In any case, given that I posted a bit about how bats launch, it seems only fair to also point out that great work has been done on how they land, as well. Dan Riskin and colleagues have a great paper in the Journal of Experimental Biology (freely available here) where they examined the tricky business of landing on ceilings. As it turns out, there are a few ways to do it, and one of them basically involves a cartwheel in the air to bring the feet, and then the hands, in contact with the inverted substrate.
This, in turn, brings me to a point that I have been making at conferences lately: landing and launching from ceilings is tricky business. The exact kinematics of ceiling-launches in bats have not been elucidated in great detail, but it is known that they are quite acrobatic. This is important, because one of the bits of rebuttal I hear rather often about large fossil flyers (mostly pterosaurs) is: "why couldn't they just drop off a cliff for speed?"
The answer is that this turns your average giant pterosaur in a rather elaborate lawn dart. Dropping head-downward to takeoff is actually a pretty inefficient way to go unless one happens to live on ceilings (like bats). Consider this: to successfully launch, one of the giant flyers (such as a large pterosaur or teratorn bird) would probably need around two g's of acceleration upwards (closer to three would be ideal). If they start by falling, then they are accelerating in the wrong direction at a g, not to mention that the poor critter is oriented in a very compromising way.
Launching like a falling stone is tough going - bats do it, but it's a very specialized trick. So, please make myself and other flight folks stop wincing: don't drop your pterosaurs.