I'm with NEiLS3LK51,
I've come across this teaser a few times in the past, both in aviation-circles (discussed it with engineers, pilots and others involved in simulation), and otherwise...
As Neil has hit on, it's all about the transition between being a vehicle on the ground and a vehicle in the air (and that's the reason a lot of the answers I've heard/seen are incorrect, no matter how credible the source might seem in cases)...
As prescribed by Neil and Bernoulli's principle, a conventional wing needs air passing above and below it at different speeds to cause an offset in pressure resulting in lift. This means that to generate lift, the wing has to be moved through the air, or the air has to be moved past the wing. The speeds required to generate lift are dependant upon a number of factors, but we'll keep it simple for now...
These aren't _strict_ aerospace numbers and practices, but it'll make it easier to see what I'm driving at...
Imagine that I build a conventional, fixed-wing aircraft and outfit it with wings which will provide lift for it above 150 knots (in these examples we'll assume that air-pressure and temperature are constant, and there's no wind)...
So, to get my aircraft to be an aircraft, and not just an odd looking car, I need to have a closing-speed between wing and air of 150 knots.
I could acheive this by pushing the aircraft through the air until I reach that speed, or I could even move the air across the wing at that speed - YES! If my aircraft was parked in a smooth-headwind of anything above 150 knots, it could theoretically be a vertical take-off aircraft too! That's unlikely though, as winds that speed aren't exactly smooth, so you'd probably roll it or flip it into the ground at the first anomaly. So we revert to plan (a), moving my aircraft through the air until it reaches 150 knots and the lift is present to get it airborne.
Now imagine a car on a rolling road once again, lets take a de-restricted Veyron for instance... Now lets say I come along and while you're not looking, weld-on some wings I've carefully designed to lift the mass of the Veyron (and the wing, driver, fuel etc.) at 150 knots (in the same atmospheric conditions as my plane)...
OK, we're ready to go... Get on the gas! Your Veyron will take off once it's moving 172.62 mph through the air... WE'RE GOING TO SEE A VEYRON FLY!!!
...Alas no, it may take up to a full tank of juice to realise, but we're not going to get your winged-Veyron moving through the air at over 170 mph until we take it off the rollers...
EDIT: And to just quickly further deal with any point regarding the wheels of the aircraft not providing the power, it's irrelevant where the power comes from until you've gained the magic 150 knots airspeed. Lets imagine that my wing as I mentioned, generates lift at 150 knots, but that's only when you have some extension of flaps, if the flaps are retracted, we need 180 knots.
Now I put some clever-engine technology on my aircraft, meaning the wheels can drive it up to oh, 180 knots. My clever engine-technology also means I can cut the jets in at 100% thrust in an instant.
I now align my aircraft to the threshold of a non-conveyor-belt runway; with the flaps retracted and apply full power to the wheels... I roll to 170 knots and extend the flaps. My aircraft now has sufficient airpseed for the wing's configuration to provide lift and I take-off. At the same moment, I apply full power to my turbofans and up I'll stay. Any method of propulsion can do it, but you still need a method of getting the aircraft from 0-150 knots before the magic of the wing's interaction with the air will stop you having to worry about such trivial things as the ground... (Perhaps a BIG catapult on a LOOOOOONG (and not very short roller-equipped) Aircraft Carrier as Neil said...
Regards,
Rob.
PS> Glad to be back with the audi-sport.net bunch again, after another little gap, all will be explained soon...