are periodic attacks of pain due to the addition, at each monthly period, to the quantity of retained menstrual fluid. Whenever, from any of these reasons, suspicion of the true character of the case has arisen, a careful vaginal examination will generally clear it up. In most cases the obstruction will be in the vagina, and is at once detected, the vaginal canal above it, as felt per rectum, being greatly distended by fluid; and we may also find the bulging and imperforate hymen protruding through the vulva. The absence of mammary changes, and of ballottement, will materially aid us in forming a diagnosis.

The engorged and enlarged uterus frequently met with in women suffering from uterine disease, might readily be mistaken for an early pregnancy, if it happened to be associated with amenorrhoea. A little time would, of course, soon clear up the point, by showing that progressive increase in size, as in pregnancy, does not take place. This mistake could only be made at an early stage of pregnancy, when a positive diagnosis is never possible. The accompanying symptoms -pain, inability to walk, and tenderness of the uterus on pressure— would prevent such an error.

Ascites, per se, could hardly be mistaken for pregnancy; for the uniform distention and evident fluctuation, the absence of any definite tumor, the site of resonance on percussion changing in accordance with alteration of the position of the woman, and the unchanged cervix and uterus, should be sufficient to clear up any doubt. Pregnancy may, however, exist with ascites, and this combination may be difficult to detect, and might readily be mistaken for ovarian disease associated with ascites. The existence of mammary changes, the presence of the softened cervix, ballottement, and auscultation-provided the sounds were not masked by the surrounding fluid-would afford the best means of diagnosing such a case.

One of the most frequent sources of difficulty is the differential diagnosis of large abdominal tumors, either fibroid or ovarian, or of some enlargements due to malignant disease of the peritoneum or abdominal viscera. The most experienced have been occasionally deceived under such circumstances. As a rule, the presence of menstruation will prevent error, as this generally continues in ovarian disease, while in fibroids it is often excessive. The character of the tumor-the fluctuation in ovarian disease, the hard nodular masses in fibroid-and the history of the case-especially the length of time the tumor has existed-will aid in diagnosis, while the absence of cervical softening (vide p. 143) and of auscultatory phenomena will further be of material value in forming a conclusion. Some of the most difficult cases to diagnose are those in which pregnancy complicates ovarian or fibroid disease. Then the tumor may more or less completely obscure the physical signs of pregnancy. The usual shape of the abdomen will generally be altered considerably, and we may be able to distinguish the gravid uterus, separated from the ovarian tumor by a distinct sulcus, or with the fibroid masses cropping out from its surface. Our chief reliance must then be placed in the alteration of the cervix, and in the auscultatory signs of pregnancy.

Spurious Pregnancy.-The condition most likely to give rise to errors is that very interesting and peculiar state known as spurious pregnancy, or pseudocyesis. In this, most of the usual phenomena of pregnancy are so strangely simulated that accurate diagnosis is often far from easy. There are hardly any of the more apparent symptoms of pregnancy which may not be present in marked cases of this kind. The abdomen may become prominent, the areolæ altered, menstruation arrested, and apparent foetal motions felt; and, unless suspicion is aroused, and a careful physical examination made, both the patient and the practitioner may easily be deceived.

There is no period of the childbearing life in which spurious pregnancy may not be met with, but it is most likely to occur in elderly women about the climacteric period, when it is generally associated with ovarian irritation connected with the change of life; or in younger women, who are either very desirous of finding themselves pregnant, or who, being unmarried, have subjected themselves to the chance of being so. In all cases the mental faculties have much to do with its production, and there is generally either very marked hysteria, or even a condition closely allied to insanity. Spurious pregnancy is by no means confined to the human race. It is well known to occur in many of the lower animals. Harvey related instances in bitches, either after unsuccessful intercourse, or in connection with their being in heat, even when no intercourse had occurred. In such cases the abdomen swelled, and milk appeared in the mammæ. Similar phenomena are also occasionally met with in the cow. In these instances, as in the human female, there is probably some morbid irritation of the ovarian system.

The physical phenomena are often very well marked. The apparent enlargement is sometimes very great, and it seems to be produced by a projection forward of the abdominal contents due to depression of the diaphragm, together with rigidity of the abdominal muscles, and may even closely simulate the uterine tumor on palpation. After the climacteric it is frequently associated, as Gooch pointed out, with an undue deposit of fat in the abdominal walls and omentum, so that there may be even some dulness on percussion, instead of resonance of the intestines. The foetal movements are curiously and exactly simulated, either by involuntary contractions of the abdominal walls, or by the movement of flatus in the intestines. The patient also generally fancies that she suffers from the usual sympathetic disorders of pregnancy, and thus her account of her symptoms will still further tend to mislead.

Not only may the supposed pregnancy continue, but, at what would be the natural term of delivery, all the phenomena of labor may supervene. Many authentic cases are on record in which regular pains came on, and continued to increase in force and frequency until the actual condition was diagnosed. Such mistakes, however, are only likely to happen when the statements of the patient have been received without further inquiry. When once an accurate examination has been made, error is no longer possible.

We shall generally find that some of the phenomena of pregnancy

are absent. Possibly menstruation, more or less irregular, may have continued. Examination per vaginam will at once clear up the case, by showing that the uterus is not enlarged, and that the cervix is unaltered. It may then be very difficult to convince the patient or her friends that her symptoms have misled her, and for this purpose the inhalation of chloroform is of great value. As consciousness is abolished, the semi-voluntary projection of the abdominal muscles is prevented, the large apparent tumor vanishes, and the bystanders can be readily convinced that none exists. As the patient recovers the tumor again appears.

Duration of Pregnancy. The duration of pregnancy in the human female has always formed a fruitful theme for discussion among obstetricians. The reasons which render the point difficult of decision are obvious. As the large majority of cases occur in married women, in whom intercourse occurs frequently, there is no means of knowing the precise period at which conception took place. The only datum which exists for the calculation of the probable date of delivery is the cessation of menstruation. It is quite possible, however, and indeed probable, that conception occurred, in a considerable number of instances, not immediately after the last period, but immediately before the proper epoch for the occurrence of the next. Hence, as the interval between the end of one menstruation and the commencement of the next averages twenty-five days, an error to that extent is always possible. Another source of fallacy is the fact, which has generally been overlooked, that even a single coitus does not fix the date of conception, but only that of insemination. It is well known that in many of the lower animals the fertilization of the ovule does not take place until several days after copulation, the spermatozoa remaining in the interval in a state of active vitality within the genital tract. has been shown by Marion Sims that living spermatozoa exist in the cervical canal in the human female some days after intercourse. It is very probable, therefore, that in the human female, as in the lower animals, a considerable but unknown interval occurs between insemination and actual impregnation, which may render calculations as to the precise duration of pregnancy altogether unreliable.

It

A large mass of statistical observations exist respecting the average duration of gestation, which have been drawn up and collated from numerous sources. It would serve no practical purpose to reprint the voluminous tables on this subject that are contained in obstetrical works. They are based on two principal methods of calculation. First, we have the length of time between the cessation of menstruation and delivery. This is found to vary very considerably, but the largest percentage of deliveries occurs between the 274th and 280th day after the cessation of menstruation, the average day being the 278th; but, in individual instances, very considerable variations both above and below these limits are found to exist. Next we have a series of cases, from various sources, in which only one coitus was believed to have taken place. These are naturally always open to some doubt, but, on the whole, they may be taken as affording tolerably fair grounds for calculation. Here, as in the other mode of calculation, there are

marked variations, the average length of time, as estimated from a considerable collection of cases, being 275 days after the single intercourse. It may, therefore, be taken as certain that there is no definite time which we can calculate on as being the being the proper duration of pregnancy, and, consequently, no method of estimating the probable date of delivery on which we can absolutely rely.

Methods of Predicting the Probable Date of Delivery.-The prediction of the time at which the confinement may be expected is, however, a point of considerable practical importance, and one on which the medical attendant is always consulted. Various methods of making the calculation have been recommended. It has been customary in this country, according to the recommendation of Montgomery, to fix upon ten lunar months, or 280 days, as the probable period of gestation, and, as conception is supposed to occur shortly after the cessation of menstruation, to add this number of days to any day within the first week after the last menstrual period as the most probable period of delivery. As, however, 278 days is found to be the average duration of gestation after the cessation of menstruation, and as the method makes the calculation vary from 281 to 287 days, it is evidently liable to fix too late a date. Naegele's method was to count seven days from the first appearance of the last menstrual period, and then reckon backward three months as the probable date. Thus, if a patient last commenced to menstruate on August 10, counting in this way from August 17 would give May 17 as the probable date of the delivery.

Matthews Duncan has paid more attention than anyone else to the prediction of the date of delivery. His method of calculating is based on the fact of 278 days being the average time between the cessation of menstruation and parturition; and he claims to have had a greater average of success in his predictions than on any other plan. His rule is as follows: "Find the day on which the female ceased to menstruate, or the first day of being what she calls well.' Take that day nine months forward as 275-unless February is included, in which case it is taken as 273-days. To this add three days in the former case, or five if February is in the count, to make up the 278. This 278th day should then be fixed on as the middle of the week, or, to make the prediction more accurate, of the fortnight in which the confinement is likely to occur, by which means allowance is made for the average variation of either excess or deficiency."

Various periodoscopes and tables for facilitating the calculation have been made. The periodoscope of Dr. Tyler Smith is very useful for reference in the consulting-room, giving at a glance a variety of information, such as the probable period of quickening, the dates for the induction of premature labor, etc. The following table, prepared by Dr. Protheroe Smith, is also easily read, and is very serviceable:

1

TABLE FOR CALCULATING THE PERIOD OF UTERO-GESTATION. 1

The date at which the quickening has been perceived is relied on by many practitioners, and still more by patients, in calculating the probable date of delivery, as it is generally supposed to occur at the middle of pregnancy. The great variations, however, of the time at which this phenomenon is first perceived, and the difficulty which is so often experienced of ascertaining its presence with any certainty, render it a very fallacious guide. The only times at which the perception of quickening is likely to prove of any real value are when impregnation has occurred during lactation (when menstruation is normally absent), or when menstruation is so uncertain and irregular that the date of its last appearance cannot be ascertained. As quickening is most commonly felt during the fourth month, more frequently in its first than in its last fortnight, it may thus afford the only guide we can obtain, and that an uncertain one, for predicting the date of delivery.

Is Protraction of Gestation Possible?-From a medico-legal point of view the question of the possible protraction of pregnancy beyond the average time, and of the limits within which such protraction can be admitted, is of very great importance. The law on this point varies considerably in different countries. Thus, in France it is laid down that legitimacy cannot be contested until 300 days have elapsed from the death of the husband, or the latest possible opportunity for sexual intercourse. This limit is also adopted by Austria, while in Prussia it is fixed at 302 days. In England and America no fixed date is admitted, but while 280 days is admitted as the "legitimum tempus pariendi," each case in which legitimacy is questioned is to be decided on its own merits. At the early part of the century the question was much discussed by the leading obstetricians in connection with the celebrated Gardner peerage case, and a considerable difference of opinion existed among them. Since that time many apparently perfectly reliable cases have been recorded, in which

1 The above obstetric "Ready Reckoner" consists of two columns, one of calendar, the other of lunar, months, and may be read as follows: A patient has ceased to menstruate on July 1: her confinement may be of ten lunar months). Another has ceased to menstruate on January 20; her confinement may be expected on September 30, plus twenty days (the end of nine calendar months), at soonest; or on October 7, plus twenty days (the end of ten lunar months), at latest.

latest on April 6 (the expected at soonest about March 31 (the end of nine calendar months); or at