416 BELL SYSTEM TECHNICAL JOURNAL 



strated, they are restricted to overhead Hnes which are not loaded. Inas- 

 much as a great and increasing number of overhead Hnes must pass through 

 cables at some point or points, the necessity for a satisfactory method of 

 carrying phantom circuits through cables has assumed great importance. 

 Inasmuch as the loading of overhead lines has already reached large pro- 

 portions and as, in view of our recent work, this importance will rapidly 

 increase, we must, unless some, method is devised of loading phantom 

 circuits, sacrifice the advantage either of loaded circuits or of phantom 

 circuits. Neither of these alternatives should be tolerated. We believe 

 and we confidentially predict that if we can place the organization of this 

 department upon a proper basis, we will develop new types of cable for carry- 

 ing phantom circuits and will devise and standardize methods for loading 

 phantom circuits, both in cables and overhead. Aside from the relation 

 which the phantom or duplex cable has to overhead phantom circuits, it is 

 important of itself, even where these overhead circuits are not to be con- 

 sidered. This is because we are entering the period of long distance under- 

 ground cables and the savings which may be made in these cables them- 

 selves, growing out of working them duplex, reach into very large figures. 

 As soon as the department is strengthened so as to respond to the extension 

 of its duties now projected, a special study will be made of physics and 

 economics of long underground cables, say between Boston and New York 

 and New York and Washington. We will also make a similar study with 

 respect to the possibilities of universal long distance service throughout 

 the United States. While we can already see as an outcome of these pro- 

 posed studies matters of very great importance and promise, and while we 

 can also see that in these directions great use will be made of loaded phan- 

 tom circuits and of phantom circuits in cables, we have thought it best 

 in the present memorandum not to count upon any savings which might 

 lie in these directions, but have restricted ourselves to those savings which 

 would follow in the natural course of events, even if our larger ideas were 

 not realized. Thus, counting upon the new toll cable construction which 

 will be done from year to year, we figure that the use of a successful duplex 

 cable would reduce our construction costs for this class of work at the rate 

 of about $180,000 a year. 



We are confident that if we are provided with the necessary ways and 

 means, we can devise a method whereby we can phantom loaded circuits. 

 By such an achievement the circuit capacity of the existing plant of the 

 American Telephone and Telegraph Company and the associate companies 

 could be increased to an extent which would be attainable under the present 

 state of the art only by an additional plant investment of $2,500,000, on 

 which the annual charges would be $250,000. Assuming that we have thus 

 increased the carrying capacity of the plant, additional circuits, by means 



