[TowerTalk] Why did adding a radial make the input Z of my inverted L go up?

[Jim Smith]

Finally got an inverted L up for 160.  The vertical portion is 30 ft of 
skinny (6.5" face) tower insulated from ground.  The horizontal portion 
is attached to the top of the tower, rises to about 40 ft near the 
middle and then down to about 30 ft at the far end.  At the moment it is 
too long but will take care of that shortly.  Autek RF1 says lowest SWR 
is around 1750 kHz.  The reading actually jumps around between 6.7 to 
7.4, presumably due to BC xmtrs.

For grounding, I only had time to bury a ground plate about 9" x 18" or 
so before the CQ 160 SSB.  No radials.  Today I measured the Z between 
the base of the tower and the ground plate.  Soil is very wet and has 
been for months.  The coax centre conductor is disconnected from the 
tower base.  The coax shield is still attached to the ground plate so I 
guess that would count as 1 radial.  The Autek RF1 Analyst says 146 
ohms.  This number doesn't jump around like the SWR.  I added 1 radial 
(#14 insulated wire) buried about 2" and 12 ft long.  Z is now 129 
ohms.  Well, I didn't expect a miracle.  Similarly buried a 42 ft radial 
at right angles to the first and hoped to see Z drop below 100 ohms (no 
scientific reason for this, just hope).  This radial is parallel to the 
coax, about 4 ft away and 2 ft lower in elevation.  (The coax runs along 
a raised garden bed).  RF1 says Z is now 161 ohms.  This is worse than 
the 146 ohms with no radials at all!!

Does anyone have any idea how this could be?

TIA for your assistance in helping me to understand this most puzzling 
result.

73 de Jim Smith   VE7FO