Jur.v, 1011. 



KNOWLEDGE. 



267 



reduced as much as possible. On tlu' ci;>ntrar\-, Mr. 

 Heaviside showed that what most telephone lines 

 required was not less, but more, inductance, to make 

 them less distortional. In short, inductance is capable 

 of neutralizini:; ca[)acit\' in cables. The reason for tiiis 

 is that capacity in cables acts as if it were a sort of 

 vacuum, into which electricity tries to rush just as air 

 rushes into a gaseous vacuum. On the other hand. 

 inductance opposes the m<i\-ement of electricit\' and 

 hence inductance in series witli capacit\- can be made. 



FlGlIRE 1. 



Mode of inserting tlie loading coils in tlie two sides of a 

 teleplione cable. 



b\' suitable adjustments, to neutralize each other. 

 The suggestion was therefore made as far back as 

 1887 that to effect an improvement in the qualities of 

 the line it was necessary to add inductance to it. .Ml 

 problems in engineering are, however, ultimateK' 

 questions of cost and detail, and until it had been 

 practically shown that tliis addition of inductance 

 was both economical and aci\'antageous, telephonic 

 engineers were liardh- justified in embarking on 

 expensive constructions. In 1899 and 1900. however. 

 Professor Pupin, in the United States, published the 

 results of some remarkable investigations on this 

 subject. He showed that if coils of \\ire (called 

 loading coils, see Figure 1) ha\ing high inductance 

 were inserted in the run of a telephone cable 

 at equal distances, and so close that nine or 

 ten of the coils were covered bv or included in 

 one wave of the current, the result was as if tlie 

 inductance were smoothly distributed over the 

 cable and also that provided the inductance were 

 large enough, a considerable impnnement in the 

 speech-transmittit^ qualities resulted. The phrase, 

 wave of current, needs a little explanation. The 

 characteristic of a wave motion of an\' kind, 

 whether in air or water, is tliat some change is 

 periodicalK' made in the medium, which is repeated 

 successively at contiguous points. If \\e have a long 

 cable, at one end of which some kintl of periodic or 

 alternating electromotive force is applied, then the 

 changes in potential are not rejiroduced instantly at 

 all points in this cable, but are propagated from point 

 to point in it with a certain \'elocit\'. The more 

 rapid these changes the faster they travel, up to 

 a limiting value which is equal to that of light. .Vt 

 certain points in the cable, separated by distances 

 called one wave-length, changes of (lotential of 

 similar type take place at the same time: tluit is to 

 say, the [jotential becomes zero or a ma.xiiuum at the 

 same instant. It is found that the mean \alue of 

 the frequency involved in ordinar\- speech is nearh- 

 eight hundred. That is, the mean pitch of the aerial 

 vibrations has this value. 



Corresponding to this frequency in a telephonic 

 cable the wa\e-length nia\ be ten to twenty miles or 

 so. Now a telephonic or telegraphic cable has four 

 specific (jualities. Its conductor has a certain 

 resistance [ler mile reckoned in ohms, and also a 

 certain inductance measured in units called a henry. 

 .\lso it has a certain cajiacits' jier mile, and a certain 

 leakage per mile. 



Mr. Heaviside first shnwtd tliat to make a cable 

 distortionless in the sense that \\'aves of all frequencies 

 would travel along it at the same rate, it is necessary 

 that the product of the resistance and of the capacity 

 per mile shall be equal to the product of the induc- 

 tance and leakage [)er mile. 



In all ordinary cables the capacit\- is too great to 

 fulfil this condition, and hence we have to increase 

 the inductance to approximate to the distortionless 

 condition. 



Pupin showetl that we can add this inductance in 

 lumps, so to speak, provided we insert these lumps 

 of inductance at such inter\-als that there are nine 

 or ten rn\h per wa\'e-length at a frequenc\- of eight 

 hundreil. A telephone cable so constructed, with 

 inductance coils inserted in it ever\- mile or so. is 

 called a locuk'd cable, and loaded cables have of late 

 years played an important part in im))ro\Mng tele- 

 phonic communication. .After the publication of 

 Pupin's researches attempts were made to put them 

 to practical test in long overhead telephone lines, 

 both in the United States and in Germany. The 

 results wen.' ver\' encouraging, and the attention of 

 t(.>leph(inic engineers was closeh' directed to the 





<if,^^^.^^ 



-dS-A -» 



,^' 





mi ' 



Figure 2. 



A loading coil for loaded telephone cables as used by the 

 National Telephone Company. 



subject. It is a comparativeh- easy matter to insert 

 such inductance coils in an overhead line, because 

 thev can be attached to the telegraph posts, a coil 

 being inserted in the line circuit every mile or so. 



The coils used consist of iron rings built up of 

 fine iron wire wliich are wound over with man\- 



