302 TRANSATLANTIC TELEPHONING. 



The ear distinguishes the difference between these much })Gtterthan 

 the eye. 



The effect of the long telephone line upon these waves is something 

 like this: 



Fig. 4. 



The little ripples that distinguish the sounds die out before the 

 main wave. Such changes as these render repeaters useless on a tele- 

 phone wire, for no repeater can restore characteristics that have 

 already- been lost. On an ocean cable this dying out occurs more 

 quickly than on a land line; and, besides, the main wave is distorted 

 and flattened so as to lose its identit}^ altogether. 



This was the situation in long-distance telephon}" when Dr. Pupin 

 attacked the problem six or seven 3'ears ago. While tramping 

 through Switzerland in 1894 he improved his spare moments by read- 

 ing Lord Rayleigh on the theor}" of sound. That part relating to the 

 vibration of strings led him to consider the telephone problem. Sup- 

 pose a long string attached to a mechanism which can only be operated 

 by transverse jerks of the string. If the end of the string- at a dis- 

 tance from the mechanism be moved back and forth, waves will travel 

 along it, and may supply the jerks required to operate the mechanism. 

 But if the string be very light, and the resistance to its motion great — 

 if, for instance, it were in a tank of water — the waves impressed upon 

 it would rapidly die out, and it might be necessary to swing the string 

 back and forth with all the violence at our command, in order that 

 the}^ should reach the mechanism at all. Substitute a heav}' string for 

 the light one. Waves imparted to it will have a much greater power 

 of persistence. It will be necessary to impart a much less violent 

 motion, and this of itself reduces very much the effect of the resist- 

 ance of the medium in which the string swings. But we need not use 

 a string that is uniformly heavy. The effect of the heavy string may 

 be closely imitated by distributing heavy masses along it at intervals. 



Dr. Pupin set himself to solve the problem of the behavior of such 

 a loaded string in a resisting medium. Its solution had not been before 

 attempted, for its tremendous intricacy would baffle anyone who had 

 not at command, as Dr. Pupin has, the resources of the ''higher 

 mathematics." Many perplexing questions are involved. Given a 

 certain amount of energy, to be transmitted by means of a string 

 swinging in a given resisting medium, how heavj^raust be the masses? 

 How near together must they be placed ? Can they be so placed and 

 proportioned that they will serve equally well for the transmission of 

 long or short waves; that is, of slow or rapid vibratory motions? 



