TELEGRAPHY — HILLIS 197 



possible telegraph service. Coaxial cable will provide this type of 

 service where large numbers of circuits are required, but the cost is 

 high and where high frequencies are used, repeaters are required at 

 frequent intervals. 



The Western Union electronic laboratory at Water Mill, Long 

 Island, has continuously investigated the possibility of radio as a 

 medium of transmission. Up to 1940, the use of radio was not advis- 

 able, for the frequencies used at the time did not provide the continuous 

 24:-hour service the year around that is required for dependable 

 telegraph circuit stability. 



The concentrated development in the ultra-short-wave spectrum for 

 radar techniques during the war disclosed that when the superhigh 

 frequencies were propagated under line-of-sight conditions, they ap- 

 peared to be quite stable. They were not affected by magnetic storms 

 or lightning discharges so it was apparent that this method of trans- 

 mission might be the solution to our problems. 



Before the war, equipment was not available to construct oscillators 

 which would generate frequencies much above 400 megacycles. Os- 

 cillator tank circuits were reduced in size until the capacity between 

 elements in the vacuum tube was used as tank capacity and a single 

 turn of wire for the tank inductance. The answer to generating still 

 higher frequencies was found in a new type of tube which utilizes the 

 speed of electron travel. Two types of tubes of this classification were 

 used during the war for radar work. They are the Magnetron and 

 Klystron. The Magnetron was developed by Dr. Hull in the General 

 Electric laboratories and was improved on from time to time. Rus- 

 sian scientists added a bit, but it was not until about 1940 that English 

 scientists made further improvements which enabled them to use the 

 tube for high-frequency generation. The Klystron tube was in- 

 vented by the Varian brothers at Stanford University. Either of 

 these tubes is capable of generating frequencies up to 30,000 mega- 

 cycles. 



It has been found that at these high frequencies, where the wave 

 lengths become several centimeters or less, they may be controlled in 

 much the same manner as light. There is still a long way to go before 

 the wave length of visible light is reached. At a frequency of 4,000 

 megacycles, the wave length is 7.5 centimeters or 3 inches, while in- 

 visible infrared rays have a wave length of about 0.01 centimeter in 

 length and the wave lengths of visible light rays range from 0.00007 

 to 0.00004 centimeter in length. 



A parabolic reflector, such as is used to concentrate the small candle- 

 power of a tiny incandescent lamp in an automobile headlight, may 

 also be used to concentrate the radiation of the microwaves. It is not 

 practical in standard broadcast wave lengths, for in order to concen- 



