274 BELL SYSTEM TECHNICAL JOURNAL 



copy of the message and the difference expressed as a percentage of 

 the latter. The nature of the errors was determined by translating 

 the errors into the corresponding Morse symbols and comparing them 

 with the Morse symbols for the correct characters. 



5. Transmission Measurements 



The amount of distortion present in the telegraph signals after 

 passing over the circuit was measured by means of a telegraph trans- 

 mission measuring set, as mentioned previously. This is a device for 

 measuring any departure in the length of signal elements from their 

 normal value. The signals used for the measurements were standard 

 signals, obtained from a mechanical interrupter. The latter consisted 

 of a printing telegraph distributor, operated at a constant speed and 

 arranged to send out a particular signal during each revolution of the 

 distributor brush arm. The procedure for the measurements con- 

 sisted in connecting the source of standard signals to the sending loop 

 of the circuit, and measuring the distortion of the various elements of 

 this standard signal at the receiving end of the circuit by means of 

 the measuring set. 



The types of standard signals used, consisted of reversals and signals 

 similar to the Morse letters C and E. Bias was measured with all 

 three signals, average characteristic distortion with the C and E 

 signals only, and maximum distortion including fortuitous with the 

 C signal only. 



The speed of signaling for transmission measurements was 15 d.p.s. 

 and the values of distortion given in the following discussion refer to 

 this speed. Since the speed of transmission by operators was 13.5 

 d.p.s., it is evident that the distortion which the circuit impressed 

 upon the operators' signals was somewhat less than the values given. 

 The distortion at the lower speed may be obtained approximately, if 

 desired, by assuming that the per cent distortions are directly propor- 

 tional to the speed of transmission. 



It should also be understood that, in addition to the distortion which 

 is due to the condition of the circuit, an appreciable amount of distor- 

 tion is introduced into the signals by the sending operator. The 

 amount of such additional distortion is a variable quantity depending 

 on the characteristic sending of a particular operator. Neither its 

 magnitude nor effect was determined in the present investigation which 

 was limited to the effect which the condition of a circuit had upon the 

 reception of signals transmitted and received by good telegraph 

 operators. 



