As already pointed out, labor commonly begins with transient discomfort in the lower part of the back. At first the uterine contractions are far apart; they last but a moment and cause only twinges of pain. Gradually, the preliminary contractions give place to others of more definite character, which appear at intervals of five to ten minutes. Estimates of the total length of labor will vary according as one counts from the first warning or from the advent of typical contractions which we hear called "pains of the right kind." These generally continue for about four hours, and this period represents the average length of time the physician remains constantly with his patient. Estimates which include the initial symptoms are longer, varying from ten to eighteen hours. Prolonged labors are rare; and extremely short labors are also infrequent, though now and again it will be only an hour or two from the very first pain until the child is born. To predict absolutely the length of labor for any particular patient is impossible. The averages calculated from large groups of cases have no more than a broad scientific interest; when applied to any individual they are apt to be very misleading. Thus, from statistics we should expect the first labor to be longer than subsequent ones, but we are often surprised by an unusually rapid delivery. To facilitate description, labor is divided into stages which are conveniently designated the first, the second, and the third. During the first stage the way is prepared for the expulsion of the child; at the end of the second stage the child is born; the third stage is occupied with the separation and the expulsion of the after-birth. The progress of labor may be ascertained from time to time by means of suitable examinations. Whereas formerly vaginal examination was the only method which served this purpose, we are now acquainted with several. For example much of the information necessary for the proper management of delivery may be gained from examination of the patient's abdomen; and this may be supplemented by observations too technical to consider here. Occasionally I have heard doctors accused of negligence because they failed to make numerous vaginal examinations. Censure of this kind generally is unjust, for discretion in limiting the number of vaginal examinations provides against infection a guarantee which cannot be overestimated. In many cases, of course, they are still invaluable toward determining what treatment should be pursued, yet they are never employed to the extent once customary. Moreover, physicians have learned to take extraordinary precautions whenever vaginal examinations must be made.