﻿602 Prof. J. A. Fleming on the Electric 



into various azimuths, and the current in a vertical receiving- 

 wire 21 feet high and at 138 feet distance was measured and 

 plotted as a polar diagram (see fig. 14, Plate XVII.) . When 

 this polar diagram is compared with the polar diagram 

 (fig. 1, Plate XVII.) for the antenna of same length but 

 having its upper part 15 feet long kept in a horizontal 

 position whilst being swivelled round, a very considerable 

 difference is seen in their form. 



The polar diagram (fig. 14, Plate XVII.) for the antenna 

 with down-sloping upper part has two deep indentations in it. 

 The ratio of the radiation in the fore and aft directions in the 

 two cases is not very different, that in the direction of the 

 free end being 72 to 76 per cent, of that in the opposite, but 

 whereas the diagram for the horizontal antenna is a smooth 

 oval curve, the other is of a distorted figure-of-8 form. 



Moreover, the absolute value of the current in the receiving- 

 wire in the two cases for the maximum readings at 0° are in 

 the ratio of about 100 : 70, so that the maximum value of 

 the current in the receiver is reduced by bending down the 

 free end so that it lies below the horizontal. 



The explanation of the distortion is not difficult. By 

 bending down the free end of the antenna we increase the 

 magnetic moment and reduce the electric moment of the 

 antenna. The polar curve for an antenna with magnetic 

 but no electric moment, viz., for a closed circuit, is a figure- 

 of-8-shaped curve. The polar diagram for an antenna with 

 electric moment but no magnetic moment, viz., for a vertical 

 wire, is a circle. 



In proportion as we lower the free end of the antenna by 

 bending down the upper part into a down-sloping position, 

 we make the magnetic moment of that antenna more 

 pronounced in its effect on the polar curve. 



We do not alter much the ratio of the fore and aft radiation 

 in the plane of the oscillator, but we cut it down immensely 

 in two regions lying between 30° and 85° reckoned from the 

 direction in which the free end points. 



These differences in the polar curves are perfectly accounted 

 for on the theory given by the writer (see Proc. Roy. Soc* 

 Lond. vol. lxxviii. A. p 9 1, 1906). It has there been shown 

 that for the bent oscillator, as used in the above experiments, 

 the magnetic force H parallel to the earth's surface, which 

 cuts across the vertical receiving wire, at the distance r, is 

 given by the expression 



H= i /y/^rmV) 2 + Uvmr- ^ mh 2 cos d\*, 



where <£ is the electric moment and M the magnetic moment 



