cannot be separated more than ten or twelve inches, and the perineal skin covered by a folded piece of iodoform gauze, or lint soaked in a ten per cent. solution of carbolic acid in castor oil. The vulva is cleaned three times a day by squeezing hot water over it from a sponge, and also each time after the patient urinates, or is catheterized. With such precautions she may urinate in a flat bed-pan from the beginning. On the third day I commence vaginal injections of hot water, substituting a two per cent. solution of carbolic acid as soon as there be found a decided odor or appearance of pus in the discharges. The parts should be inspected in a good light on the third and each succeeding day or two. On the third or fourth day a dose of castor oil or a mild saline is given. On the fifth all silk stitches are removed and the knees liberated from the bandage, but the patient forbidden to separate them. If silkworm-gut stitches have been used, they need not be removed for several days or a week later, unless they are too tight and commence to ulcerate.

Frequently, however, there is more to do than this. The inflammatory reaction may go on to suppuration. If so, after each carbolated douche the labia should be separated, and the edges of the wound washed and touched with a five per cent. solution of carbolic acid or sprinkled with iodoform. If pus come from the vagina a three per cent. solution may be thrown in upon the stitches with a little piston syringe, after having placed a little cotton on the meatus urinarius to protect it. A very small strip of lint dipped in carbolated oil, but squeezed out so that the oil will not get on the urethra, should be laid over the edges of the wound in the vulva, and on the external cutaneous surface. The parts should be thus dressed twice a day except that a three per cent. solution of carbolic acid will be strong enough after the first dressing or two. If the suppuration increases the parts should be dressed every eight hours.

After the stitches are removed the wound should be cleansed with the carbolated water and protected with the carbolated oil or lint three times a day, until suppuration has pretty well ceased, then twice a day. In this way, even when the condition is not favorable, I always get union of the deeper and important tissues by first intention, and usually of the whole. In most cases in which the stitches include much skin, as in Fig. 127, there will be a little suppuration in the fatty tissue about the external cutaneous edges, and occasionally a little about the bruised edges near the hymen, which scarcely ever diminishes the length, but may slightly diminish the thickness of the resulting perineal body.

I have twice introduced a deep stitch to hold granulating perineal surfaces together, but have only produced irritation and increased suppuration, and now content myself with binding the knees and dressing the surface as an open wound. If, however, we have a nurse

who will thoroughly and frequently syringe out the depression or gutter between the wounded surfaces, the granulations may be expected to meet and unite more quickly, and draw the parts in better shape, than without the stitch; but without such attention the inclosed pus decomposes and does harm.

Lacerations into the Rectum.

I have not yet had an opportunity to sew up a laceration opening into the rectum by immediate operation, but consider that the advantages of an immediate operation are greater for such a lesion than for the incomplete variety. As such a laceration usually occurs rapidly, and before much dilatation of the inferior parts, the probabilities are that the amount of bruising will not usually be sufficient to prevent union by first intention. The edges should be trimmed perfectly smooth, the parts drawn together by hooks, and the shape of the tear accurately determined. The rectal mucous membrane is then united accurately by a continuous catgut suture, or a series of silkworm-gut interrupted sutures, which include but little beside the rectal mucous membrane. The remainder of the rent is then united as directed for lacerations not extending into the rectum. It must be borne in mind that no traction is allowable on the rectal stitches; the vaginal and cutaneous must be depended upon for holding the parts together.

The after-treatment is the same as for the lesser lacerations, except that the bowels are kept constipated for four or five days at least, and not disturbed by a laxative unless a rectal pressure is complained of by the patient. The less opium that accomplishes the purpose the better. In finally moving the bowels I prefer to give five or six grains of blue mass, followed, if necessary, in twenty-four hours by a mild saline, so as to give time for a softening of the fæces. If lumps are felt in the rectum they should be broken up against the sacrum by the well-oiled finger introduced along the posterior rectal wall. As rectal tubes or catheters are liable to be directed by the rectal promontory forward against the wound, they should not be used except by the physician. (See Figs. 31 and 54.) The silkworm-gut vaginal stitches should be allowed to remain for two or three weeks, and if not easily accessible without stretching the parts, may be left two or three weeks longer.

Secondary Perineorraphy.

An ideal secondary perineorraphy should be the same as the immediate operation, with the additional preliminary step of cutting out the cicatrices, and denuding the tissues that were exposed at the time of the laceration. That the older methods of restoring the vulvovaginal outlet and forming a new perineal body were unsatisfactory, is attested by the number, complication, and confusion of methods that

have been recommended. The first and fatal fault consisted, and still consists, in treating the perineum as so much plastic tissue to be cut and fitted as a tailor fits a coat. For the sake of simplicity it is also customary to recommend one form or fashion of perineorraphy as the usual operation. It would be much more reasonable to recommend, for the sake of simplicity, one amputation of the leg for all kinds of injuries requiring an amputation, for the leg is a much simpler structure than the perineum.

What is to be Accomplished.

It is not only necessary to remove a cicatricial tissue and unite torn surfaces in performing perineorraphy, but to so unite them that the characteristics of the perineal body will be restored. The recto-vaginal promontory must normally close the pelvic outlet. The V-shape of the edge of the levator ani, the sling shape of the levator vagina, the convergence of the labial tissues at the fourchette, the size and pyramidal shape of the perineal body, and the approximation of the median line attachments of the levator vagina and constrictor cunni to the perineal septum, are all to be restored.

When to Operate.

The operation should be performed as soon as the parts can be brought into a healthy state and the patient's general health will permit, for the longer the delay the greater the reaction and atrophy of tissue, and the less the chance of restoring the contiguous unsupported deeper parts to their normal place and condition. (See Prognosis.)

Methods of Restoring the Perineum when the Rectum is not Opened. From the time when perineorrhaphy meant the denudation and uniting of a narrow strip of labial tissue there has been a long series of operations devised, many of which still survive as useful therapeutic measures. Yet none of them has, or can, become the one ideal operation.

The Median Triangular Operation.

The oldest of the surviving methods of closing the rent is by a triangular denudation. A line is drawn along the edge of the skin external to the laceration from a point on one labium major above the lacerated portion to a corresponding point on the other labium, and two other lines joining the ends of this line to a point in the median line of the posterior vaginal wall above the cicatrix or relaxed portion (Fig. 130). The surface included in these lines is to be denuded. Two denuded triangles are thus formed whose common base (the dotted line) is the median line. They to be brought together

so that the labial angles (11) will meet, and be so stitched by vaginal and cutaneous stitches. By comparing Fig. 130 with Figs. 114 and 115 it will be seen that the denudation corresponds with the appearance of certain median lacerations after labor. This is the ideal operation in median lacerations of the vulva with but little or no extension into the vagina. But as such lacerations seldom require attention after having cicatrized, the operation is seldom to be performed.

The Modified Triangular Operation.

In extending the denuded triangles far enough up the posterior vaginal wall to cover a median laceration extending through the levator vaginæ, it has been found that the traction upon the stitches at the introitus vaginæ prevents primary union between them. In consequence a pus pocket forms at the recto-vaginal promontory and a subsequent depression remains at or in front of the recto-vaginal promontory

In order to obviate this the triangles are made to extend only to the levator vaginæ, and a smaller triangle or notch is denuded on the posterior vaginal wall as far as desirable. Fig. 131 shows the triangle thus modified, Fig. 132, the surface as it appears between the labia. The reason why the raw surface in the secondary operation is narrower than that found immediately after the laceration has occurred, is because the parts are, in the latter case, all drawn apart to an equal degree for inspection, whereas when cicatrization occurs the edges of the shallow and but slightly retracted vaginal portion are drawn over the wounded surface in a proportionately greater extent than those of the many times deeper and strongly retracted vulval portion.

This operation is then the ideal one for median lacerations extending up the posterior vaginal wall. But the proportion of such extended median lacerations requiring a secondary operation is small.

The Bilateral Operation.

Having noticed from a study of the cicatrices that lacerations extending beyond the vaginal entrance assumed a diagonal direction on one or both sides instead of following the median line, W. A. Freund recommended to extend the vaginal triangles or tongues along the cicatrices on either side, and leave the sound vaginal wall about the

median line. As the cicatrix does not always represent the entire extent of the tear, he removes sufficient tissue around it to normally close the vaginal orifice. The resulting raw surfaces have almost the same shape as that which is found immediately after the laceration. Compare Fig. 117 with Fig. 133. The edges of the vaginal denuded strips are first drawn together (Fig. 134), and afterwards the resulting vulval triangles. Even when one of the arms of the Y is almost or entirely wanting in the cicatrix, a short strip must be denuded in order to bring the parts together symmetrically. For instance, if the shorter vaginal strip were entirely gone in Fig. 133, and the edge of the denuded figure were at the dotted line instead, it will readily be seen