I thought you might find the following of interest. This is taken from a paper presented to The British Academy by archaeologist Julian Richards and engineer Mark Whitby.

"Whatever the precise route, the varied nature of the terrain, with undulating topography and potentially wet areas, helps dictate the approach to the development of a method of transportation. In this development a number of ideas were considered including the use of rollers, forming a timber cylinder around the stone and rolling it, Crabbing the stone along using levers and the use of ice as a medium across which the stone could be slid.

The orthodox method using rollers to move the stones was considered but rejected. Subsequent experiments in moving the 10 ton lintel proved that it is a practical system, but has limitations. The direction of the stone is difficult to control on all but the most level ground and the method involves high risk as rollers have to be placed ahead of the moving object. As the load goes up, the system becomes prone to binding as the weight of the whole load will at times bear on only one or two of the rollers due to uneveness either in the rollers or in the ground surface. The latter can be overcome by running the rollers on a flat, possibily timber track, and the former by selecting rollers of a uniform diameter. However, directional control remains an issue, as any roller placed out of true to the track will cause the load to veer off.