Every physical thing has an infinite number of natural frequencies. As others have said, this depends not only on the material (specifically, the mass and the stiffness of the material) but also the physical shape of the object.
You can show the effect of shape by holding a stiff metal ruler so it overhangs the edge of a desk. Twang it, and it will make a noise. Now move it so it doesn’t overhang so much (the “shape” changed–it’s shorter) and twang it again. Higher pitch, higher frequency.
But why do they have natural frequencies? Imagine a brick bouncing up and down on a spring (Java applet simulation). It takes time for the spring to stretch and slow down the brick–time for the energy transfer, in other words. The brick is mass, the spring is stiffness, and time is frequency. Change the mass or stiffness, and the frequency (time) changes.
Fairly simple objects–a string, a rod, a bowl-- have relatively simple natural freqencies that can be calculated. These frequencies are related to each other, and get progressively higher. Here’s a graphical picture of the first few frequencies of a vibrating string. When someone talks about “the” natural frequency, they’re typically talking about the first, or fundamental natural frequency (remember the ruler over the edge of a desk?). This is physically menaingful because the fundamental frequency is usually the “strongest”–exhibiting the most energy and highest amplitudes (cool video [scroll down] of the fundamental frequency of a wine glass).
If you want to demonstrate fundamental frequencies and overtones to your nephews, head to the back yard and grab a garden hose. With one person on each end, swing it back-and-forth to make a single loop. That’s the fundamental. Swing it faster, and you can make two loops, with a stationary point in the center. Faster yet, three loops. Or, if you’ve got a Slinky, stretch that out on the floor between two people and shuck it back-and-forth.
More complicated, three-dimensional, objects will have mode shapes that are very different-- swinging back-and-forth in all directions. Here’s a set of animations of a manifold system. In these complicated systems, saying “the” natural frequency is less meaningful, because there may be little relationship between frequencies with different mode shapes. Very large, complex structures will have multiple natural frequncies, and it’s hard to point to which one might be the “worst”–the infamous Tacoma Narrows Bridge being an example (download a a movie thereof here).
And one last thing–materials also have some internal damping that tends to reduce the amplitudes of resonance. Materials like glass and steel have low internal damping, so they will “ring,” or continue to oscillate well after you tump them. Other materials will have a higher internal damping, and will tend to “thud” when you thump them. Two items made out of materials with the same mass and stiffness but different damping might have very similar natural frequencies (not exactly the same, but close enough), but you would perceive them as different because one would ring longer than the other.
-zut, whose PhD thesis included a fair amount of calculation of natural frequencies.