﻿Radiation from Bent Antem 593 



thus providing a tellurium-bismuth the rmoj auction in vacuum, 

 one junction in contact with a fine resistance wire which 

 could be heated by electric oscillations sent through i: - 



fig- 5). 



The thermojunction was connected by long wires with a 

 sensitive movable-eoil galvanometer placed in a room about 

 40 feet from the receiving antenna. The fine wire circuit was 

 included in the receiving antenna between the base and the 

 earth-plate. This thermal receiver could be calibrated by 

 passing a continuous current through the heating wire, and 

 in the case of the instrument mostly used in these experiments 

 the sensitivene^> was such that 1 milliampere passed through 

 the heating wire produced a derlexion of 1 cm. of the spot of 

 light on the galvanometer scale. 



Observation showed that the deflexion of the spot of light 

 on the scale was almost exactly proportional to the square of 

 the current through the heating wire. Hence an oscillatory 

 current, the root-mean-square value of which was as small as 

 300 microamperes, could be read by the spot deflexion. This 

 sensitiveness cannot be called great, but it was .sufficient for 

 the purpose, and such resistance-wire combined with a vacuum 

 tellurium-bismuth couple is not only more easily calibrated bur 

 less troublesome to use than a bolometer bridge arrangement. 



Xo attempt was made to tune the receiving antenna to the 

 transmitting antenna with any great exactness. The actual 

 experiment consisted in placing the bent transmitting antenna 

 at a certain distance from the fixed vertical receiver antenna 

 and then swivelling round the horizontal part of the transmitter 

 into various directions and taking readings of the current in 

 the receiving antenna in each ca>e. To avoid errors due to 

 variation in the sending antenna current, a comparison measure- 

 ment was always made in each case between the current in the 

 receiver with the free end of the bent transmittino; antenna 

 pointing directly away from the receiver and the current when 

 the transmittino- antenna had anv other assigned azimuth. 



In the first set of experiments the sending antenna con- 

 sisted of a double copper wire 20 feet in length having at the 

 free end a terminal capacity plate, the wire being bent over at 

 various heights fro,m the ground or earthed end, so as to 

 make a bent antenna with 1 foot vertical and 19 feet hori- 

 zontal. 2 feet vertical and 18 feet horizontal. 3 feet vertical 

 and 17 feet horizontal, 1 feet vertical 16 feet horizontal, 5 feet 

 vertical 15 feet horizontal, Occ. &c; and in each ca-e obser- 

 vations were taken of the current in a receiving antenna at the 

 same distance but at various angular positions round the trans- 

 mitter. The currents produced in the receiving antenna, as 

 calculated from the galvanometer deflexion, were plotted as 



