following experience, which has probably not been paralleled in the literature of this subject:

A boy four years of age, while playing on the street, fell against an iron bar. Being unable to rise again, he was taken up and carried to St. Mark's Hospital, where in the first instance moderate pain was noted besides the functional disturbance. There was neither any difference in level or any other deformity, nor any shortening or the typical equinus position. A photograph taken two days after the injury only showed a very moderate and uniform swelling of the leg.* Abnormal mobility and crepitus, in accordance, could be produced only by very rough manipulations.

On the day following the injury two skiagraphs were made in different positions; one of them (Fig. 137) in the dorsal and the other (Fig. 139, A) in the lateral position. To my surprise, figure 137-which had been skiagraphed by a direct irradiation, the center of the platinum disc of the tube being perpendicular to the anterior surface of the leg-did not show the slightest indication of a fracture, while figure 139, A (also compare Frontispiece), which represents the leg irradiated from the outer aspect of the tibia, shows a marked fracture-line.

The fracture presented the typical oblique type in the middle of the tibia, the fracture-line running from below anteriorly to above posteriorly, the upper, tapering fragment overlapping the lower end. No sideward displacement having been present, it can be understood why the rays, reaching the long axis of the tibia in a vertical direction, do not show the fracture-line.

* Photograph published in "New York Medical Journal," January 6,

A very slight change in the position, where the inclination toward the fibular direction amounts to less than one millimeter, brought out the fracture distinctly.

Now, if I had, as is the custom in general, taken a skiagraph in the anteroposterior direction only, and if the manipulations made during the first examination were carried out as gently as they properly should be, the fracture might have been overlooked entirely. And if, in view of the local pain and tenderness, the swelling, and the functional disturbance, the possibility of a fracture would have been seriously considered, the skiagraph (Fig. 137) might have silenced the uneasy conscience.

This experience teaches the necessity of adopting the principle of always taking at least two skiagraphs in two different positions in all cases of suspected fracture.

The medicolegal aspects of a case of this kind need no further comment.

In taking skiagraphs of foreign bodies it must be considered that their size varies according to the distance from the tube. (Compare p. 305.) In oblong bodies great errors as to their extent may be committed. The author once was very much surprised in a case where a needle-fragment had entered the palm of the hand in a perpendicular direction. The plate, while indicating the presence of the needle distinctly, created the impression that the fragment was only about two millimeters in length. When extracted it was found to be more than an inch long, the rays having reached the hand in a perpendicular direction, so that the circumference of the fragment was reproduced rather than its length. A side view, of course, would have cleared up the error at once.

Misinterpretations have also arisen from unavoidable mechanical and chemic defects, causing markings in the photographic plate, the significance of which must be well known to the skiagraphic interpreter.

Blemishes may be produced by spots caused by pus from wounds or by perspiration.

As

In the location of foreign bodies, especially in the skull, many errors were and are still committed. to their avoidance, compare pages 265 and 306.

In drawing conclusions from skiagraphs it should especially not be lost sight of that a skiagraph is by no means a photograph of an object, but a silhouette —that is, a photograph of its shadow.