5 Conclusions

We have compiled a large data base of rotational measures for a broad range of molecular cloud types, derived from observations of velocity gradients taken primarily at millimetric wavelengths. As a consequence, it has proved possible to undertake a detailed investigation of cloud rotational parameters, and to evaluate the importance of angular momentum in maintaining cloud stability.

We have demonstrated that two measures of stability based on the ratio between angular momentum and turbulent/thermal virial terms (parameter $\beta$), and between rotation and gravity ($\alpha$) lead to somewhat differing conclusions. Given the potentially large systematic errors to which $\beta$ is prone, however, it seems likely that a substantial fraction of clouds may be appreciably stabilised by rotation. Employing only parameter $\alpha$, for instance, we find that $\sim 50\%$ of clouds and disks are to some degree stabilised through rotation, although clumps and condensations appear more dependant upon turbulent support. High levels of angular momentum may also be responsible for departures in cloud sphericity, and we note that structures with large values of $\alpha$ and $\beta$ are characterised by typically larger aspect ratios $\Gamma$.

The orientation of $\Omega$ is found to vary markedly between various cloud groups, with clumps and condensations displaying a more-or-less random distribution with respect to the galactic plane. Isolated clouds (and perhaps disks), on the other hand, appear to favour orientations towards the north and south galactic poles.

These disparities presumably imply that rotation in larger cloud structures and disks derives from galactic shear, whilst vectors $\Omega$ for clumps and condensations have either been randomised through dynamical and/or magnetic interactions, or derive from turbulent vorticity.

A comparison between observed functional trends in $\Omega$, J/M, and J, and those predicted through models of cloud rotation has proven particularly interesting. Thus, it appears that derived trends in J(M) are broadly similar to those predicted through simple clump merger models. On the other hand, the variation of J/M with radius appears not to be consistent with models of isothermal rotating clouds, whilst dln$\Omega$/dlnR is inconsistent with either conservation of angular momentum or models of turbulent vorticity (although in this latter case, clumps and condensations follow predictive trends rather more closely than is the case for the sample as a whole). Whilst magnetic braking may account for at least some of the observed decrement in J/M, the present results imply gradients dln(J/M)/dM which are less than would normally be expected through such a mechanism; a disparity which may arise through a variety of causes, including the trapping of Alfven waves during periods of rapid cloud contraction.

Whilst most of the functional trends for cloud rotation display moderately high levels of correlation, there appear also to be certain exceptions to this rule. Thus, the trend between J and M has a correlation coefficient r = 0.98, whilst the coefficient for $\Omega(M)$ varies between 0.03 and 0.41 (depending upon cloud type). Whilst it seems likely that such differences must arise from the nature of angular momentum transfer from large to small scales, we have as yet little understanding of the root origins of these disparities.

We have, finally, noted that gradients dln$\Omega$/dlnR, dln(J/M)/dlnR, and dlnM/dlnR for disks appear significantly different from those characterising most other cloud subgroups; a disparity which presumably derives from their markedly differing spatio-kinematic structures, and proportionately high fractional stellar mass contents.