I learned a lot while solving this problem. Thanks however to Ian for the offer, which is much appreciated.
I carefully inspected my camera, a wooden field type, for loose fitting items, bad alignment and worn seals and put black tape and black paint here and there for a totally light-tight job. I tested that with a small hand-torch placed inside the bellows and went into a darkroom to look for any escaping light. It surprised me just how much there was when I started.
Then I reviewed and practiced my method of loading film into holders, placing holders into the camera back and removing and reinserting darkslides without nudging anything open which shouldn't be. I checked my method of loading film into spirals and handling the developing tank.
At one stage, I developed an unexposed sheet, like Henri Becquerel, to see if I had a bad batch, which reproduced the error, thus eliminating the camera and holders. But it wasn't the film, because I tried the same with a sheet from a different batch and got the same result. It turned out to be the developing tank, which I had bought second hand, since my existing tank wasn't big enough. It had a very fine fissure in the sidewall, which also explained the greater than usual shrinkage in the quantities of developer and fixer I was getting.
This also taught me that we see what we are looking for. I saw a tank, and assumed it was a good one. I bought a brand new tank, which I have since tested.
Now I can get on with taking photographs of the very fine scenery in and around Madrid, and other things.
Bill Dixon