128 



PROF. A. E. H. LOVE ON THE TRANSMISSION OF 



that they are proportional to the values of the square root of HOGAN'S "audibility 

 factor," but this is not stated expressly. As the received antenna current was 

 inferred in some way from the value of this factor, it seems unfortunate that the 

 value of the factor was not recorded. The behaviour of the detectors used for the 

 shunt readings is not thoroughly understood by electricians. For example, FLEMING 

 (op. cit. ante, p. 397) states concerning the Fessenden liquid barretter, used by HOGAN, 

 that its action is probably electrolytic, and due to annulment of polarization, rather 

 than purely thermal. HOGAN does not state how the approximate law of squares 

 connecting his audibility factor with the received current was verified, but FLEMING'S 

 statement would seem to indicate that the action of his detectors would approximate 

 to that of the detector used by AUSTIN. Now AUSTIN has investigated the action of 

 his detector. In a series of experiments performed at Brant Rock he measured the 

 same antenna currents in two ways : (l) by means of the shunted telephone and 

 detector ; (2) by means of a rectifier, which was " connected in a secondary circuit 

 coupled to the antenna, and calibrated by means of a thermo-element in the antenna 

 and an exciting buzzer circuit which could be tuned to the wave-length used." 

 Apparently the rectifier had to be used because it was desired to measure currents too 

 weak to be measured by the thermo-element, but it is not clear how the rectifier was 

 calibrated for such weak currents. The results of these experiments are recorded by 

 him in a table (p. 319), which does not support the conclusion that the current is 

 proportional to the square root of the audibility factor. The quantities recorded in the 

 table are the values of the shunt [SJ in ohms, and the antenna current [I] in 

 micro-amperes, the resistance [R] of the telephone being 600 ohms. From this table 

 I find that (R + S)/S is nearly proportional to I 2 for large values of I, nearly 

 proportional to I for small values. This result appears from the following table- 

 compiled from AUSTIN'S. 



TABLE IV. 



SI 



In view of this result, and the fact that most of the numbers recorded by AUSTIN 

 as values of received antenna current at distances exceeding 500 sea miles are less 



