370 BELL SYSTEM TECHNICAL JOURNAL 



either by boat or airplane. It is probable that the wave structure of 

 the ocean surface would produce irregularities in reception if the 

 receiving antenna be too near this surface, as on a boat, and for this 

 type of receiver transport the time occupied by an experiment is 

 rather long. Naturally the slow motion makes a fine grained record 

 possible. In an airplane the time of transit is very much reduced 

 but the vibration and unsteadiness are not favorable for accurate 

 recording and the electrical noise level is high. There is also an 

 increase in range necessary to get the same angular distance below the 

 horizon as with a boat. This range extension, however, is relatively 

 not as great as might appear since the falling away below the horizon 

 is proportional to the square of the distance. For example, a line 

 tangent to the earth is 100 feet up at 14 miles and 1000 feet up at 45 

 miles from the tangent point. ^ If these be boat and airplane antenna 

 altitudes respectively, the airplane must always travel 31 miles farther 

 to get the same angle of refraction below the horizon as the boat does. 

 Since it is necessary to travel about 92 miles to get one degree below 

 the former horizon (angle between the two earth radii) and the trans- 

 mitting antenna height will further increase the range for a given dif- 

 fraction geometry, it is evident that the difiference in antenna altitude 

 as between a boat and an airplane, is not of serious effect either in 

 space covered or in accompanying signal attenuation. 



We were fortunate in being located so that a land plane could be 

 used to give us an over-water transmission. As a glance at the map 

 (Fig. 1) will show, it is possible so to locate a transmitter on the New 

 Jersey shore that there is an over-water path for an airplane flying 

 along the Long Island shore. Owing to the curvature of the Long 

 Island beach, an over-water path for the entire distance to Montauk 

 Point requires a location of the transmitter at or south of Long Branch, 

 New Jersey, and such a location makes the minimum path length 

 possible (Long Branch to Rockaway Beach) about 20 miles. This 

 was too great a distance to be satisfactory to us and we elected to 

 locate north of Long Branch. Although the curvature of the Long 

 Island shore then interposed land between Montauk Point and the 

 transmitter, this land lay well below the horizon, as viewed from the 

 transmitter, and it was thought, therefore, that its effect might be 

 small or negligible. 



North of Long Branch the favorable shore transmitter sites are 

 restricted to the Sandy Hook region and the stretch between Sea 



2 Air refraction is included by increasing the apparent radius of the earth to 5260 

 miles. See Schelleng, Burrows and Ferrell, Proc. L R. E. 21, 427, 1933; Bell Sys. 

 Tech. Jour.,X\\, 125, 1933. 



