the result was just as striking, and as in the fetus we note the narrowing of the thorax on the convex side and the descent and separation of the ribs.

Summary: So far then, the points to which your attention has been called as showing the forces implied in normal postures are, first, the conception of the segmentation of the body and the shifting of these segments about the line of support to maintain balance, secondly, the fact that the tube of tissues connecting superimposed segments must be subject to a constant change of

FIG. 3. Röntgengram of fetus twisted.

tension, and thirdly, that if we regard the thorax as a segment its diversion must mean that its walls come under strain on account of the peripheral tension of the soft parts, so that the ribs tend to distort, changing their shape and their direction.

A deduction as to the effect of rib strain on the vertebræ: We have proved that the ribs are under strain; consequently, we

may deduce that such strain must be communicated to whatever the ribs are attached. Now the ribs do have an intimate attachment to the posterior part of the dorsal vertebræ, therefore the strain which enters the ribs must be conveyed back into this part of the column, viz., the row of arches. So you can readily comprehend that if the dorsal column is moving primarily with the thorax, its posterior part, the row of arches, must be strained in a direction necessitated by the potential strain in the ribs, which tendency as compared with the primary dorsal movement must be regarded as a separate one of reaction. This of course must apply both in lateral bend and in twist (Fig. 4).

twist

·lateral bend

FIG. 4. Arrows represent direction of force as due to lateral bend or twist (implied in potential strain of rib) with reference to arch of vertebra.

Demonstration of models: Let me next illustrate the normal tendencies of posture as so far deduced, with mechanical models which I have had constructed for the purpose. These models cannot prove anything that I have stated but they can illustrate my views, and, in this fashion, I believe I can make myself clearer as to what has come before.

Model of the thorax: The first model (Fig. 5) represents the thorax as a unity, the dorsal column being regarded as an integral part of the thorax and moving with it. I have, therefore, represented the column by a curved stave of wood and attached to it these leather ribs and sternum, to imitate nature. The parietal muscular system is based upon the conception of Meyer and I have used tapes to carry out the different pulls. The posterior spinal system, the anterior system and the two oblique systems are represented, the transverse system has been left out for the sake of simplicity. These pulls end in the base of the model representing a fixed basic resistance. Movement is enabled by two joints suitably contrived for the purpose. We have then a model simulating the normal thorax with the exception of the elasticity of the ribs which are here purposely made more flexible

FIG. 5. Model of thorax, seen from behind. A, base; B, metal arc; C, block, gliding on arc; D, stave representing the dorsal column; E, setscrew fixing the ball and socket joint of dorsal column in block; F, setscrew fixing block to arc; GG, leather ribs; HH, tapes representing muscles; KK, adjustable leather straps.

than in nature, in order to exaggerate the distorting tendency for purposes of clearer illustration.

Illustrating first side bend (Fig. 6) with the center of motion at the center of the base you will note the tension on the convex side so that the leather ribs descend and separate, thus causing contraction of that side, whereas, on the other side there is no strain and the tapes are relaxed. Such a bend however is not the usual bend in nature, because, on account of the balance of segments around the line of support, the base of the thorax normally swings one way while the apex swings the other. We illustrate this by swinging the model in a similar fashion (Fig. 7). This kind of a bend, however, brings out exactly the same points as the previous position; therefore, whatever the nature of the

side bend the points illustrated in the model are the same, the important thing being the contraction of the thorax on the convex side implying an approximation of the lateral rib wall to the column.

To illustrate twist we rotate the thorax on its base (Fig. 8), the result of peripheral tension is to be seen in the general spiral arrangement of the tapes and a general distortion of all the leather ribs, although the path of this distortion is not quite as distinct as with side bend.

Cross-section model: The question now is, what is the result of these forces in the individual dorsal vertebra? To illustrate such tendencies I had another model made illustrating the mechanical arrangement of the ribs and column and muscular pulls as seen in cross-section (Fig. 9). The tapes representing