He then tested how much friction there was when a ceramic balland a polyethylene socket were immersed in a liquid with the same properties as synovial fluid.
It turned out that the handle had compressed the eyeball in and pushed it up in the socket (think about squeezing a rubber ball), but it remained attached.