500 BELL SYSTEAI TECHNICAL JOURNAL 



The amplifier with each of the filters taken separately was calibrated 

 with input supplied by the thermal agitation in standard resistances 

 ranging from 50 to 250 ohms. The calibration temperature was 

 approximately 23° C. Check calibrations were made weekly and at 

 such times as changes were made in the apparatus. The stability of 

 the entire system was such that over periods of months measurements 

 were made with an accuracy closer than ± 1/2 decibel. 



In interpreting data on atmospherics of low amplitude, such as 

 received on submarine cables, it is necessary to take into account the 

 random voltages generated in the amplifier circuits and the thermal 

 agitation voltages of the conductor connected to the amplifier input. 

 Both of these voltages appear in the output circuits mingled with the 

 amplified atmospherics. The former originate principally in the first 

 stage of the vacuum tube amplifier. Thermal agitation produces a 

 random voltage, uniformly effective at all frequencies. The r.m.s. 

 amplitude of this voltage is dependent upon the frequency range 

 considered, the resistive component of the impedance of the conductor 

 and the temperature of the conductor. ^ The conductor in this case 

 is the cable or antenna circuit. The r.m.s. values of these voltages, 

 when integrated over periods of time comparable to those occupied 

 in taking data on atmospherics, are substantially steady; therefore, 

 their separation from the atmospheric voltages is not difficult. Correc- 

 tions for both amplifier and thermal noises have been made on the data 

 presented. 



Observations of audio frequency atmospherics received on long 

 antennas and loop aerials have been reported by several observers.^ 

 Their accounts describe the general characteristics, although some 

 confusion has occurred in identification of the musical atmospherics. 

 In view of the fact that the apparatus used by us was particularly 

 adapted to reception and analysis of frequencies in the audio range, 

 it appears that our data may add considerably to the information 

 previously disclosed. 



Types of Atmospherics 



Audio-frequency atmospherics observed on submarine cables are 

 essentially the same as those received from a long antenna except for 

 high attenuation and frequency discrimination attributable to the 

 cable characteristics and to the shielding effect of sea water.* The low 



2 J. B. Johnson, "Thermal Agitation of Electricity in Conductors," Pliys. Rev., 32, 

 p. 97, July, 1928. 



3 H. Barkhausen, "Whistling Tones from the Earth," Phys. Zeits., 20, p. 401, 1919. 

 T. L. Eckerslev, "Electrical Constitution of the Upper Atmosphere," Nature, 117, p. 

 821, June 12, 1926. 



^ John R. Carson and J. J. Gilbert, "Transmission Characteristics of Submarine 

 Cables," Jour. Franklin Inst., 192, p. 705, December, 1921. 



