The kinematics of the LMC are now fairly well understood, with velocities of thousands of individual tracers of various types having been fitted in considerable detail with (thick) disk models. Questions that remain open for further study include the reality and origin of kinematical differences between different stellar tracer populations, the differences between the gaseous and stellar kinematics, and the amount and origin of non-equilibrium features in the kinematics. The kinematics of the SMC are understood more poorly, but appear generally consistent with being a spheroidal system of old stars with an embedded irregular disk of gas and young stars.
The HST PM work has provided the most surprising results in recent years, with important implications for both the history of the Magellanic System and the origin of the Magellanic Stream. Of course, it is natural in discussions about this to wonder about the robustness of the observational results. It should be noted in this context that many experimental features and consistency checks are built in that support the general validity of the HST PM results. These include: (1) the use of random telescope orientations causes systematic errors tied to the detector frame to cancel out when averaging over all fields; (2) the final PM errors are based on the observed scatter between fields, with no assumptions about the source and nature of the underlying errors; (3) two groups used different methods to analyze the same data and obtained consistent results; (4) P08 managed to measure a PM rotation curve for the LMC that is broadly consistent with expectation, which would have been impossible if the PM errors were in reality larger than claimed; (5) the difference between the LMC and SMC PMs is more or less consistent with expectation for a binary orbit, which would not generally have been the case if the measurements suffer from unknown systematics; (6) the LMC PM is consistent with expectation based on the line-of-sight velocity field of carbon stars (see Section 2.4); and (7) the LMC PM leads to an HI velocity field with a straight zero-velocity curve, by contrast to previously assumed values (see Section 2.4).
One interesting feature in the observational PM results is that with the P08 PM values, there are no bound LMC-SMC orbits, given their different µW and smaller error bars for the SMC compared to the K06b results. However, the SMC PM is significantly less certain that that for the LMC, due to the smaller number of fields observed with HST, and the fact that most of them were observed at a similar telescope orientation (which implies that potential systematic errors that are fixed in the detector frame do not average out when the results from different fields are combined). This underscores the need for additional PM observations. A third epoch of observations for most fields has already been obtained with HST/WFPC2, and preliminary analysis supports the validity of the results based on the first two epochs (Kallivayalil et al., these proceedings). A fourth epoch is planned with HST/ACS and HST/WFC3 in 2009. With the increased time baselines and use of multiple different instruments it will be possible to further reduce random errors and constrain possible systematic errors. In turn, this will allow new scientific problems to be addressed, such as the internal proper motion kinematics of the Clouds, and their rotational parallax distances (the distances obtained by equating the line-of-sight and proper motion rotation curves).