214 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1957 



,xtO-'* 



UJ 



-J 



2 



< 



u 



1- 



< 



Z 



cr 



UJ 

 Q. 



in 



CO 

 u 



UJ 

 Q 



-120 



-160 



40 60 8C 100 12 



FREQUENCY IN KILOCYCLES PER 



140 



SECOND 



160 180 



Fig. 13 — Deviation of measured cable attentuation from design characteris- 

 tic as a function of frequenc.v. Typical average values for Types A, B, and D for 

 37°F and atmospheric pressure. 



tiire. The measurements indicated that the derived temperature co- 

 efficients were accurate to within ±10 percent. 



Measurements also were made to determine the effect of pressure on 

 the primary constants of the cable. These measurements indicated that 

 capacitance was the only parameter affected by pressure. The capaci- 

 tance increased linearly 0.1 percent for each 500 pounds per square inch 

 of applied pressure. Since the attenuation, a, is inversely proportional 

 to the impedance, it is evident that if C is the only parameter affected 

 by pressure, a will also be affected b}^ pressure to an amount equal to 

 approximately one half the pressure effect on C. The pressure coefficient 

 of oc was therefore established as 0.05 percent per 500 pounds per square 

 inch of pressure. 



LAVING EFFECT 



Analysis of the Florida-Puerto Rico cable data indicated that the 

 measured ocean bottom attenuation was less than the attenuation 

 predicted from factory measurements. The differences were large enough 

 to warrant study and indicated that the measurements were in doubt or 

 sea bottom conditions w^re not known accurately or that some un- 

 explained phenomenon was taking place. 



In March of 1955, approximately 22 miles of cable of the transatlantic 

 design were laid in 300 fathoms of water off the coast of Spain in the 

 Bay of Cadiz, and another equivalent length was laid in 2,300 fathoms 

 off Casablanca. Precise measurements of attenuation were made in 

 both cases, and it w^as established that a difference did in fact exist 



