PROGKES8 IN RADIOGRAPHY. 153 



necessarily be taken into consideration, these operators groped their 

 way to an ability to make radiographs, more or less clear to be sure, 

 but often deplorable as indications of actual conditions, for the pic- 

 tures failed to show what ought to be shown, and sometimes even 

 showed what was not existent. 



The experimenters could of course make nothing of these deceptive 

 errors, for in the medical world particularly, questions of pure 

 physics and geometry are by no means clearly understood. An emi- 

 nent member of the Academy of Medicine voiced, therefore, the senti- 

 ments of a great number of his confreres when he declared that the 

 radiograph was likely to err gravely and that its indications were of 

 little value. He even w^ent so far as to say that no two radiographers 

 could take similar radiographs of the same fracture. That this last 

 statement is true in practice, although theoretically erroneous, has 

 been demonstrated experimentally before another learned assembly 

 by the production of radiographs of a fracture, taken at various 

 angles, in some of which the incidence was normal, in others oblique. 

 To understand the real fallacy of the case in question it is only neces- 

 sary to realize that since the radiograph is only a conic projection of 

 certain shadows these shadows will necessarily vary according to the 

 angle at which they are projected, a fact which makes all radio- 

 graphic deformations and exaggerations easily appreciable. For in- 

 stance, when walking along the street in the evening your shadow 

 from some gaslight grows longer and longer as you increase your dis- 

 tance from the lamp-post. About noon, shadows are short ; as the sun 

 sinks toward the horizon the rays strike objects more obliquely and 

 the shadows are elongated. These elongations are due to the angle of 

 the rays of light ; they illustrate exactly how two fragments of frac- 

 tured bone radiographed at too oblique an angle may unite to form 

 a single silhouette, the elongation of the shadows making a continu- 

 ous image, and thus hiding the fracture. 



The distance between the fractured member and the photographic 

 plate, and between the member and the Crookes tube, also plays an 

 important part in the fornuition of the radiographic image. In 

 throwing " Chinese shadows " on the wall with the hands it is 

 quickly evident that the silhouette is very small Avhen the hand is 

 close to the wall, and that it grows rapidly when the hand is moved 

 toAvard the light. The same principle is true in radiography. If, 

 as often happens, the ends of the fracture are far apart, the fragment 

 nearer the light gives an exaggerated shadow in the radiograph, 

 which joins the normal shadows of the more distant portion and 

 effectually conceals the break. Or if, as they more frequently do, the 

 two ends overlap, the radiograph will, of course, should the fracture 

 have been reproduced from a position beneath which the bones over- 

 SM 1905 14 



