Late cases of tuberculosis of joints, in which destruction has taken place, offer comparatively little difficulty in diagnosis, although not a few of these, so-called, at operation prove to be some other pathologic condition. Early cases, however, by which is meant those in the first year or two of the disease, are exceedingly difficult to diagnose, and are the source of a high percentage of error. The experience in this clinic has led to the belief that in many of these cases a positive diagnosis cannot be made by the ordinary methods so far in general use. For this reason, it has become our routine to do exploratory operations in all doubtful cases. Not only are cases of tuberculosis commonly mistaken for some other condition, such as chronic arthritis or synovitis, but other conditions are considered as tuberculosis and treated as such. It is difficult to decide which error causes