92 BELL SYSTEM TECHNICAL JOURNAL 



forms part of a circuit which includes the repeater, the other being 

 simply a reference circuit. An amplifier-detector combination am- 

 plifies the voltages across these resistances and then rectifies them so 

 as to obtain an indication on a d.c. galvanometer. Equality of volt- 

 ages across the two resistances, which are designated as R and R' in 

 the figure, is thus indicated by equal deflections of the galvanometer. 

 When this condition is secured, the repeater gain is read directly from 

 the dials A and B. By means of this device, it is readily possible to 

 measure the gain of a repeater within a few tenths of a mile. Owing 

 to the fact that the measuring circuits are comprised entirely of 

 resistances, the readings of the set are independent of frequency, so 

 that gains can be measured at all important telephone frequencies. 

 As pointed out above, transmission measurements over the com- 

 plete circuits including the telephone repeaters are required at periodic 

 intervals in order to insure that proper transmission standards are 

 being maintained. By means of such measurements, the variations 

 in the overall equivalent of the circuits due to the cumulative effect 

 of small gain variations, slight variations which remain after the auto- 

 matic transmissicn regulators have compensated for the major varia- 

 tions in the conductors and variations from other causes including 

 the effect of different conditions of humidity on the wiring in the offices, 

 are determined and compensated for. These measurements are made 

 by applying a known electromotive force through a known resistance 

 to one end of the circuit and receiving the current at the distant end 

 with a suitable calibrated arrangement employing an indicating meter. 

 Since this type of measurement is similar in principle to the method 

 employed for measuring the gains of the individual repeaters, it will 

 not be described. 



IX. Automatic Transmission Regulators 



Since the resistance of long cable circuits employing small gauge 

 conductors is comparatively large, it is, of course, evident that changes 

 in this resistance caused by temperature changes to which the cable 

 circuits are subject will have a large effect on transmission. For 

 example, in the case of an X.L.L. 19-gauge 1000-mile circuit (1600 

 kilometers) in aerial cable, the total attenuation changes more than 

 110 transmission miles during the course of a year. This corresponds 

 to a variation in the received power of more than 10 10 or ten billion 

 times. 



It is, of course, essential to provide special means to counteract 

 these effects. Furthermore, since the temperature changes which 



