﻿600 Prof. J. A. Fleming on the Elect 



© 



ric 



The experiments here detailed were made at distances equal 

 to about one and one-third of a wave-length between the 

 radiator and receiver. Hence for this case \/r = 0*72, and 

 therefore cos 6 = 0-27 or 0=74° 20' or 0=105° 40', according 

 to the way of reckoning the angle. It is not a little remark- 

 able that in the case both of Mr. Marconi's measurements 

 and mine, made with bent oscillators having the ratio of the 

 horizontal to the vertical parts varying from 40 : 1 to 3, the 

 angle of minimum radiation always lies near to 110°, as 

 shown by fig. 2 in Mr. Marconi's paper (loc. cit.) and the 

 diagrams in Plate XVIII. in this present paper. 



It is also very interesting that the actual results obtained 

 from bent antennae of such different proportions as those 

 mentioned should agree so well with the theory evolved for 

 a doubly bent antenna with three equal branches. 



One other interesting point remains to be considered. 

 The experiments above mentioned were all made with 

 antennae bent at a right angle. The question arises, what 

 would be the effect of inclining the upper part of the bent 

 antennae at various angles ? A series of experiments was 

 accordingly made with a radiating antenna 20 feet in total 

 length, the vertical part of which was kept always at 5 feet 

 and the remainder tilted at various angles from 20° above 

 the horizontal to 20° below the horizontal. The receiving- 

 wire was 21 feet vertical and at a distance of 138 feet. The 

 current in the receiving wire as read by the thermoelectric 

 receiver was taken for various inclinations of the upper part 

 of the sending antenna, both with the free end pointing away 

 from the receiving wire and with the free end pointing 

 towards it. The ratio of these two currents is given in the 

 following Table III. and plotted out as ordinates in terms of 

 the inclination angle as abscissae in fig. 6. 



The figures of Table III. when plotted out into a curve 

 (see fig. 6) show a gradual increase in ordinate as the free 

 end of the antenna is elevated above the ground, but no decided 

 minimum value of the ratio of the fore and aft radiation. One 

 very striking difference was, however, found between a trans- 

 mitting antenna with the upper part horizontal and one with 

 the upper part placed in a down-sloping position when the 

 complete polar diagram was taken in the two cases. The 

 following experiment shows this difference. An antenna 

 was formed with the 20-foot wire and capacity-plate above 

 described, and was fixed to a wooden frame so that 5 feet of 

 it was vertical and contained the spark-gap, and 15 feet was 

 bent in a down-slanting direction so that the free end was 

 onlv 2 feet from the ground. This was then swivelled round 



