2 Observations

  Observations of all molecular lines in this survey were made using the James Clerk Maxwell Telescope (JCMT) in September 1995 and April 1996. We used the common-user receivers A2 (230 GHz band) and B3i (345 GHz band) coupled to the DAS (Dutch Autocorrelation Spectrometer). The bandwidth was a nominal 760 MHz, giving a frequency resolution of 0.756 MHz. Each spectrum was divided into channels of width 0.625 MHz. In September 1995, we tuned to the frequencies of lines of H₂S 2(2,0)-2(2,1) (216.710 GHz), CH₃CN J=13-12 K=0 (239.138 GHz), ${\rm C}^{17}\hbox{O } J=3-2$ (337.061 GHz) and CH₃CN J=19-18 (349.350 GHz). All data were taken in dual sideband mode, with sidebands centred $\pm 3$ GHz from the main band. Counting both main and image (upper or lower) sidebands, we observed between two and ten 760 MHz-wide bands in each source. Pointing was checked on G34.26 and was accurate to within 5$^{\prime\prime}$ throughout the observations. System temperatures were in the range 400-650 K with RxA2 and 900-1300 K with RxB3i.

Additional spectra of C¹⁷O J=2-1 (224.714 GHz) and C¹⁸O J=3-2 (329.331 GHz) were taken in April 1996. System temperatures were higher at 600-700 K with RxA and $4500-10\,000$ K with RxB. The resulting rms noise at 329 GHz was between 0.4 and 0.9 K, and 0.05 K at 225 GHz.

The spectra of G5.89 were obtained during March 1996 as part of a 300-360 GHz molecular line survey of this source (Thompson & Macdonald 1998, in preparation). These are by far the noisiest spectra with rms noises of between 0.08 and 0.18 K in 0.625 MHz channels.

With many lines visible in each spectrum, it can be difficult to attribute lines to main and image sidebands. In some cases we changed the local oscillator frequency by 10 MHz and observed a second spectrum in which the image sideband lines appeared shifted. In many cases we observed with lower and then upper sideband centred on the tuning frequency, so that different lines appeared in the image sideband. Line identification was performed with the help of laboratory and theoretical lists of millimetre transitions (Lovas 1985; Poynter & Pickett 1985; Anderson et al. 1990a,b; Boucher et al. 1980, Herbst 1996) All spectra and line temperatures quoted in this paper have been corrected for telescope forward scattering and spillover efficiency $\eta_{\rm fss}$ to give the main beam brightness temperature $T_{\rm R}^*$ in kelvin appropriate for extended sources. $\eta_{\rm fss}$ was taken to be 0.8 in the 230 GHz band and 0.7 in the 345 GHz band. Although the hot cores are smaller than the JCMT beam (which is $\sim\!21''$ in the RxA band and $\sim$14'' with RxB) we are also observing ambient cloud emission on larger angular scales, hence our correction for $\eta_{\rm fss}$ as measured on the moon. The conversion from $T_{\rm R}^*$ to flux S in Jy is given for the JCMT by $(S/{\rm Jy}) \simeq 13 (T_{\rm R}^*/{\rm K})$.

The observed sources are listed in Table 1 with the positions, distances (Churchwell et al. 1990) and LSR velocities (from ammonia; Cesaroni et al. 1992; Olmi et al. 1993) assumed. In Table 2, the frequency bands in which each object was observed are shown. Not all sources were observed in all frequency bands. In general, the sources which showed the most line emission were observed at more frequencies. Spectra were taken towards the HII region positions in all sources. Spectra were also observed at an offset position 20$^{\prime\prime}$ to the north of the HII region in five cores: G9.62, G10.47, G29.96, G31.41 and G34.26.

  

Table 1: Positions, velocities and distances for observed objects

  

Table 2: Observed frequency bands for each source, with important molecules with transitions in each band given