NATURE 



26^1 



THURSDAY, NOVEMBER 15-, 1917. 



ELECTRICAL ENGINEERING. 



(i) The Theory of the Submarine Telegraph and 

 Telephone Cable. By Dr. H. W. Malcolm, 

 Pp. xi + 565. (London: The Electrician Print-- 

 ing and Publishing Co., Ltd., n.d.) Price 185. 

 net. 

 (2) Alternating-current Electricity and its Applica- 

 tions to Industry. Second Course. By W. H. 

 Timbie and Prof. H. H. Higbie. Pp. ix + 729. 

 (New York: John Wiley and Sons, Inc. > 

 London: Chapman and Hall, Ltd., 1916.) 

 Price 135. 6d. net. 

 ^i) TN the summer of 1850 a small party of 

 • ' -^ engineers arrived at Dover in order to lay 

 a cable across the Channel. One of them — Wil- 

 loughby Smith, who was afterwards president of 

 the Institution of Electrical Engineers — has left 

 an interesting record of their adventures. At that 

 period it was considered absolutely unnecessary to 

 test copper wire. All copper wires were supposed 

 to have the same conductivity. The cable was 

 twenty-five miles long, made up of short lengths 

 of wire which purported to be No. 14 Birmingham 

 Wire Gauge, but which varied in diameter. It 

 was covered with gutta-percha, so that the out- 

 side diameter was about half an inch. No 

 armouring of any kind was used. The cable 

 was coiled oa board the tug Goliath,. and the party 

 had to wait some weeks for a calm day. The pro- 

 ject excited much good-natured ridicule amongst 

 the town folk. A man was found cutting the 

 cable with his knife to show his friends that there 

 was a wire inside. A spectator was heard ex- 

 plaining to interested listeners that it was im- 

 possible to pull a cable of this kind 25 yds. long 

 resting at the bottom of the sea. It was, therefore, 

 absolutely impossible to pull one twenty-five miles 

 long. He evidently thought that the function of a 

 cable was similar to that of a bell-pull. On an 

 ideal calm day the pioneers laid the cable from 

 Dover to Grisnez, but they were destined to bitter 

 disappointment. The letters printed by the type- 

 writing instriimfent at Grisnez were so mixed that 

 the few messages received were quite undecipher- 

 able. To make their discomfiture complete, the 

 anchor or the trawl of a fishing-smack cut the 

 cable in two not many hours after it was laid. 

 They were thus prevented from carrying out ex- 

 periments which would probably have enlightened 

 them considerably on the laws governing the 

 transmission of submarine signals. As it was, 

 thev had no conception that their failure was 

 mainlv due to ignorance of the laws of electrical 

 capacity and induction. 



In 1866, when world-wide interest and en- 

 thusiasm were aroused by the laving of the 

 Atlantic cable, one would have anticipated that a 

 "book on the theory of the subject would be pub- 

 lished, at least in a ^ few years' time. The 

 development of the theory, however, was difficult, 

 and needed laborious investigations by mathe- 

 NO. 2507, VOL. 100] 



matical physicists. Most of the cable companies 

 train their own enigineers, giving them an insig^it 



j into both their technical and commercial activi- 

 ties. Few cable experts, therefore, have sufficient 



I mathematical knowledge to understand the writ- 

 ings of Kelvin, Heaviside, and Pupin, and so there 

 was little demand for a text-book on the subject. 



When, however, submarine telephony began to 

 be studied the importance of the labours of Heavi- 

 side and Pupin were appreciated. The engineers 

 of the Post Office and of the late National Tele- 

 phone Company studied thfl theory enthusiastically 

 and carried out most painstaking work in their 

 research laboratories. They were helped in no 

 small measure by the papers of Kennelly. Flem- 

 ing, also, by numerous papers and lectures, ren- 

 dered invaluable help to our telephone engineers. 

 The recent great advances in submarine telephony 

 are due to the recognition by engineers of the 

 importance of the work done by the mathematical 

 physicists. The application of these results to 

 practice, however, was a triumph for which all 

 the engineers concerned deserve the greatest 

 credit. 



Let us compare, for instance, the simple order 

 for twenty-five miles of cable given to the Gutta- 

 percha Company in 1850 with the specification 

 for the sixty-four-mile Howth and Aber Geirch 

 (Ireland and Wales) submarine telephone cable 

 laid in 1913. The specification says that the 

 attenuation constant of the cable must not exceed 

 o*oi6 per naut (nautical mile) for sine-shaped 

 waves of frequency 800. Considering that the 

 value of the attenuation constant depends on many 

 factors, this clause proves the confidence of the 

 practical engineer in his ability to gauge the pro- 

 perties of the materials he uses and his faith in 

 theory. Experiment later proved that the actual 

 value of the attenuation constant was 0*015 ^^ 

 the specified frequency. The engineers had few, 

 if any, misgivings that the cable might prove a 

 failure. Once they had determined the physical 

 constants of the cable, they knew from their 

 laboratory experiments that they could calculate 

 the quality of the speech transmitted. The only 

 evidence of lack of faith, perhaps, is that they 

 connected Aber Geirch with Manchester, and 

 Howth with Dublin, by aerial lines of copper 

 having the abnormal weight of 600 lb. per mile. 

 We know that, if the position of the "loading" 

 coils and their sizes had not been calculated by 

 elaborate and lengthy mathematical forrnulfe, 

 speech between Manchester and Dublin would 

 have been impossible. ' 



At first sight it is not obvious why the theory 

 of the submarine telephone cable should be 

 simpler than that of the submarine telegraph cable. 

 The reason is that speech-sounds can be treated 

 as periodic waves. The microphone transmitter 

 also is admirably adapted for producing these 

 waves, and the telephone receiver is a marvellously 

 sensitive instrument. The working of the receiv- 

 ing apparatus of a submarine cable, however, 



I depends on what electricians call transient pheno- 



I mena. The mathematical solution is given in a 



M 



