An investigation was made of 1,001 uncomplicated full-term parturitions occurring in 1,621 obstetric admissions. Of these, 40 per cent enjoyed intact membranes until complete dilatation and were used as controls. Thirty-one per cent spontaneously ruptured the membranes before the onset of regular pains at twenty-minute intervals, and were classed as dry labors. The remaining 29 per cent was made up of those whose membranes were intact only to a dilatation of from 4 to 7 cm., 11 per cent by spontaneous rupture and 18 per cent by deliberate rupture.No light was thrown on the etiology of rupture of the membranes, although it occurred disproportionately more often in primiparas. Occipitoposterior position was not a factor.The length of labor was studied from four points of view: There was a larger percentage of prolonged labors in the control group than in the dry and partially dry groups. The cumulative distribution curves of percentage completed labors at two-hour intervals, was most favorable for completely dry labors, least favorable for the “wet” labors, the partially dry labors falling between. In the partially dry labors there was a positive rather than a negative correlation between the length of the “wet” portion of the labor and the total length of labor. The statistically significant averages of the length of labor showed dry labor to be shorter than the controls by 2.0 (±0.25) hours for primiparas and 2.4 (±0.24) hours in multiparas.Concenrning the safety of labor in the absence of the membranes: The incidence of forceps intervention was 16.3 per cent in the control group, 10.6 per cent in the dry labor group, and 12.6 per cent in the partially dry labor group. The morbidity was essentially the same in all groups, i.e., 6.2 per cent in the controls, 5.5 per cent in the dry labor cases, and 6.7 per cent in the partially dry labors. There was a slightly lower incidence of cervical lacerations in the dry labor group than in the controls, with the partially dry labors falling between. The fetal mortality was unaffected by rupture of the membranes either spontaneously or deliberately.The literature was exhaustively reviewed and reports in the last fifteen years amply confirm the above findings.These results offer support to the explanation of cervical dilatation suggested by Dewees, that in labor the cervix is gradually retracted over the head by muscular action alone. On the other hand, that dilatation is accomplished by the hydrostatic wedge, a theory which has been disputed repeatedly since the eighteenth century, appears to be incompatible with the results of this experiment, inasmuch as the membranes proved to be unnecessary for a safe, easy, and short labor.