Second, the knowledge and skill of the operator and the perfection of technique undoubtedly influence surgical statistics. The ordinary practitioner or surgeon cannot hope for as favorable results in many operations on cancer as can those who are past masters in this line, and these latter are the ones who furnish the favorable statistics. Third, the class of cases operated on affects surgical statistics very greatly. While epithelioma of the face, and even of the lip, when well removed, may yield most favorable statistics, cancer of the breast, uterus, stomach, intestines, gall bladder, etc., yield increasingly unfavorable statistics, as will be presently seen. Fourth, the duration of observation after operation affects very seriously the validity of statistics. Not long ago three years’ freedom from disease was considered the time to regard a cancer as permanently cured; but this time has been lengthened more and more, by the observation of any number of cases where the disease has recurred even long afterwards, and reliable observers are now very chary in expressing an opinion as to the final cure of a cancer. This will be more fully considered in another lecture. Finally, the optimism of the reporter seems often to have something to do with the reliability of surgical statistics. This need hardly be discussed. The older and more experienced the surgeon the less confident he is of having actually cured cancer with the knife. At a discussion in the New York Academy of Medicine, some years ago, Dr. Robert F. Weir said that the late Dr. Agnew, a celebrated surgeon of Philadelphia, had remarked, just before his death, that he doubted if he had ever been justified in an operation upon cancer, and he, Dr. Weir, stated that he could almost say the same. Turning now to the actual statistics of operative surgery on cancer, we will find that the percentage of reported cures varies very greatly, in accordance with the points just stated. It is understood, of course, that no accurate statements can be made from statistics in reference to the actual mortality of cancer in any location, partly owing to the paucity of figures, and partly because the stages and extent of the disease differ so greatly, and the results vary with the previous duration of the lesion and the period of observation after the operation.