Gordon, your orginal idea for the logs was to have a ripping taken off each side to make them of a uniform thickness and to give them a greater bearing surface. But, as you are aware, this is not likely to be historically accurate. As an alternative and because of the need to stabilise the tower you suggested that we notch out the logs so that they interlock with each other. This would be very time-consuming work and since the logs are likely to be of varying diameters, each joint would have to be tailored to the log it was mating with. This requires that the timbers are all sorted and numbered so they can be assembled in the correct sequence.

Suppose we just use raw logs and accommodate the inherent taper by laying them head to tail to maintain a level platform. We can easily make a plumb-line level out of primitive materials and use it to check that level is being maintained. If we find that one corner is a bit lower than another, we put the fat end of the logs to that corner for a layer or two. The weight of the stone would cause the surface of the logs to flatten where they meet and form a natural interlock. Raw logs would be much easier to insert into the crib than notched ones. If one or two of the logs below had twised under load, the notches might fail to engage properly and we'd be faced with making on site adjustments at a most critical time. We can still use birdmouth noggings as I suggested previously for the layers where the shoring props are attached. These are only required to take a compressive load, so they are self-retaining.

Speaking of shores, when I was talking about stopping the stone when it reaches vertical, the structure I had in mind is called a "flying shore". With a stop-log lashed to the top of the block, the shore would convert the rotational momentum into linear momentum in an upwards vertical direction. This would attempt to lift the stone back out of the ground, so we are using the stone's own weight to resist its momentum rather than brute force, which tends to smash things.