////■ iK.i\s.\nssio\ (xrr ./.v/) Hi.tiiRiixcn svsi i-.ms 405 



line is i-(|iial to tho s(|uarc of tlu- ratio of tlir rorri'spoiidin^; rnrri-iits. 

 On the basis of this relation, anil berause it is in jjeneral more con- 

 venient to measure or compute currents than powers, liic ciirrcnt 

 ratio has often lK>en used in (letermining the e(iui\alenl ol an\ piece 

 of apixiratus or line in terms of standard cable. It should be noted, 

 however, that such a current ratio can be properK- used as an index 

 of the transmission efficiency of a part of a circuit only when it is 

 et)ual to the square root of the ratio of the corresponding [lowers. 

 .Also, of course, the voltage ratio can In- siinilarK' used wlicii il meets 

 the s;ime reciuirement. 



Limitations in Isi-: ok .St.vndard ("Ani.h; I'nii 



.As has been shown ab(>\e, the attenuation, either of currcMi or 

 piiwer, correspontling to the mile of standard cable is directly pro- 

 portional to the square root of the frequency of the current under 

 consideration. This means that the standard cable mile corresponds 

 not only to a certain \-olume change in the reproduced speech sounds, 

 but also to a distortion change. For comparisons between the stand- 

 ard cable circuit and commercial circuits with talking tests and as 

 long as most of the commercial circuits had distortion comparable 

 to that of standard cable, this two-fold effect of standard cable was 

 desirable. At present, however, many types of circuits are being 

 used which have much less distortion than standard cable. .Also, the 

 use of voice testing has been largely given up in the plant and it is 

 now the general practise to determine the efficiency of circuits and 

 apparatus on the basis of measurements and computations for single- 

 fretjuency currents, a correlation having been established between 

 these latter results and those of %-oice tests. These factors have 

 made it tiesirable to have a unit for expressing transmission efficiencies 

 which is distortionless, that is, is not a fimction of fre(iuonc>-. 



(JlAI-IFICATIONS OI" A NlCW I'MT 



The consideration of a new unit for measuring transmission efficiency 

 brought out the following desirable qualifications: 



1. Logarithmic in Character. Some of the reasons for this h;j\-e 

 alreadv been discussed. In addition, the application of such a unit 

 in measurements of sound make a logarithmic unit desirable, since 

 the sensation of loudness in the ear is a logarithmic function of the 

 energ>' of the sound. 



2. Distortionless. The advantages of a unit which is independent 

 of frequency have l)een referred to above. In expressing the effi- 



