Well, by popular request...

On the Earth, we've got something called the Coriolis effect, inherent to the rotation of the Earth. It gets stronger as you move poleward; it is very weak at the equator, gradually increasing through the tropics, then ramping up around 45 N/S. It is a right-of-motion deflection, meaning anything moving to the west is going to tend to be deflected towards the north. We quantify this effect with the Coriolis parameter, termed f in meteorology. It is generally much smaller than the vorticity (spin) found within a tropical cyclone by 100+ times, but is nevertheless significant.

As a storm grows larger -- or, alternatively, as it grows stronger and its peripheral impacts reach over a larger area -- it is able to reach both higher and lower values to the north and south (respectively) of the Coriolis parameter than were it at a normal (or its original) size.

Now, picture the flow around a hurricane in the Northern Hemisphere (you can apply these to the southern hemisphere, but for simplicity and practicality's sake, I'll stick to the NH here). Air -- we call it a parcel of air -- due west of the center is going to move south, while that due east of the center is going to move north. So now, you have two parcels displaced from the latitude of the center, one in the SW quadrant of the storm, another in the NE quadrant.

In general, vorticity (here, the spin of the atmosphere plus the Coriolis effect) is a conserved quantity, meaning if one goes up, the other has to go down to compensate. So, picture the air parcel on the SW side of the storm. Since f decreases towards the south, the value of the vorticity of the parcel is greater than that of its environment (where the value of f is smaller). Similarly, on the NE side of the storm, the value of the vorticity of the parcel is less than that of its environment (since the value of f increases to the north, this is where the value of f is larger than that of the parcel of air).

On the SW side, we call this greater value of vorticity a positive perturbation, whereas on the NE side, the smaller value of vorticity is termed a negative perturbation. In meteorology, positive values of vorticity -- again, spin -- are associated with cyclones and cyclonic flow. Thus, on the SW side of the storm, you get this weak cyclonic eddy; this results in a flow from the southeast to the northwest across the storm. On the NE side of the storm, you get a weak anticyclonic eddy, also resulting in the same flow across the storm.

Note that further away from the storm with these eddies, you get flow to the southeast, but that affects the outside environment and not the storm. It's hard to describe with words, but I can't find a suitable picture.

The stronger (larger) a storm is, the stronger these eddies become. Thus, in general, the natural tendency of a symmetric tropical cyclone is to move to the northwest; the larger flow pattern, such as a strong subtropical ridge, usually overwhelms this effect. Nevertheless, these eddies can impart a northwestward motion upon the storm of about 2-3mph....trending more to the north with a larger/stronger storm. Not large, and generally negligible, but it is an effect included in the Beta advection models -- in fact, this is exactly what Beta advection is, the so-called beta effect.

I've tried to make this as non-technical as possible, but I'm sure it is probably still too much, for which I apologize. I'll try to refine it a bit if need be, though. As always, this is somewhat of a simplified model, as it assumes a symmetric storm in an ideal environment; the same principles hold for shearing environments and/or asymmetric storms, but just to different magnitudes and perhaps slightly different directions.

Hope this helps, at least a little!