Instead, you need "non-destructive testing" -- the proverbial better idea. One non-destructive technique, X-ray fluorescence, shares much with neutron activation analysis. You zap the object with energy and the radiation is characteristic of the elements in the sample.

But while neutron activation measures gamma rays, X-ray fluoresence measures X-rays. More important, X-ray fluorescence can operate outside a reactor. And it works on a small area, so if the irradiated section of a painting does change color, the problem is small.

X-ray fluorescence comes in several flavors. Generally, the object is bathed in high-energy X-rays. A second approach, featuring a beam of high-energy protons, is being used at the Louvre in Paris, and elsewhere.

The principle of this particle-induced X-ray emission is simple: create high-energy protons in an accelerator, whack them against the art object, and measure the X-rays produced when the protons pass through the electron clouds in the object. (Technically, that's "excitation," not "whacking," but you get the point).

As with neutron activation, the energy levels are the tip-off. "The energy of the X-ray is very well-known for the elements," says Gregory Norton, vice-president for marketing at National Electrostatics, which makes the proton accelerators.

He says the ease of detecting elements increases with weight. Since only 10 elements are lighter than sodium, the lightest detectable element, most of the periodic table is fair game. He says the particle-generated method penetrates deeper than traditional X-ray fluoresence, which reads only the surface atoms.

While the use of modern materials in a supposedly antique artwork is one way to detect forgery, you can also compare the results to data from a known sample. Did the questioned painting use the same pigments as the real one? Do inconsistencies indicate massive refurbishing after the painted was completed?