We present a new large-scale survey of J=3-2 12CO emission covering 4.8 square degrees around the Rosette Nebula. Approximately 2000 compact clumps are identified, with a spatially-invariant power law mass distribution index of -1.8. Most of the inner clumps show velocity gradients of 1-3 km/s/pc, directed away from the exciting nebula. The gradients decrease with distance from the central O stars, and are consistent with a photoionised gas acceleration model, assuming clump lifetimes of a few 10^5 yrs. However, in one clear case, the observed near-constant velocity gradient is difficult to explain with simple models. Most blue-shifted but very few of the red-shifted clumps are associated with dark absorbing optical globules, confirming that the dominant molecular gas motion is expansion away from the central nebula and O stars. Many clumps also lie in a molecular ring, having an expansion velocity of 30 km/s, radius 11pc, and dynamical lifetime of ~1Myr. The J=3-2/1-0 12CO line ratios of the clumps decrease with distance from the O stars, implying a gradient in their surface temperatures; the results are consistent with a simple model of clump surface heating due to the central stars. Seven high-velocity molecular flows are found in the region, with a close correspondence between these flows and embedded young clusters. These outflows are sufficiently energetic to drive local gas turbulence within each cluster. We find 14 clear examples of association between embedded young stars seen at 24um and CO clumps; these are thought to be photoevaporating molecular envelopes. The CO clumps without evidence of embedded stars tend to have lower velocity gradients, and it is suggested that the presence of the young star may extend the lifespan of the photoevaporating envelope.

The evidence for a rotation of the epsilon Eridani debris disc is examined. Data at 850 micron wavelength were previously obtained using the Submillimetre Common User Bolometer Array (SCUBA) over periods in 1997-1998 and 2000-2002. By chi-square fitting after shift and rotation operations, images from these two epochs were compared to recover proper motion and orbital motion of the disc. The same procedures were then performed on simulated images to estimate the accuracy of the results. Minima in the chi-square plots indicate a motion of the disc of approximately 0.6'' per year in the direction of the star's proper motion. This underestimates the true value of 1'' per year, implying that some of the structure in the disc region is not associated with epsilon Eridani, originating instead from background galaxies. From the chi-square fitting for orbital motion, a counterclockwise rotation rate of ~2.75 degrees per year is deduced. Comparisons with simulated data in which the disc is not rotating show that noise and background galaxies result in approximately Gaussian fluctuations with a standard deviation +/-1.5 degrees per year. Thus counterclockwise rotation of disc features is supported at approximately a 2-sigma level, after a 4-year time difference. This rate is faster than the Keplerian rate of 0.65 degrees per year for features at ~65 AU from the star, suggesting their motion is tracking a planet inside the dust ring. Future observations with SCUBA-2 can rule out no rotation of the epsilon Eridani dust clumps with ~4-sigma confidence. Assuming a rate of about 2.75 degrees per year, the rotation of the features after a 10-year period could be shown to be >1 degree per year at the 3-sigma level.