AUSTIN: APPARATUS FOR RADIOTELEGRAPHY 285 



over the surface of the earth than the damped wave trains pro- 

 duced by spark sending. Several attempts have been made to 

 settle this question by experiment, but over the moderate dis- 

 tances employed no difference in absorption has been observed. 

 In order to extend these experiments to greater distances, a 30- 

 kw. arc operated with 500-volt direct current was installed at 

 the high power station at Arlington, Virginia. At a wave length 

 of 4100 meters, the arc gave an antenna current of from 47 to 53 

 amperes. Comparisons were made of the received current from 

 this arc and from the 500 cycle spark set giving from 100 to 120 

 amperes in the antenna. A very careful set of observations of 

 the received currents from the two types of apparatus was made 

 at St. Augustine, Florida, the measurements being taken by the 

 calibrated detector and galvanometer method. The distance 

 between the two stations was 530 nautical miles. The received 

 currents were found to be simply proportional to the radiation 

 currents at Arlington with an error not greater than 10 per cent; 

 that is, at this distance there was no evidence of a difference 

 in the absorption. These results were verified by the shunted 

 telephone method using the slipping contact detector, ^ at New 

 Orleans and also at Key West, both places being approximately 

 900 miles from Washington. 



The receiving apparatus was then placed on the U.S.S. Ar- 

 kansas and taken to Colon, 1800 nautical miles from Arlington. 

 During the two days available for observation at Colon the 

 receiving apparatus was taken to the Naval Radiotelegraphic 

 Station. During these two days, the arc signals were heard at 

 each schedule both day and night, while the spark signals were 

 heard only at night. These observations indicated that at 1800 

 miles the continuous waves show a smaller degree of absorption 

 than the damped waves. It was not possible, however, to draw 

 this conclusion with certainty, since at the season of the year in 

 which the observations were taken, late December, exceptional 

 days occur which might perhaps affect the continuous oscillations 



in a different manner from those of the spark. 



/ 



» Journ. Wash. Acad. 1: 8. 1911. 



