2 Observations and data reduction

The observations were performed during several extended sessions ranging from May 95 to June 98 with the radome-enclosed Onsala 20 m telescope. The telescope is equipped with a cooled SIS receiver. The single sideband system temperatures as measured outside the atmosphere were 250 - 1500 K. As backends we simultaneously used a 256 channel filterbank (bandwidth 64 MHz) and a 1600 channel autocorrelator (configured to a bandwidth of 40 MHz). The resulting velocity resolutions are 0.76 km s^-1 and 0.076 km s^-1respectively. We used the standard chopper wheel method for calibration. All spectra were taken in a dual beam switching mode (11' beam separation) with the exception of the Orion-close sources and NGC 1333 in the CS(2-1) line which were frequency switched (14-18 MHz frequency throw). Frequent pointing checks were made towards SiO masers, and pointing offset for all of the observations was confined within 5'' from the pointing model. We mapped the CS(J=2-1) (98.0 GHz), SO( $J_{\rm N}=3_2-2_1$) (99.3 GHz) lines together with transitions of the less abundant ³⁴S isotopomers, C³⁴S and ³⁴SO, at certain positions. The telescope FWHM beamsize is $\sim 40''$ and main beam efficiency is 0.56 at these frequencies. Only correlator data is presented here. The integration time was initially 120 s per position but in all cases the observations were repeated until we achieved satisfactory signal to noise ratio. The data have been reduced by the DRP package written by M. Olberg, Onsala Space Observatory, Gaussian fitting, where relevant, and displaying was made with the XS package written by P. Bergman, Onsala Space Observatory. A linear baseline was removed from all spectra taken in beam switching mode while a low order polynomial baseline was subtracted from the frequency switched data. Since the observations spanned over a long range in time extra effort has been put on checking consistency between scans taken at same position, line and backend at different occasions.