EXTENSION OF THE TELEPHONE SYSTEM TO SHIPS 143 



edge obtained in the laboratory as a means for enabling the system 

 to be tried out. A preliminary survey of the purely technical prob- 

 lems convinced us that the more important ones were the develop- 

 ment of two-way radio telephone apparatus and of multi-channel 

 systems which would operate from a.single transmitting station without 

 interference between channels; the design of transmitting apparatus 

 which would satisfy the requirements and which could be built with 

 the vacuum tubes available and the development of a type of receiv- 

 ing system which would provide sufficient selectivity to allow an 

 economical use of the frequency range and at the same time fit in 

 with the two-way system most likely to be adopted. It was decided 

 that during the laboratory development work preparations should 

 be made in the field for providing the necessary experimental stations. 

 This field work as it developed included the location of station sites, 

 the actual construction of the station buildings and the antennas, 

 the equipping of the stations with the apparatus as developed in the 

 laboratory and as further developed in the station, the equipment of 

 the ships, and, finally, the operation and tests of the overall system. 



System Design Considerations 



In the beginning it was thought that to cover the required 200 

 miles (320 km.) range about one or one and one-half kilowatts in 

 the antenna would be necessary. It was not known that wave lengths 

 would be made available for this work by the Department of Com- 

 merce. To produce this amount of power in the antenna there were 

 available Western Electric 250 watt tubes which it was decided to 

 employ. The question then arose as to the particular type of trans- 

 mission systems most suitable for the work. The points of import- 

 ance in solving this problem are as follows: 



The greatest economy both in power and in wave length range 

 may be secured by transmitting only one side band of the modulated 

 wave. Moreover, this method has the great advantage that varia- 

 tions in the transmission characteristics of the medium do not cause 

 as great fluctuations in the received signal. This is because the re- 

 ceived signal is proportional to the product of the carrier and side 

 bands and if the carrier is supplied locally instead of being trans- 

 mitted, it is not affected by transmission factors. The use of such a 

 system, however, or of one in which only the carrier is suppressed, 

 throw upon the receiving set the burden of maintaining a constant 

 oscillator frequency not only comnlirpting it but also making reception 

 impossible for the great majority of ships which are equipped with 

 only straight detectors. This would defeat general inter-communica- 



