These secondary motions are of vast importance.  Beyond the limit of normal motion lies danger of lesion; and lesions almost invariably occur through exaggeration of secondary motions.  They are also important in the correction of lesions; for it is necessary to carry a joint to the limit of its normal motion to get any tension upon it.  The character of the tension produced depends on this secondary motion.

    Extensive laboratory work needs to be done on this subject.  A beginning of work of this character was made by the classes of the A. S. O., to be described later.

    In the spine as a whole, extension and flexion* occurs in the lumbar region, rotation in the dorsal, both in the cervical, except in the axis, in which rotation only occurs.
 

*NOTE:  Extension and flexion are loosely defined.  Accurate definition is needed for this work.  As here used, flexion means bending toward the umbilicus, or toward lines drawn vertically and horizontally through it; extension means ;moving away from the umbilicus or vertical and horizontal lines drawn through it.  Thus, flexion of the spine involves opening out or extension of individual vertebral joints; but to avoid confusion, such motion will be referred to as flexion of the individual joint, the umbilicus being regarded as the basis of comparison.  The ribs are extended in inspiration, flexed in expiration.

 
    Side-bending occurs in all vertebrae except the axis.
 
    The work of the students in the classes of January and June, 1917, of the A. S. O., referred to, was on the subject of axes of rotation.  That of Messrs. (now Doctors) Fish and Lawrence was published in the A. S. O. Journal in 1916.  The same discovery was made by each of these gentlemen working independently of the other, and both of them deserve credit for the discovery.  The work of one student, Mr. (now Dr.) Louis E. Browne, alone, was on the subject of axes of flexion and extension, and the credit for the verification of those axes belongs to him alone.

    Mr. Browne’s work showed that the axes of flexion and extension in the dorsal region coincided closely with the position of the head of the ribs as they articulate with the facets on the bodies of the vertebrae; the axes in the lumbar vertebrae and the lower two or three dorsal vertebrae were in the same relative position, coinciding with the position of the semi-solid core of the intervertebral discs in those regions.  This theory had been advanced in class; but “he who proves discovers,” and Mr. Browne’s was the first work aiming to prove this point.

    The work of the other two students (Fish and Lawrence) came, however, as a complete surprise.  Their work dealt with the axes of rotation; and proved that the axes of rotation of lumbar vertebrae lay at varying distances behind the articular surfaces, in some even behind the tips of the spines of the vertebrae.  When a profile drawing of the spine was made and these centers were placed at the proper distances behind, a further discovery was made; that a line drawn through all of these centers formed a continuous curve.  The center of rotation of the fifth lumbar was about three inches behind the tip of th spine; and the line beginning there swept forward to opposite the second and third vertebrae, then backward again until opposite the eleventh dorsal it was practically at infinity; in other words there was no rotation in the eleventh dorsal spine—the surfaces lay in the same plane; there was slight rotation in the twelfth (inferior articular surface of the twelfth with the superior articular surface of the first lumbar); most  and sharpest  rotation between the second and third lumbar, less again as we approach the sacrum.  Doubtless these facts conform to some natural law oaf mechanics.  We find also that the spine of the second lumbar is the largest, usually, corresponding with this fact of greater rotation and also being the point where the lines of tensions from iliac crests to ribs cross each other.

    The centers of rotation of dorsal vertebrae, on the other hand, were shown by these drawings to be in front.

    The center of rotation of the eleventh dorsal was neither in front nor behind; those of the tenth dorsal and of all other dorsal vertebrae were, however, in front of the articular surfaces.  Here again when these centers were placed on a profile drawing of the spine, and a line was drawn joining them, it described a continuous curve, farthest away in the lower dorsals, nearest in the mid dorsals, farther away again in the upper dorsals, until with the first  dorsal or seventh cervical it again reached infinity.

    At this point the centers of rotation again jumped across to the rear of the spinal column and again described a continuous curve, convex toward the spine as before.

    This work was subsequently verified by the other students in the classes, though there was much variation in the results.  For instance it was found that the centers of rotation for the upper dorsals, determined with the most careful work, were very irregular in many spines.  This means merely that the amount of motion there is so slight that nature did not develop the articular surfaces in accurate conformation with the dynamic or mechanic law, whatever it is, which led to the fairly accurate relations of surfaces in other parts of the spine.  It also implies the opposite, namely, that articular surfaces are developed normally in automatic adjustment to the mechanical laws applying to their motion.

    Variations were found also in the lumbar region.  In some the articular surfaces of even the fifth lumbar vertebra faced each other so sharply as to throw the center of rotation near to or in the tip of the spine of that vertebra (which means that gross lesion might exist here with no evidence thereof in the relation of the spines); in some they were not curves at all, with no point of rotation, but lay at an angle—almost at right angle—with each other, indicating that there was no rotation here, that the only motion possible in these joints was that of bilateral or unilateral flexion and extension (the latter meaning side-bending).  In many the two articular surfaces on either side of the same vertebra did not have the same centre of rotation, but that of each was centered around a point lying in or directly posterior to the opposite surface—each surface determined the center of rotation of the opposite surface.  This fact probably indicates that rotation was not a proper function in those vertebrae, but was accomplished by a primary side-bending and secondary rotation.

    In some spines the centers for the lower three or four were found to lie behind, those of the upper three or four to lie in front; but again at such relative distances as to form continuous curves, as in the dorsal and lumbar regions.

    In the work done by Mr. Fish, for instance, it was found that in the spine he examined the centers of rotation for the facets in the lumbar region did not coincide on the two sides, being farthest apart in the fifth lumbar (where the articular processes are farthest apart) and approaching each other evenly until they coincided opposite the tenth or eleventh dorsal.  These were centers for rotation; but it was the side-bending involved in rotation that gave these different centers, as will be shown later.

    The centers for side-bending or tilting to right or left were not determined.  The center probably varies with each instant of motion, being first the center of the body of the vertebra, then its inferior surface, then moving to the point of greatest resistance, which appears to be usually the articulation of the convex side.  Side-bending occurs as said to a slight degree in all vertebrae, abut apparently chiefly in the lumbar region.

    In other words, in the lumbar region the lateral motion is usually probably primarily side-bending, secondarily rotation; whereas in the dorsal region it is probably primarily rotation, secondarily side-bending.  The distinction is merely theoretical, for they both occur together, and one hardly occurs without the other.  The intervertebral disc in the dorsal region is, however, not thick enough to allow much side-bending.  The angle of motion in the dorsal region, in its various parts, may be quickly learned by taking each vertebra in turn and sighting with the eye so that the articular surfaces of the interior costal facet and of the inferior articular process will both lie in the same plane.

    In the dorsal region and the ribs the smallness of the articular surfaces indicates that the amount of actual motion is very slight.  Articular surfaces in the spine never move over the full range of their surfaces; they probably rarely even approach each other’s limits, or even move so that half of either surface is uncovered.  One articular surface is always larger than the other; and the actual range of normal motion is probably comprised within the larger of the two—limited to the distance that the smaller can slide around without leaving the surface of the larger.  This for the reason that articular surface forms automatically when bone slides on bone, and if the range of motion were exceeded more articular surface would tend to form.  The lightness of the motion in the dorsal region is still further indicated in the smallness of the costal facets on the vertebral bodies, and in the flatness of the superior and inferior surfaces of the vertebral bodies with the thinness of the intervertebral discs.  A mere cartilaginous yielding distributed through the whole spine amounts to a very large total of motion.

    The proper motion of the ribs is practically limited to that as seen ;In respiration, which, variously adjusted and combined, can allow all of the actual motion that is observed.  Examining the facets on the transverse processes we find that they are in many if not most spines not facts at all, but deep grooves, covering in some cases the half of a circle.  The only proper motion possible here is a turning motion—a sliding in and out being prevented by the articulation of the head of the bone with the bodies of the vertebrae.  The head of the bone is carried with one or the other of the two vertebrae that it touches (the one body in the case of the twelfth and eleventh ribs) and slides on the facet of the other; but this motion is so slight at the facet on the transverse process as to cause little more than an elastic yielding.  “Bucket-bail” motions of ribs do not seem to be possible, judging by the anatomical conformation of the facets.

    Some work done by Mr. (now Dr.) Schoonmaker in the A. S. O. shows that the facets on the transverse processes of vertebrae also vary in continuous and graded series.  In the vertical plane they face toward a point in front of the lower part of the chest; in the transverse plane they face directly forward in the first dorsal and turn gradually to face forty-five degrees out in the tenth dorsal.  The axis of turning (flexion and extension) is of course the center of each rib itself.  These facets represent the direction in which normal pressure is brought to bear on the articulations, from muscular action and from atmospheric pressure without, and from blowing, etc., within the chest; for of course articular surfaces must be directly perpendicular to the pressure that bears against them, or they would slide to the limit of their motion and stay there.

    The proper motions of cervical vertebrae are direct extension and flexion in the median line, and unilateral extension and flexion.  Experiments made on each other by students in the A. S. O. classes referred to showed that here also one side moved at a time.  In the first stages of this unilateral extension, the concave side is fully extended or approximated before the opposite side begins to move, but the limit of approximation is soon reached, and then the opposite side begins to open out, or flex, while the approximated side remains stationary with only the slight turning on its own axis that is necessary.  Each side becomes the center of rotation for the other side.  This fact is not expressed in the articular surfaces for the reason that they must be smooth to allow the other motions, that require gliding in other directions.

    The proper motion of the atlas on the axis is a pure rotation around the odontoid process.

    The proper motion of the occiput on the atlas is extension-flexion, or nodding; with slight side-bending and slighter rotation.  The atlanto-axial is almost a ball-and-socket joint with very slight ranges of motion.