Q: How many light bulbs does it take to screw with your head?

A: (See below.)

I am at home, right now, staring balefully at the light bulb in the ceiling above my desk in the room that I use as my office. I took this room because no one else wanted it. Two of the four walls are exposed brick, but not the sort of exposed brick suitable for a classy urban living room; it is the sort of exposed brick suitable for the sewers beneath Bratislava.

This is a very old house, and when we moved in 20 months ago, we had to rewire it. That meant replacing all the lighting fixtures, which meant purchasing all new light bulbs. There were 25 of them, all screwed in on the same day, all identical, except for the one I am now staring at. It has bothered me from the get-go, because the workers accidentally got a small smudge of paint on it.

It's bothering me more now. That is because all the other 24 bulbs -- same brand, same batch -- burned out long ago; the longest lasted seven months. Most have been replaced two or three times. But this one, used just as much, has been burning for 20 months.

Now, I know what you are thinking. You are thinking: This is one boring column.

Not so fast. I am on a mission to prove that the moguls of the light bulb industry are engaged in a vast conspiracy of planned obsolescence, a monstrous collusion to keep us buying their products when they know that a simple dab of paint will indefinitely increase a bulb's longevity, offering some small solace in these difficult economic times.

Not so boring anymore, is it, Mr. and/or Mrs. John Q. Public?

The bulb in question is made by Feit Electric, in Los Angeles. I called president Aaron Feit, explained the situation and asked him what might account for it.

"Luck," he said.

As well he might.

Of course, the average person might accept this explanation. But the average person has some sense of proportion and maturity. I contacted Cornell mathematics professor Richard Durrett, one of America's leading experts on probability theory. I asked him the odds that this was simply coincidental. He said there were too many unknown variables to make a responsible estimate. So I observed how that was a shame -- because if that was his final position, then I would not be able to twice mention his name, Richard Durrett, or point out that Cornell is a great institution of higher education. At this point, Durrett reconsidered.

"Okay, I suppose you could assume a mean life of five months and a standard deviation of one month, which is consistent with the behavior of your other 24 bulbs. So this represents a deviation of 15 standard deviations. The normal distribution has the shape E to the minus x squared. If we extrapolate, we get 10 to the minus 8 for this specific one bulb to survive this long. So the odds of it being pure coincidence would be . . ."

I held my breath.

". . . a hundred million to one."

Now we're talking!

Actually, now we're talking to Lou Floyd, executive director of Paint Research Associates Laboratories in Ypsilanti, Mich. Floyd is one of the world's most knowledgeable scientists on the subject of paint. Floyd, too, immediately raised the coincidence conjecture, but then I raised the Lou-Floyd-is-one-of-the-world's-smartest-humans-as-reported-by-the-Washington-Post conjecture, and he, too, reconsidered.

The organic materials in basic ceiling paint, he said -- "surfactants, defoamers, thickeners, pigment dispersants, latex binders" -- might thermally decompose slowly from the heat of the bulb. If they do, they could draw heat energy from inside the bulb. That could help reduce the maximum heat within the bulb, the same way that boiling water, which is also thermally decomposing, maintains itself at 212 degrees Fahrenheit regardless of how much heat is applied to it. The less heat in the bulb, the less stress on the filament, the longer the bulb might last.

Is this possible?

"Well, it's not impossible," he said.

Good enough for me!

And that's when, sitting in my home office, with the light bulb above my head, I had a bright idea. I felt like a little man in a cartoon. I was going to create a grand experiment in which thousands of readers would screw in two new light bulbs, put a dab of paint on one of them, and report back a few months later.

Alas, the D.C. fire marshal expressed the opinion that paint on a light bulb can pose a fire hazard, and the Post lawyers expressed the preference that I not enlist the newspaper in a plot to incinerate the Washington metropolitan area.

The cartoon bulb went out. The real one didn't. It may never.

Gene Weingarten's e-mail address is weingarten@washpost.com.

Chat with him online Tuesdays at noon at www.washingtonpost.com.