208 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1952 



constructed. The terminal transmitter converts either message or television 

 signals to a frequency-modulated signal centered on 70 mc and the terminal 

 frequency-modulation receiver recovers these signals, thus providing a link 

 between the relay system and other telephone facilities. 



Observer Reaction to Low-Frequency Interference in Television Pictures* 

 A. D. FowLERi. p^oc. I.R.E., 39, pp. 1332-1336, Oct., 1951. 



This paper presents results of tests to determine how much low-frequency 

 interference can be tolerated in black-and-white television pictures. Various 

 levels of single low-frequency interference were superimposed on a locally trans- 

 mitted television picture. Observers viewed the picture and rate the disturbing 

 effect of each level of the interference. Ratings were made in terms of preworded 

 comments ranging from "not perceptible" to "unusable." Interfering frequencies 

 from 48 to 90 cycles per second were employed. 



Just visible interference appears as a flicker. The rate of flicker is the differ- 

 ence between interfering and 60-cycle field frequencies. The most disturbing 

 interference produced a flicker rate of 5 or 6 cycles per second. To be tolerated, 

 peak-to-peak amplitude of this interference had to be 54 db weaker than the 

 peak-to-peak amplitude of the television signal (including synchronizing pulse). 

 For flicker rates of 0.5 and 12 cycles per second, the amount of interference 

 which could be tolerated was larger by 14 and 3 db, respectively. 



Arcing at Electrical Contacts on Closure. Part II. The Initiation of an 

 Arc* L. H. Germeri. Jl. Appl. Phys., 22, pp. 1133-1139, Sept., 1951. 



The capacity of the plates of an oscilloscope charged to 35 or 40 volts is dis- 

 charged repeatedly by approaching electrodes of carbon, active silver, and inac- 

 tive silver. Facts about the discharges, which are arcs of very short duration, 

 are inferred from resulting open circuit potentials and calculated electrode 

 separations. 



The separation at the first arc varies in different experiments but corresponds 

 on the average to a nominal electric field of 0.6 X 10* volts/cm for carbon or 

 active silver and to 2 X 10^ volts/cm for inactive silver. Each arc is initiated 

 by a very small number of field emission electrons. The hypothesis that a single 

 electron may perhaps be sufficient is consistent with observations at later stages 

 of each closure when the electrodes are closer and the field much higher. 



The earlier observation, that the potential across a short arc is constant and 

 independent of current, is not true if the arc time is sufficiently short. For 

 active silver a time comparable with 2 X 10~^ sec is required to establish the 

 steady arc voltage characteristic of later stages of arcs which last longer than 

 this. The initial time during which the potential is decreasing toward its final 

 steady value is 100 times the transit time of a silver ion across the gap. 



* A reprint of this article may be obtained on request. 

 iBell Tel. Labs. 



