spontaneous evacuation has not already taken place: 1st. When shall the abscess be opened? 2d. Where shall the opening be made? and 3d. How shall the operation be done? {224} The first of these questions has been answered in a general way by the preceding remarks, and it is only necessary to add here, by way of recapitulation, that the time for opening the abscess will depend upon its location and the condition of the patient. If the pus is near the surface and can be easily and safely reached, whether pointing has occurred or not, it is ripe for evacuation and should be liberated at once, even though the patient be in the best possible condition and show no evidence of deleterious effect from its presence. Nothing whatever can be gained by permitting it to open spontaneously, but much may be lost. If, however, the situation of the abscess be such that it would be necessary to traverse healthy tissues to a considerable extent in order to reach it, and the patient shows no evidence of septic absorption, it would be highly injudicious to attempt to open the abscess: first, because under the circumstances you could not be positively certain that a collection of pus existed; and, secondly, because it is doing no harm. Delay, with careful observation, is now the proper course. Within a few days the apparent abscess tumor may either show decided signs that it is diminishing in size and undergoing resolution, or it may approach the surface, so that evacuation will become safe. On the other hand, should symptoms of blood-poisoning develop and the patient show signs of rapid exhaustion, our attitude must be one of action instead of delay. The pus must then be liberated even at some risk. I still insist, however, that a positive diagnosis must be established, and that the operative measure shall be in no sense exploratory.