Steve,

I have some comments as to "where does the current go"?  Comments are below:

Today, I made some more antenna current measurements on my 12 m
vertical (and my shiny new loading coil), using a thermocouple
meter.

At resonance, the current into the base of the loading coil, was
2.2 A.  But when measuring the current into the antenna at the
hot end of the coil, the meter only indicated 1.8 A.  I had
expected the reading to be the same at both ends of the coil.

The loading coil has many more turns than is required to resonate
the 12 m vertical, so the connection to the antenna uses a tap
which is several turns down from the top of the coil.  I've been
thinking that the excess turns at the top of the coil may be
causing the difference in current readings.  (This may sound
silly, but in my mind's eye, I can sort of picture some of the
current in the coil going into the capacitance of the antenna;
and some current going into the capacitance of the unused turns.)

Before I try removing the excess turns, I'd like to know whether
the current should indeed be the same at both ends of the loading
coil.  Can anyone help?

Self-capacitance between turns (each one to all others) involves a myriad of
situations that has an equivalent net capacitance.  Also there is a myriad
of capacitance situations to "ground" potential.  It is not possible to have
a loading coil with no self or stray capacitance.

I agree with the comment from Andy that it is undesirable to have "unused
turns" at the "hot" end of the loading coil.  It is generally bad practice
to have any uneccesary capacitance to ground in the vertical "up wire",
which is why commercial LF NDB antennas with multiple wire top loading taper
to a single feedpoint, using bare wire (no dielectric coating that
needlessly adds to self capacitance).

The current meter reading at the "cold" side of the loading coil can be
regarded as supplying current for antenna radiation as well as for parasitic
losses in the loading coil.  I am assuming that these currents are in the
form of quadrature components (the antenna has significant far field
radiation, the loading coil does not).  To reconcile your readings of the
"cold end" ammeter reading showing a vector resultant of 2.2 A, and a single
current of 1.8 A going up the wire from the "hot end", doing a vector
calculation shows that the current components that satisfy your observed
meter readings are 1.8 +j1.26 amps for the resultant "cold end" current of
2.2 amps.  The calculation of the quadrature (j) current is fairly sensitive
to error in ammeter readings, but even with the figures advised, and if the
explanation is reasonable, the current needed to furnish the self and stray
capacitance in a loading coil is rather more than I had ever imagined before
doing the sums.  However, it is substantially "reactive current" so it is
not a big factor in the overall loss budget of the antenna.

Further investigation of the situation could be done with a dual channel
oscilloscope, low capacitance probes, and a low level signal generator (so
as to not fry a scope probe) to observe any phase difference between
voltages at the "cold" and "hot" ends of the loading coil for a known tuned
antenna.  However, the source impedance also has an influence (practical
transmitters present much lower impedances than 50 ohm test equipment) on
the relative phasing of loading coil currents and voltages so it is
preferable to test with a step down transformer to say 5 ohms if a 50 ohm
signal generator is used..

This radio stuff is fascinating as to the ways in which RF currents can
disappear into thin air :o)

73, Bob ZL2CA