12 L. Schwendler — On the General Theory of Duplex Telegraphy. [No. 1, 



For a tolerably good line I" + p" as well as V -f- p can be taken as small 

 in comparison with i ; hence approximately 



£ 1_ 1 



K'~~ V + I" + p' + p n ~ L + p' + P " an 



From which it follows that also for the compensation method p and p" 

 should be selected as large as possible. 



But p = a -f-jfdoes not give a condition besides that we know we 

 should select a and /"absolutely not small. 



Further we see that the disturbance D 3 ' has v' for its factor, while D t ' 

 has X'v' for its factor. 



Hence for a given A.' v', the best will he to make v as small as possible. 

 The regularity of the signals is therefore obtained if we fulfil the fol- 

 lowing conditions in either station. 

 B = K 

 a = o 



p as large as possible 

 v as small as possible. 

 Knowing this we may now consider that balance in either station is 

 rigidly obtained, or that 



H \/a — K X v v b = o 

 but B = K 



and a = b 



we have X v = 1 



The absolute value of a may now be determined by considering that it 

 is advisable to produce the signals in either station in the most economical 

 manner. 



Maximum Magnetic Moment. 

 We have 



^ = ^77+7^'^ 

 **- 7+7+7 *VV? 



But 



[i! p." i 



a" +f" + c" a ' + f + c Q 

 where Q = * (L + p' + p") + (T + P ') (I" + p") 



P = P' + P" = i e " q ' ^ * *' g " ^ a " 



which has a maximum for a' and a" taken as independent variables. 



