CHARACTERISTICS OF TOLL TELEPHONE CABLES 311 



is the value of the quantity G/2C = conductance divided by twice the 

 capacitance, both measured at the room temperature in the factory. The 

 quantity G/2C is used because it is the coefficient in the leakage component 

 of attenuation as explained in connection with the formula (12) below for 

 high-frequency attenuation. The average value of G/2C for 1000 cycles 

 at 70° F. is about 8.3. This quantity increases with frequency and at the 

 same time decreases with temperature in the same way G changes, since the 

 capacitance changes are relatively so much smaller than the conductance 

 changes. 



Layer to Layer Variations of Primary Parameters 



The values of the primary parameters vary from inside layers to outside 

 layers of cables, in addition to variations mentioned under specific param- 

 eters above. There are three basic reasons for this variation with location 

 in the cable. The first is that the length of an outside pair is usually 

 considerably greater than the length of an inside pair. Unusual twisting 

 of the inside layers might make up for this difference but in the ordinary 

 construction this is not done. This increase in length amounts to as much as 

 1 or 2 per cent and is reflected at once in the d-c. resistance as well as in the 

 a-c. parameters. 



The second reason is that, particularly in the outside layer, the sheath 

 being made of lead-antimony, has electrical properties considerably different 

 from the properties of copper wires. The large size of the sheath relative 

 to the wires is an important factor. High-frequency currents in the wires 

 near the sheath produce fields cutting the sheath which affects the fields in 

 a different fashion from the way adjacent copper wires affect the field of a 

 pair of conductors near the center. 



The third reason, closely allied to the second, is that the conductors in 

 the core of the cable are surrounded by a practically symmetrical mass of 

 copper conductors and paper plus the sheath, while conductors in any other 

 layer are surrounded by an unsymmetrical arrangement of conductors and 

 paper. 



A fourth factor is the variation in the amount of space allowed pairs in 

 the core by the pressure of the outside layers. 



The magnitude of these effects is indicated by the curves of Fig. 13, show- 

 ing layer-to-layer variations in per cent for Resistance, Inductance, Con- 

 ductance and Capacitance. Such large variations would be of considerable 

 importance were it not for the fact that in the process of splicing pairs are 

 made to pass, in effect, from inside to outside of the cable and vice versa. 

 A long study of these variations will be found in the paper by Wuckel'. 

 The effects of splicing together sections slightly different in their character- 



