LarryS said:
Assume that our LC circuit under discussion consists of a parallel plate capacitor, a coil inductor and a battery all in series and that, obviously, it is running at its resonant frequency.
Now I’m assuming that the frequency of any EM radiation generated (free space), near or far, will be the resonant frequency of the circuit.
There is no such thing as a lossless LC resonance. So your statement that "obviously, it is running at its resonant frequency" isn't quite right. Transients, which you didn't specify, will die out. Granted, broadband transients, like a step function will excite a greater response at resonance, which will appear to die out slower. So, it's only running at its resonant frequency if it's driven at that frequency. An antenna absolutely can radiate when driven off resonance, in fact that's a common scenario. And yes, the frequency of the radiated waves will match the frequency of the electron motion in the circuit.
LarryS said:
If that is the case, then the matching free space wavelength for that frequency can be calculated from the geometry of the two components alone and is independent of the length of the wires connecting them.
As
@tech99 said, it's not the geometry of the components, it's the whole circuit, which is mostly wires. One (kind of sloppy) way to think of this is that the L & C components determine how the circuit works, and the wires between them act as the antenna. The resonant frequency of the circuit may not match the most effective antenna size for launching the radiation. I could make a very small circuit that resonates at, let's say, 30MHz but have really low radiation because my antenna, the wires connected to the LC tank, is way too small. This would kind of confirm your suspicion that the wavelength must be comparable to the size of the antenna. 1/4 wave dipoles can work OK, 1/128 wave dipoles don't.
At high frequencies there is the added complexity that the wires themselves have, or are, the L & C, you can have resonant circuits that don't really have isolated "components".
Also, I'm not at all sure about the relationship between the free space wavelength and the geometry of the antenna. My intuition says that the area near/at the antenna doesn't count as free space propagation. But, honestly, EM fields right near antennas are not my specialty, it's pretty complicated and depends on lots of specific details. I'm sure others can discuss this better. Anyway, I wouldn't assume that the wavelength doesn't change or that it's even well defined in that region. Free space propagation is a really useful simplistic benchmark case, it's great for understanding the fundamentals, and a really good model for talking to a spacecraft going to Jupiter, but it isn't always the case in real life.