STEADILY ALTERNATING CURRENT ON A LONG LINE. 343 



at the crests of successive alternations. In such instances, the 

 waves will change abruptly from one sinusoid to another of dif- 

 ferent amplitude. The resulting waves may therefore appear to be 

 distorted, or nonsinusoidal, until they are analyzed. An example 

 of this kind is presented in Fig. 13 (Film No. loi). Here the full- 

 length artificial line (1255 km.) is freed at the distant end, and the 

 recording oscillographs of e.m.f. and current are connected to the 

 line at the half-way point (627.5 km. from each end). It will be 

 seen that the impressed e.m.f. eg is applied near an instant of zero 

 voltage, and without perceptible splash. The voltage wave ^390 at 

 the midlength point, commencing about 2 milliseconds later, rises 

 also without splash, along an approximately sinusoidal curve. Just 

 about the crest of this wave, the first reflection (positive) has had time 

 to return from the distant free end. This reflected wave adds itself 

 to the first arriving wave at the midway point, and so apparently 

 distorts the wave form. These distortions can be noticed for several 

 alternations in ^390, and are due almost wholly, if not entirely, to the 

 superposition of new reflections. The same considerations apply 

 to the current /'390 in the line, at the midway point. The electrical 

 connections are indicated in Fig. 14. The first reflected wave of 



o 628 km _^ 6 26 km 



H @>- 



A ^1 B 



0= 



Fig. 14. Electrical Connections of Oscillograph and its Leak, half-way along 



the Artificial Line freed at B. 



current arriving at P, 4.3 milliseconds after the first arrival at P, 

 is in the negative direction from the distant free end B, and so 

 produces a more marked apparent distortion in ^90, than occurs in 

 ^39o> where the first reflection is positive. Similar distortions due 

 to further incoming reflections are noticeable in the curve ^go for 

 several alternations. At the end of the oscillogram, 125 milli- 

 seconds from the start, the waves of ^390 and j'390 have very nearly 

 attained their final steady and practically sinusoidal values. 



