pelvis. The tissue lining the pelvis has been regarded as having little influence, although in German text-books the pressure of the tissue lining the posterior wall of the pelvis has always been held to be a factor in determining it. Now, mere change of shape in the bony pelvis will not cause rotation; the only result can be a distortion of the head, i.e., the head will mould itself to the varying capacity of the bony pelvic walls, but not necessarily rotate. This holds even more strongly of the breech. Another very misleading term has been posterior rotation of the occiput in head cases, or chin in face presentations. As I hope to show, no part in the pelvis ever rotates primarily backwards. Rotation is always a forward and inward movement.

2. THE ANATOMY OF THE PARTS CONCERNED IN ROTATION. This must be considered in relation to—

(a.) The Canalized Pelvic and Pelvic Floor.

(b.) The Fatal Body, especially the Head.

(a.) The Canalized Pelvis and Pelvic Floor.-This may be considered as made up of two parts, one in front of the transverse of the brim and true pelvis, the other behind it. The first is the anterior half of the pelvis and tissues, the second is the posterior half with its tissues.

During the second stage of labour the anterior half of the canalized pelvis is practically the same as that of the bony pelvis, inasmuch as the bladder and retropubic fat are partially drawn out of the pelvis, and it is lined symmetrically with vaginal and cervical tissue.

The posterior pelvic wall, chiefly sacrum and coccyx, is different, however, inasmuch as it has springing from it a segment of the pelvic floor, the sacral segment, attached to the sacrum, coccyx, and edges of the sciatic notches.

This sacral segment is the most important factor in bringing about rotation, and as such must be fully considered.

It is fixed above to the sacrum and coccyx, extends downwards from the bony outlet of the pelvis posterior to the ischial tuberosities, and is unattached at the lower end. It thus includes the posterior vaginal wall and tissues behind it, and has imbedded in it parts of the obturator internus, coccygeus and levator ani muscles, as well as the transversus perinei. Part of the strong gluteus maximus also enters into its formation (Figs. 1 and 2, Plate II.)

From its attachments it can be pulled back and made to recoil, e.g., by Sims' speculum, or driven back by the part of the foetus engaging in the pelvis. It is the strong segment of the pelvic floor, as will be understood by the following measurements:

Length from tip of coccyx,
Thickness at level of cervix uteri,

3 inches.

2 inches.

2 inches.

6 P

The length is of course increased during labour. The sacro-sciatic ligaments, coccygeus, and levator ani muscles all help to restrain undue driving back of the lower portion of the sacrum and coccyx as well as of the sacral segment.

It is important to consider it divided into two portions—a right lateral and left lateral-lying respectively to the right and left of an imaginary vertical mesial line on the lower portion of the sacrum coccyx and posterior vaginal wall (Fig. 2, Plate II.) The anterior pelvic wall may be considered as the concave aspect of a segment of a circle, on part of which the rotating head or body glides.

One specially important point to be kept in mind is that the anterior and posterior walls of the canalizing pelvis are of very unequal lengths, and that consequently the anterior portion of the presenting part at the brim will touch the sacral segment, and be compressed or pushed forward by it very much sooner than the presenting part lying posteriorly. The recoil of the sacral segment is increased as we pass towards its lower end, as it acts like a lever of the first kind (Fig. 1, Plate II.)

b. The Fatal Body, especially the Head.-The only point I wish to note here, is the relation between the head and canalized pelvic floor so far as mere fitting is concerned. One factor in rotation depends on whether or not the head, for instance, passes through the pelvis easily or the reverse. In the production of malrotated head cases this is an important factor. Another way to put this is as follows. The question as to whether the occiput or sinciput will be deeper in the pelvis depends on whether the relation between the head and pelvis is such that the latter accommodates the suboccipito - bregmatic diameter or the suboccipito - frontal. The head is here considered as made up of the occiput and sinciput the one begins behind, the other in front of, a coronal plane drawn at the posterior margin of the anterior fontanelle.

3. A DESCRIPTION OF THE MOVEMENT OF INTERNAL ROTATION IN THE VARIOUS PRESENTATIONS.

I briefly remark, first, that while the path of the head during rotation is a complex one, and is not exactly known, it is sufficient to describe it as a rotation of the head on an imaginary and varying vertical axis, occurring when the head is at a varying part of the true pelvis, usually below the level of the ischial spines, i.e., when the leading part comes within the sphere of the sacral segment.

Internal rotation takes place in head, face, forehead, breech, and spontaneous expulsion cases. The clinical facts as to internal rotation in each of these is briefly as follows:

HEAD. In the L.O.A., the occiput, known by its small fontanelle, lies at the brim, opposite the left ilio-pectineal eminence, becomes driven down, and when it strikes the left lateral half of the sacral segment, i.e., when asymmetric to the sacral segment, is rotated through quarter of a circle to the front and right, so that it lies wholly

or partially within the pubic arch, and as it now lies symmetrically to the sacral segment, rotates no further (a, Fig. 1, Plate I.)

In the L.O.A. the occiput rotates to the front and right, i.e., the part deep in the pelvis, and first striking the left half of the sacral segment, becomes rotated to the right and front.

In the R.O.A. the occiput rotates to the front and left through quarter of a circle, i.e., in the R.O.A., the part deep in the pelvis, and first striking the right lateral half of the sacral segment, is rotated to the front and left (Fig. 1, b, Plate I.)

In the R.O.P. we may have two rotations-(a.) The normal or long, and (b.) The abnormal or short.

(a.) The normal or long takes place so that the occiput passes from the region of the right sacro-iliac synchondrosis to the pubic arch. Clinically, the occiput is felt leading, and the occiput fits the pelvis accurately-the head is never a small one. In the R.O.P., therefore, when the occiput leads and strikes the right half of the sacral segment before the sinciput touches the left lateral half, it is rotated through threequarters of a circle to the front and left (Fig. 1, e, Plate I.) (b.) The abnormal or short rotation has been greatly misunderstood, owing to erroneous terminology. In R.O.P. cases where the head is small, the sinciput is deep in the pelvis, consequently strikes the left half of the sacral segment first, and is rotated by it to the front and right through quarter of a circle. Thus the occiput lies in the hollow of the sacrum, and no further rotation of the head takes place, inasmuch as the head is now symmetrically placed to the whole sacral segment (Fig. 1, g, Plate I.) The malrotated occipito-posterior cases are really rotations forward of the sinciput, occur in cases where the head is small, and can easily be predicted during labour by noticing the deep position of the sinciput, i.e., deep position of anterior fontanelle. They are best termed, therefore, after rotation, sinciput to pubis cases.

I need not detail the L.O.P., as we have either (a.) long rotation of the occiput from left to right (Fig. 1, ƒ, Plate I.), or (b.) short rotation of sinciput from right to left (Fig. 1, d, Plate I.)

The rotation of the shoulders is simple. In the L.O.A. the right shoulder is anterior, strikes the right half of the sacral segment first, and is rotated by the pressure on its anterior aspect, to the front and left, through quarter a circle. The head is therefore externally rotated to the left (restitution).

In the R.O.A. the left shoulder is anterior, and is, for reasons already given, rotated to the front and right, causing external rotation to the right (restitution).

In the L.O.P. and R.O.P. either shoulder may strike a lateral half of the sacral segment, and thus rotation of the shoulders may occur to right and front or left and front in each (v. Plate I., Fig. 1).

Internal Rotation in Face Cases.-We name face cases according to the position of the chin (mentum) as L.M.A., R.M.A., R.M.P.,