42 Messrs. Barton and Bryan on the Absorption of 



was also found easy to adjust the resistance to a few ohms 

 and to restore it nearly to its original value after the change 

 which spoDtaneously occurred in it from day to day. Three 

 bridges of this type were used, of resistances 261, 549 to 560, 

 and 1336 to 1355 ohms. Let these be denoted respectively 

 by the letters F, G, and H. Finally, since the pencil- 

 markings on the glass form an extremely thin sheet, the 

 resistance may be assumed to be practically the same to high- 

 frequency waves as to the steady current by which they 

 were measured on the Post-office box. 



. Experiments. — The oscillator emits rapidly-damped electric 

 vibrations, only the first dozen, say, being appreciable. We 

 thus have, advancing along the line with the speed of light, 

 a damped wave-train, its large end, or head, leading and its tail, 

 after about twelve waves, being negligibly small. Suppose 

 now one needle of the electrometer to be connected to the 

 line and a bridge of no resistance to be placed at the end of 

 the line a little beyond the electrometer. We shall then have, 

 at the electrometer, stationary waves due to interference 

 between the incident waves and those reflected at the bridge. 

 Hence, if a series of readings be taken with the electrometer 

 at different distances from the end of the line the throws will 

 be found to periodically wax and wane. But when the bridge 

 is distant a few wave-lengths from the electrometer we have 

 the head of the wave-train interfering with the tail only. 

 And with somewhat greater distances between electrometer 

 and bridge the interferences of the waves and, consequently, 

 the waxing and waning of the electrometer-throws cease to 

 be appreciable. Thus, if the distances between electro- 

 meter and terminal bridge are plotted as abscissae and elec- 

 trometer throws as ordinates, we should expect the experiment 

 to yield a damped wavy curve. And this is the case, as first 

 shown by Mr. V. Bjerknes*, and utilized by him and his 

 successors to determine the wave-length of the oscillations in 

 use. The result of this experiment in our case is shown by 

 the full-line curve E on the diagram. 



In the case just considered the reflexion coefficient, p, of 

 the bridge is —1, as shown by equation (2). The waves 

 are therefore unchanged in magnitude by the act of reflexion. 

 Turn now to the general case of a precisely similar experiment 

 with a terminal bridge for which p is finite and less than 

 unity. In this case, obviously, theory predicts a curve of 

 similar form, wavy and damped, but lying between narrower 

 limits, since the reflected wave is now always smaller than the 

 * Wied. Ann. vol. xliv. pp. 522^523 (1891). 



