Weird behavior with high voltage LCD driver

[Red Squirrel]

I'm looking to see if I can determine if the backlight of an LCD itself is shot (it flickers a lot, my hunch is it's in fact the bulb) or if it's the actual driver.  But since I'm dealing with unknown high voltage I don't want to just stick my scope on it.  My goal is to have the scope on it using a voltage divider and wait till it starts flickering to see if anything changes. 

So I built a basic voltage divider, a 100k and 10k resistor.  I had my hand held multimeter (note: not grounded) connected to the divider across the 100k resistor, and as soon as I place only ONE wire from the divider on the high voltage output, I get an arc and the LCD actually dims!  What is it arcing to if the other end is not connected to a return path?   It's also drawing a decent amount of power if it's enough to make the LCD dim, but how?  Where is that power going?  If I take a short piece of wire and touch one end to it I get the same thing.   (insulated wire of course).  I'm kinda stumped.  I know high voltage can arc through the air but is that really what's happening here?  How does the circuit even work then, the traces are not shielded or anything like that.

[bktemp]

The voltages used for driving CCFL tubes are around 1kV and the frequency is somewhere between 30kHz und 70kHz. Since the tubes are driven using a constant current, even tiny amounts of capacity (<10pF) load the output signficantly.
Measuring the output voltage of a CCFL inverter can be difficult, because you need a high impedance at the operating frequency. Even a standard 100:1 probe often has too much capacity and affects the output voltage.

[Red Squirrel]

But where is the current going though when I put a wire on it?  Even a short wire with nothing on the other end loads it down!  Is it basically acting as a radio transmitter because of the high voltage and high frequency and the wire is an antenna? 

[Red Squirrel]

lol so I just blew a 10k resistor, I decided to put my voltage divider with dmm across and see what happens, figured maybe if it's all connected up and then I turn it on it won't be arcing or doing anything weird, did not figure I'd blow a 10k resistor this way, there must really be a large amount of current coming out of there  .  My multimeter seems to have survived. I think.     I should have tried without it connected first. 

I'm almost convinced it's the bulb that's busted though and probably not worth replacing.  This was more just for fun, and think I've gotten beyond voltage levels that I'm equipped to work with safely.  I can always make a voltage divider in the megaohm range and try again I suppose.