276 BELL SYSTEM TECHNICAL JOURNAL 



claimed, required an unusual amount of concentration on their part, 

 causing fatigue and making it difficult to receive over the circuit for a 

 long period of time. 



The type of distortion for which the disagreement between opinion 

 and accuracy of reception was most pronounced, is negative character- 

 istic distortion. The operators considered 25 per cent of this type of 

 distortion to be about the maximum allowable value for commercial 

 operation, whereas about 50 per cent was required to cause an ap- 

 preciable reduction in their accuracy of reception. These values were 

 checked a number of times with different operators and appear to be 

 well established. A similar condition occurred in the case of fortuitous 

 distortion for which the corresponding values are 50 per cent and 85 

 per cent, respectively. In this case, however, the infrequent occur- 

 rence (about four times a minute) of the maximum value of this type of 

 distortion probably accounts for the small effect on the accuracy of 

 reception. 



These results would seem to indicate that the accuracy of reception 

 is not a complete criterion of the allowable amount of distortion in a 

 telegraph circuit. It is undoubtedly true that operators who work over 

 a circuit, often do not wait until they make errors before calling the 

 repeater attendant to fix the circuit. They usually take such steps as 

 soon as they notice any appreciable distortion or have any difficulty in 

 receiving signals. 



In explanation of this point it may be said that when an operator is 

 receiving perfect signals at a speed below the maximum rate at which 

 he can work, he has what may be termed a "margin of attention" 

 which may be utilized in interpreting defective signals. Before the 

 point is reached where his accuracy of reception is impaired, he ex- 

 periences a reduction in the "margin of attention " and on that account 

 may pronounce a circuit unsatisfactory, since operation with little or 

 no "margin of attention" soon produces a mental strain. The recep- 

 tion of signals having frequent distortions of say 30 per cent may very 

 likely decrease the "margin of attention" to a greater extent than the 

 reception of signals having only occasional distortions of 60 per cent or 

 even greater. Also, if the speed of transmission during the tests had 

 been slower or faster, the "margin of attention" would have been 

 greater or less, and the point at which the operators considered the 

 circuit unsatisfactory would have been different. 



Because of the above it is thought that in establishing allowable 

 limits of distortion for commercial telegraph circuits, the opinion of 

 operators must be given consideration in addition to the accuracy of 

 reception. 



