15m earth-electrode

Well, because of my stroke and bad weather, a few days ago was the first chance I had to go into the garden to inspect the far grounded electrode. To my great surprise I discovered  my wire was not connected to the ground rod at all and was as my son had left it last September - the wire was just connected to  a rolled up piece of coax laying on the ground, in effect a capacitive ground connection: that was why I needed series inductance then! The feed wire is about 1.5m off the ground along a fence. Since the start of the weekend I have been using the far end grounded (as I thought it was originally) to the ground stake. Impedance is still very high (around 200-300 ohms) but I no longer need the series inductance. As for results, it seems perhaps more directional and (maybe) a few dB down on the  previous (unintentional) arrangement. Whether the antenna was working as an NVIS wire is not clear but at the old QTH running the wire back through the grass made little or no difference to performance of the earth-electrode antenna, confirming the " loop in the ground" behaviour at MF/LF.  I may next try extending the poorly grounded wire further along the fence top to compare performance. The maximum run possible is about 30-35m.