THE NATURAL PERIOD OF LINEAR CONDUCTORS 407 



accuracy of determination of the distance between two consecutive 

 values of dlldl = will not be sufficient unless very great care be 

 taken, a large line current supplied, and an indicator responding well 

 to dljdl (such as a square law thermocouple) be used. For the 

 attainment of greater accuracy an average over a number of nodal 

 distances must be used. In short, this method of measurement 

 requires a relatively long line for accuracy, up to the limit where a 

 deterioration of the maxima and minima has become pronounced due 

 to attenuation. At the current minima dlldl is great enough for good 

 settings but no meters of requisite sensitivity at the zero end of their 

 scales exist. 



Another and more sensitive method of observation of nodal distances 

 is to make use of the variation of line current as the total line length 

 is varied, particularly if both ends of the line be pure reactances and 

 the line conductors have an adequate copper content and very good 

 insulation. Coupling such a line weakly to a generator by merely 

 placing the generator in the neighborhood makes it possible to build 

 up very sharply resonant standing waves so that settings without 

 any particular precautions can be made to one part in 3,500. Of 

 course the point located is again one where dljdl = but the value 

 of A/, for a given value of A/, is very much larger. Moreover the 

 accuracy is not decreased by shortening the line * and, since the 

 resonance energy is then dissipated in a shorter length of line, the 

 corresponding increase in the resonance current makes the antinode 

 easier to locate. Although this method is very sensitive to energy 

 losses it is by far the best method of using a Lecher system. Essentially 

 it is nothing but a sharp "tune" observed and interpreted as space 

 resonance. In the present work the length of the Lecher system was 

 sufficient to observe four resonance maxima throughout the range of 

 wave lengths used, giving, by difference, three wave length readings. 

 These readings could readily be duplicated to one part in 3,500; 

 it is improbable however that the velocity of propagation along the 

 wires is within one part in 3,500 of that of free space so that this 

 precision is unusable, though comforting. This accuracy of setting 

 would be useful where small frequency differences or line constant 

 changes are to be observed. Since the measurements checked ad- 

 mirably from day to day and the line attenuation was low, it was 

 assumed that the line velocity was not affected by the adjacent ground 

 (110 cms. at lowest point) and was near enough to that of the velocity 

 of light to allow the line to be used as the basic wave length standard. 



^ A. Hund, Sci. Paper No. 491, Bureau of Standards, 1924, points out the same 

 fact. 



27 



