Fig. 3) confirmed that there was no place of cleavage between the lining of the bursa and the adjacent muscles and tendons, as would have been the case if the cysts had been synovial herniations from the FIG. 3.-Wall of semimembranous bursa. Giemsa. ( A 40.)knee. In no instance was a natural communication with the joint found. It would be easy to establish a false hiatus in endeavouring to ensure a radical excision of the deep surface of the bursa.