). Additionally, we find 6 UCHII candidates in our radio continuum survey without an associated IRAS source. As discussed by Becker et al. (1994), this is probably due to the confusion limit of the PSC on the galactic plane, and the generally lower radio luminosity of these sources. However, we note that in our field only unresolved thermal radio sources are not present in the IRAS-PSC, while all resolved HII regions are detected in the far-infrared. Incidentally, we note that all the compact thermal radio sources in our survey not associated with IRAS PSC sources are fainter at centimeter wavelengths than those detected in the far infrared, and thus they may be associated with stars of type later than O. However, without knowing the distances it is impossible to draw a final conclusion.

In our surveyed region, the percentage of IRAS sources satisfying WC89 color criteria is (5/$43 \sim 12$%). This is consistent with the percentage found accross the entire inner galactic plane ($\vert l\vert\le 90^\circ$, $\vert b\vert\le 0\hbox{$.\!\!^\circ$}6$, $\sim 8$%). The fraction of WC89 sources in the IRAS-PSC database drops to much lower values outside the inner galactic plane (WC89).

4.3 Continuum emission from the H₂O maser

During an incomplete low spatial resolution (2$^\prime$) single dish survey of the $l=+45^{\circ}$ field in the H₂O 22 GHz maser line, a new maser was detected. The masing gas is probably coincident with a 15 $\mu$m source ($F_{15~{\mu}\rm m}$ = 370 mJy) located at $\alpha(2000)=19^{\rm h}12^{\rm m}46^{\rm s}$ $\delta(2000)=10^\circ45^\prime30^{\prime\prime}$,and was interpreted as a candidate young stellar object (Testi et al. 1997). Therefore, it was interesting to see if any radio continuum emission from an associated UC HII region could be detected.

From a careful inspection of the area around the maser, no radio continuum emission was seen above the (local) 3$\sigma$ level (0.6 mJy/beam at 3.6 cm and 1.2 mJy/beam at 6 cm). With the young stellar object hypothesis in mind, there are two possible explanations: 1) the putative UCHII region is intrinsically too weak to be detected or absent because the eventual exciting star is of late spectral type; or 2) there is an UCHII region, but it is in such an early evolutionary phase that it is optically thick even at 3.6 cm. The lack of radio continuum emission close to H₂O masers in high luminosity star forming regions has been amply demonstrated by a survey of a large number of maser in the radio continuum, which showed that many maser associated with high luminosity sources do not have any close-by radio continuum source (Tofani et al. 1995). Subsequent molecular observations of the masers without continuum emission has indeed confirmed that these are associated with very young star forming regions since in all cases a hot molecular core was found at the same position (Cesaroni et al. 1999).

To settle the nature of the new maser-15 $\mu$m source, molecular observations in high density tracers are needed, as well as an estimate of its distance and luminosity.