﻿EQUIPMENT AND WORK OF AN AERO-PHYSICAL OBSERVATORY. 23 



states that from 1883 to 1888 fifteen cases of Saint Elmo's fire had been 

 observed. All occurred during the night-time, indicating that there is 

 no electrometer at the observatory and no means of determination other 

 than, visibility, from September to February. On one occasion it was 

 heard during the daytime. All cases were noticed during the prevalence 

 of well-marked lows, generally about six hours after the center had passed. 

 Observations of the Potential. — The first source of error in determining 

 the true potential of a point in air is to be found in the bending of the 

 equipotential lines by the walls of buildings, the sides of mountains, hills, 

 etc. In a rectangular court fifteen metres wide, with walls twenty-five 

 metres high and forty metres long, it was found by Exner ; (see " Reper- 

 torium der Physik," xxii, heft 7) using a collector suspended by a silken 

 cord and connected with a quadrant electrometer in such a way as to 

 be readily moved up or down and from one side to another, that the 

 value of the potential varied. The contour of the potential surfaces 

 was approximately that shown in the accompanying diagram. With 

 the collector two metres from the wall, at heights of 5, 10, 15, and 20 



N^\\\'^\\V 



metres, the potentials w^ere 2, 7, 17, and 48 volts respectively. With the 

 collector in the centre of the court, at tlie same heights, the values of the 

 potential were 5, 11, 32, and 68 volts. The determination is open to 

 criticism, however, in this : that, the potential being at times exceedingly 

 variable, no method in which but one collector is emploj^ed can give 

 conclusive results. The experiment should be tried with two similar 

 collectors. The experiments made in Washington a few years ago by 

 the Signal Service (see Mendenhall, Memoirs of the National Academy) 

 iire therefore preferable, inasmuch as two electrometers calibrated to 



