OK so I was screwing around with my F30T8 magnetic preheat single lamp NPF desk lamp. I have a couple of moderately strong ceramic magnets (nothing crazy but more powerful than your typical fridge magnet...) and I noticed when I hold a magnet over one of the lamp electrodes it changes the pitch of the ballast to a higher pitch. If you stick another magnet over the other electrode the buzz gets even higher-pitched. I assume a stronger magnet on each end would amplify the effect too. It's very easy to notice on this ballast since the ballast is pretty loud. I suppose on your average ballast that only emits a muffled hum you might not notice it as much.

Anyone know what is happening here? It only happens with the magnets over the electrodes of the lamp. Anywhere else on the tube naturally has no effect. Is the electrode material magnetic? I wonder if placing the magnets on the ballast would cause any changes in the ballast's specs? Magnetism is almost as interesting as electricity. 

Well, I mean... It involves electricity... So I mean.. Like your indirectly interested in electricity now mate congrats, lol... I betchya Medved or Ash likely have a good explaination.

I'd think it has something to do with the magnet affecting the arc, I've heard of arcs being affected by magnets like I think you can push them in certain directions with a strong enough magnet, wonder if you're doing that and it's affecting the amount of work the ballast has to do to power the lamp

Sorry for the very simple sounding explaination but I hate using my phone to type and don't feel like using big words. Lol

I thought that might've been the case too but it only happens at the electrodes, not the center of the lamp. The arc spans the whole lamp. The magnet must effect the electrode itself somehow.

The magnet applies sideways force to the arc. But the force direction depends on the current direction, so it changes 60 times/sec together with the current

From there my guess is :

Normally when the filament is negative, the current exits from it in a small hot spot, so the lamp fill around the filament is ionized around this spot. When shortly after that the filament is positive, the hot spot location is not important but the ionized path is still near that spot, so current still takes the same path

With the magnet when the filament is negative, the current still exits in some hot spot (the spot might be pulled to another place on the filament by the induced force pulling the arc, but i think it will still be a spot). When the filament is positive, the induced force pulls the arc in another direction and away from the existing ionized path, so it might strike a new path (which needs higher overshoot voltage). Meanwhile the negative path is fading out, so when the filament becomes negative again, it again needs higher overshoot voltage to restrike that path, and same for the positive path

The higher overshoots might mean change in the crest factor, and that definitely would affect the sound of the ballast..

In the center of the lamp the arc is fairly diffused (its not a thin arc like in HID) so the entire volume of the lamp in that area is ionized anyway, so the magnet does not change as much

Ash has given a good explanation and if anyone knows he will. If you look at diodegonewilde channel on YouTube he does an experiment where he brings a magnet near a CFL and you can see the wattage increasing as he does