A molecular line survey in the 345 GHz window has been performed toward three sources in the W 3 Giant Molecular Cloud with the James Clerk Maxwell Telescope. The analysis not only uses the rotation diagram method, but also statistical equilibrium calculations to determine beam-averaged column densities and physical parameters such as kinetic temperature and density. The column densities are accurate within a factor of two, but the abundances are much more uncertain, however. Clear physical and chemical differences between the three sources have been found.
W 3 IRS4 is the simplest of the three lines of sight studied in this
work. The region is well characterised by a temperature of about 55 K
and a density of 
. Only the sulfur-bearing
molecules seem to favour slightly higher temperatures and densities. It
is proposed that the star has already broken free of the molecular cloud
and has created a PDR on the back side. In addition, a shock, perhaps
connected with the penetration of the ionization front into the dense
cloud, may be responsible for the creation of the sulfur-bearing
molecules and for the elevated temperature and density.
W 3 IRS5 is more complicated, since several physical components are
present in the beam. The bulk of the gas is well characterised by a
temperature of approximately 100 K and a density of 
, but gradients are present. Again the sulfur-bearing
molecules seem to trace the denser and hotter gas, which are proposed to
be formed in the shock connected with the very powerful outflow. The
hottest region is traced by SiO emission. Since this emission is
expected to be very compact, interferometer measurements are needed to
properly study this component.
Multiple components are also present toward W 3(
), with the
molecular emission arising partly from the extended core surrounding
both W 3(
) and W 3(OH), and partly from the compact clump near
W 3(
). The latter gas is warm (220 K) and dense (
) and contains complex organic molecules.
This survey has provided an unique view on three star-forming clumps within the same parent cloud. Chemical modeling will be the natural next step. In addition, further constraints on the physical structure of the IRS5 region must be obtained through a multi-component excitation analysis of both low- and high excitation lines. Mapping and interferometer data at subarcsec resolution, together with infrared absorption line data, are urgently needed to complete the picture of this high-mass star-forming cloud.
Acknowledgements
The authors are indebted to Fred Baas, Remo Tilanus, David Jansen and
the JCMT staff for assistance with the observing runs, and to Todd
Groesbeck, Peter Schilke and Eric Herbst for help with updating the line
catalogs. They thank Geoff Blake, Lee Mundy, Tom Wilson and Malcolm
Walmsley for useful discussions, David Jansen for his efforts in setting
up the excitation calculations, and Per Bergman for providing them with
his 
collisional rate coefficients. This research was
supported by the Netherlands Organisation for Scientific Research (NWO)
through a PIONIER grant, and by the Space Research Organisation
Netherlands (SRON).
Table 7: Gaussian fit parameters
Table 7: (continued)
Table 7: (continued)
Table 7: (continued)
Table 7: (continued)
Table 7: (continued)
Table 7: (continued)
Table 8: 
Gaussian fit parameters W 3 IRS4
Table 9: 
Gaussian fit parameters W 3 IRS5
Table 10: 
Gaussian fit parameters W 3(
)
Table 10: (continued)
Table 10: (continued)
Table 11: 
Gaussian fit parameters W 3(
)
Table 12: 
Gaussian fit parameters W 3(
)
Table 12: continued
Figure 4:
Four panels containing double side band spectra for W 3 IRS4 (upper
spectrum), W 3 IRS5 (middle spectrum) and W 3(
) (lower spectrum).
Identifications in the lower side band are given with solid tickmarks,
in the upper side band with dotted tickmarks. The lower side band is
chosen as the abscissa. Due to different 
for the three
sources the upper side bands do not line up. To ease the
identification, the frequency ranges for the spectra are given at the
top of each panel. The molecular identifications are only given in
W 3 IRS5 but apply to all three spectra. The corresponding molecular
parameters are found in Tables 7 (click here)-12 (click here)
