340 BELL SYSTEM TECHNICAL JOURNAL 



the intelligence-carrying component contains no outstanding narrow- 

 range of frec]ucncies to which tuning can be applied should make 

 obvious the error in this assumption. 



The other assumption is that a \va\e, which is ideal for the trans- 

 mission of power, is also ideal for the transmission of intelligence. 

 As a matter of fact, the transmission of intelligence inherently involves 

 rapid and unpredictable changes in the current, whereas the trans- 

 mission of power is best brought about by steady current, either 

 direct or alternating. These two conditions are, of course, incom- 

 patible. 



.APPENDIX A 



Use has been made of the following two principles: 



1. In a telegraph circuit in which the line speed is near the maxi- 

 mum, the shape of the received dot is substantially independent of 

 the shape of the impressed dot, and 



2. The magnitude of the received current is approximately |)ro- 

 portional to the area under the transmitted voltage curve. 



The following general discussion of these principles has been fur- 

 nished by J. R. Carson. 



Let the arrival curve, due to suddenh- imijrossetl unit battery be 

 denoted by A (/); then the received signal 5 (/), due to the elementary 

 dot impressed signal / (/) is given by ' 



S(t)= ffix)A'(l-x)(lx (I) 



the upix'r limit of integration being / for t<T and 7" for t^T. The 

 latter case will alone be considered since the conclusions arrived at 

 in this case are conser\-ative. 



Expanding .1' (/— .v) in (1), we get 



S{t) = [>1 ' (/) - ^'^A "(/) +~A '"(/) . . ■]/ /(-vW.v (2) 



rxj{x)dx 



where //2 = 



,//(.v)..v 



-,//f.v)./.v 



'J. R. Carson. "Theory of the Trnnsient Oscillations of Electrical Networks 

 and Transmission Systems." A. 1. E. E. Trans., Vol. XXXV'III, 1919, p. 345. 



".1— J 2 



