THE NATURAL PERIOD OF LINEAR CONDUCTORS 



413 



occurring somewhere in the generator circuit. With the advent of 

 inclement weather the source of these resonances could not be looked 

 for. These dips in the resonance curve are not evident to an observer 

 watching the resonance rod meter as the generator wave length is 

 varied; they become evident only on plotting a carefully taken reso- 

 nance curve. The results both for preliminary eye settings and final 

 resonance curve are: 



A check eye setting on the 250 cm. rod mounted vertically agreed 

 with the other eye settings. The eye settings, or eye estimates of the 

 top of the rod resonance curve, were made first, the resonance curves 

 were run last. On discovering the "dips" in the 250 cm. rod curve 

 it became useless to run a 227.1 resonance curve before eliminating 

 these dips, the preliminary curve being discarded. It is not certain 

 however that these "dips" were present in most of the eye settings 

 as these were chiefly made with the generator nearer ground and 

 with shorter power leads. They are given for what they are worth. 



Evidently experiment more nearly checks Abraham ^ than Mac- 

 donald, the rods operating as if their effective lengths were 6-7 per 

 cent greater than their physical lengths. Whether this "end cor- 

 rection" varies with the rod diameter was not investigated owing to 

 bad weather. Several rods were however bent into circles, cut to 

 250 cms. perimeter (outside length), and their natural periods deter- 

 mined. The results follow below. With the bending, the radiation 

 resistance of the rods went down pronouncedly and the knife blade 

 contacts were shortened to span a distance of 10.2 cms. A preliminary 

 test showed the natural period of such rings to be independent of 

 their orientations; they were therefore hung up in the knife edges 

 with open gap downwards. The top edge of the ring was always 



^ Abraham gives a correction term of the form 

 X 2 r . , ^ / 1 



1 



['"'• (.-i|)T 



4 log I 



where "»" is the order of the harmonic and C„ a compHcated integral. For "»" = 1 

 and the 300 cm. rod, X// = 2.018, which is too small. 



