Richard,

Wow. You spent a lot of time on this. Thanks.

Let's abandon antennas and postulate a twin-lead vacuum-dielectric 100
ohm transmission line that is ideal in the respects we consider
important (except loss). There may be (resistive) loss, but no
coupling to outside sources or objects, no hysteresis, no
electromagnetic radiation. The characteristic impedance is exactly 100
ohms real over our frequency of interest. The velocity factor, for
convenience, is 0.5.

We feed it on one end with a sinewave generator whose impedance is
matched to the line. On the other end we terminate the line with a
floating load of arbitrary impedance. The only requirement of the load
is that it be perfectly linear and can be described completely as a
real and imaginary impedance at any given frequency of interest.

Would you consider this system linear? We are not talking antennas
now...nothing is radiating.

I'm not sure I'm going to go anywhere with this, and I'm not setting a
trap. I'm just curious what conditions would have to be set before you
would consider a transmission line system linear (you are welcome to
add any conditions I might have forgotten). By the way, if you want to
discuss any really subtle effects such a Stokes shifting, I define
linearity as obeying the law of superposition within a reasonable
dynamic range, say 140 dB, which is about 20 dB better than the input
dynamic range of our best HF receivers.

I understand you are in an argumentative mood with others in the group,
but I am taking no sides (I do happen to like Cecil's motorbike...) and
will try to keep things civil. If I go anywhere with this, I hope to
explain it clearly enough and with enough supporting material that
there will be no arguments. I am not a guru so don't expect anyone to
believe anything on my word alone.

So what say you Richard? Do we have linearity?

73,
Glenn AC7ZN