The most obvious result is that the CO emission in the 30Dor Complex comes from well defined clouds. The sizes of the clouds shown in this study are in the tens of parsecs. Maps of the spectra have many characteristics of similar maps in Milky Way molecular clouds (except for the line strengths). All of this suggests that is reasonable to talk about the molecular cloud population of the LMC and to compare it with that in the Milky Way. Moreover, the line profiles result from a complex interplay of a variety of cloud conditions, including internal cloud kinematics, as well as temperature, density and abundance variations along the line of sight. The fact that the LMC line profiles appear similar to those in Milky Way molecular clouds suggests that many of these physical conditions are comparable in the LMC and the Milky Way.
Before discussing the individual cloud properties, we first look at the arrangement of clouds in each region. In the Central region, the clouds are arranged in an elongated structure, approximately 600 pc long. Optical images show that this coincides with a dust lane. The IRAS point sources in this region follow along the peak of the CO emission. In addition Caldwell & Kutner (1992) and Caldwell et al. (1993) have shown that there is extended FIR emission associated with this structure.
The clouds in this structure appear close together. In fact there is so much overlap that the individual clouds are only apparent in the maps covering restricted velocity ranges. This compact arrangement is in contrast to that of the N11 region, in which there are very clear gaps between most of the clouds (Johansson 1991; de Graauw et al. 1996). This difference in arrangement suggests that the local physical conditions may be very different in the complex south of 30Dor and in the N11 region.
The emission in the Southern region appears to be separate from that in the Central region. (If there is a connection between the two regions, it would probably be about 4 minutes east of the fully sampled N-S strip. That strip, does show some weak emission between the two regions, but we have not yet followed that up). The Southern region appears to have a much simpler structure than the Central region. The clouds there still have some overlap. There is a single strong FIR source at the peak of the CO emission. In addition, there is extended FIR emission covering the whole area in which CO emission is seen (Caldwell & Kutner 1992, 1996; Caldwell et al. 1993).
We have noted that, while the line profiles appear similar to those in Milky Way molecular clouds, the LMC lines are systematically weaker. We first ask whether this is could result from the effects of beam dilution at the greater distances of the LMC clouds. That is, if we took a typical Milky Way cloud complex and moved it to a distance of 50 kpc, how strong would the emission appear?
To answer this question, we can look at fully sampled maps of the
Orion region, and degrade the resolution to 10 pc. In the
undegraded CO(
) maps (with 0.1 pc linear resolution),
at the
peak is about 75 K, and in the extended cloud envelopes is about 5
K. In the degraded maps, the 
at the peak is about 20 K, and
in the envelopes is about 4 K. (Note, the peak only falls by a
factor of 
4 because there is still strong emission 5 pc from the
peak, and the envelopes do not change by even that much, because
there is little variation in intensity on the 5 pc scale). In the
LMC, the strongest peaks in the 30 Dor Central and South regions
(Table 1) are 5.2 K. These are a factor of 
4 weaker than in the
degraded Milky Way maps. In the extended envelopes (not near the
peaks or too close to the cloud edges), the lines are 1 to 2 K.
Again, these average a factor of 
4 weaker than in the degraded
Milky Way maps.
It therefore appears that the difference between the LMC and Milky Way line strengths is a real one. This means that there are some physical conditions that affect line strength, and are different in the LMC and the Milky Way. The most obvious of these are:
1. Abundances. The lower metallicity of the LMC results in lower
[C]/[H] and [O]/[H]. We would therefore expect a lower [CO]/[H
].
This lower abundance would produce lower CO column densities and
therefore weaker CO lines. One might think that, since the
CO(1 
0) line is optically thick in most Milky Way molecular
clouds, then dropping the column density should not have a
significant effect on line strength. However, we know that
trapping plays an important role in CO excitation and reducing the
column density can reduce the excitation temperature. This effect
is most pronounced in the lower density parts of the cloud where
the line is less easily thermalized.
2. Excitation. In addition to the trapping effects, mentioned above, it is possible that the temperatures and densities in the cloud are lower than in their Milky Way counterparts. These would result in lower excitation temperatures and weaker lines. Heating at the strongest peaks should be by embedded O stars. Caldwell & Kutner find that the FIR luminosities of these clouds are comparable to active clouds in the Milky Way, suggesting comparable internal heating. The extended envelopes of the clouds are heated by the interstellar radiation field (ISRF). Evidence suggests that the ISRF is not weaker in the LMC than in the Milky Way.
3. Clumping effects. There is growing evidence in Milky Way cloud cores and extended envelopes, that the CO emission is coming from clumps (on roughly the 0.1 pc scale) that fill a certain fraction of the cloud volume. The temperatures, densities and abundances that we measure from molecular line studies are within the clumps. The velocity dispersions that we measure are often dominated by the clump-clump motions. The properties of the lines that we see depend on the physical conditions in the clumps as well as the filling factor of the clumps. It is possible that in the LMC, the clump properties are different, or that they fill a smaller fraction of the volume than in the Milky Way.
We present these ideas to explore the range of plausible explanations for the weaker lines in the LMC. Further analysis of these possibilities requires observations of different transitions and different isotopic species, as well as a line formation analysis that takes clumping and trapping into account. Some discussion of these issues for the SMC are in Paper IV.
The Central region is associated with clouds #35 and #36 identified with the large scale CO survey of the LMC by Cohen et al. (1988), made with a 1.2 m radio telescope (FWHP beam of 8.8'). The Southern region lies projected toward cloud #34 in Cohen et al. It is interesting to determine the degree to which clouds we find with high angular resolution account for the CO luminosity seen by the 1.2 m telescope. If there is significant extended low level emission beyond the boundaries of the clouds that we have mapped with the SEST, then we would expect the total luminosity of the 1.2 m telescope to be larger than the sum of the SEST luminosities. That is because the larger beam includes the luminosities of the clouds and the extended component. If there is no significant extended component, then we would expect both telescopes to measure the same total luminosity.
In order to compare the low and high resolution luminosities, we have added all the spectra taken with SEST toward both the Central and Southern regions and those taken with the 1.2 m dish at CTIO covering the same areas. In Table 2, we summarize the parameters obtained from a single gaussian fit to the integrated spectra. As can be seen in Col. 5, the luminosities derived from the SEST and CTIO dishes are in good agreement, suggesting that essentially all the CO emission detected with the low angular resolution is coming from the clouds mapped with the SEST. In fact, the SEST luminosities are about 40% larger than those observed with the 1.2 m telescope, making very unlikely the presence of extended low level component of the CO emission. The lower values of the luminosities observed with the 1.2 m dish are probably due to beam efficiency effects, since the arc like structure generally runs off the main beam central position.
Table 2: Luminosity comparison
In understanding the evolution of the interstellar medium, and in particular the formation and destruction of the molecular clouds, it is important to examine the relationship between the atomic and the molecular gas. The most sensitive HI maps of this section of the LMC (but with a 15' beam, and 12' sampling) are by Rohlfs et al. (1984) and Luks & Rohlfs (1992). At this resolution, we can only look for general trends relative to our data. (We would hope, now that the Australia Telescope is available that HI maps with resolution comparable to our CO maps will soon be available).
In comparing the HI maps (especially the ``disk'' component identified by Luks & Rholfs) with the CO maps of Cohen et al., we see that for the whole 30 Dor Complex, the HI generally follows the CO. The HI is generally more extended, but the strongest emission is generally in the same place. The Central region corresponds with a distinct peak in the HI. The Southern region is part of an extended HI sub-peak.
In comparing velocity components, for the Central region, the HI
has two components. The higher velocity component is at 
ranging from 253 to 275 km s
. There is no CO emission from the
Central region in this velocity range. However, there is a lower
velocity HI component, which very closely follows the CO velocity
in the Central region. It even shows the shifts with position
evidenced in the CO (e.g. Figs. 5c, d). In the direction of the
Southern region, the HI also has two components. The higher
component is at 261 km s
, and there is one CO cloud at
approximately that velocity. The lower velocity component is at
234 km s
, which is the same as that of most of the CO emission in
this region (e.g. Figs. 5a, b).
We also note other features found in the two regions. The Central
region has six IRAS point sources falling within the boundaries of
the CO emission. The 100 
m flux densities range from 26 to
44 Jy. Each of the sources appears close to a CO peak. The
Southern region has one IRAS point source, with a 100 
m flux
density of 112 Jy. It appears located at the peak of the CO
emission. We should note that, in between the two regions, east
of the N-S strip, there are two IRAS point sources. This region
has not yet been surveyed by SEST.