DEPARTMENT OF TERRESTRIAL MAGNETISM. 301 



The best type of earth-current line would, from our investigations, seem to 

 be one having subterranean lines consisting of leaded rubber-covered cable 

 laid about 18 inches deep in insulating conduit. In addition to its much 

 longer life, this type of installation should be nearly free from thermoelectric 

 effects, should have better line insulation, etc. To offset these advantages, 

 however, the first cost is considerably greater than for the more common 

 aerial type. Furthermore, since no direct comparisons between aerial and 

 subterranean lines have thus far been reported by other investigators, it seemed 

 advisable that at the outset of our earth-current studies we should provide 

 lines for obtaining such comparisons. These should be of value not only for 

 our guidance in the future, but also should aid us in more correctly appraising 

 the results already obtained by other investigators. For the furtherance of 

 this purpose two miles, one mile on each arm of the earth-current system 

 at the Watheroo Observatory, are to be provided with both aerial and sub- 

 terranean lines, so that either pair of earth-plates may be connected through 

 the recording instrument by overhead and underground lines alternately at 

 short intervals. Another feature is a second and longer aerial line which 

 extends along each arm of the system for a distance of 2 miles instead of 1. 

 This will serve two purposes. One is to compare the relative merits of the 

 longer and shorter aerial lines and the other is that, when used in conjunction 

 with the shorter line, it will furnish the evidence necessary for determining 

 whether the simpler "three-point" systems will suffice, or whether a more 

 elaborate distribution of earthed points is needed. 



The recording instrument to be employed is a Leeds and Northup multiple- 

 point curve-printing recorder which can be used to full advantage on the 

 simpler three-point system, but is also applicable to systems of as many as 

 12 points. With special features which have been provided, this instrument 

 should meet all the requirements of the earth-current potential measurements 

 at Watheroo, unless the magnitudes found there prove to be very much 

 different from those that have been obtained elsewhere. The potentiometer, 

 employing as it does a null method, will practically eliminate polarization of 

 the electrodes. 



The selection of earth-plates or electrodes will involve further experimental 

 investigation. Nearly as man}'- types have been recommended as the number 

 of investigators using them. The nonpolarizing electrode, as our preliminary 

 tests indicate, probably introduces more spurious effects than it eliminates 

 when used in a fixed system. Some preliminary field tests were made in 

 August at the Cheltenham Magnetic Observatory, Maryland, primarily to 

 determine the requirements of instruments for special investigations in the 

 field but the opportunity was used also to observe electrode characteristics. 

 It was noted that marked changes in the potential difference between earthed 

 electrodes could be produced by the addition of water to the soil in their 

 immediate vicinity and by mechanical agitation of the electrodes. 



Comparisons in the laboratory of various electrode materials and of types 

 of electrodes, though as yet not definitive, would indicate that metallic 

 electrodes, probably of lead or iron, will prove most satisfactory in fixed 

 systems designed for continuous recording at observatories. Such elec- 

 trodes if used must, however, be checked periodically against some standard. 

 Investigations leading to the design of such a standard and the method of 

 using it are now under way. 



