At Berkeley in the 70's we used a giant sand box on inner tubes to stabilize a holography setup that had to keep the vibration of the mirrors down to just a few wavelengths of the HeNe laser's output frequency for something like 13 seconds at a time in order to get a good exposure. The added challenge was that the laser setup was on the third floor of the building. One would set up a Michelson interferometer on the table and project the interference pattern on the wall, and then count how long the average period was before someone on the third floor would walk down the hall, close a door, or sit down at their desk and disturb the pattern. I thought it was pretty amazing that the thing worked so well. IIRC the recipe was a 4'x8' sand filled table with inner tubes on top, with another sand filled 4'x8' box on top of that. I don't recall the top of the top box floating on the sand, I think it was attached to the sides of the box. Might have been a piece of slate from a pool table, or my old brain may just be embellishing...