236 



SCIENCE. 



[Vol. IX., No. 214 



ments made by me in May, 1886, confirm this general 

 statement, although, the actual values would vary 

 greatly from day to day. Thus, with two collectors, 

 on one date I obtained as mean values, for 80 feet 

 elevation, 150 volts ; for 55 feet elevation, 40 volts ; 

 while on another date tho. values for the same eleva- 

 tions were resijectively 300 volts and 100 volts. 



Professor Exner (^Repertorium der Physik, xxii. heft 

 8, 451) gives the results of some experiments of a 

 similar nature made about the same time, which show 

 the potential gradient to be of uncertain value, and 

 influenced largely by the proximity of buildings and 

 walls. The following values for the potential were 

 obtained with a water-dropper in an enclosed 

 court : — 



Two TTipitrAqfrnm wan (Height (metres) 5 10 15 30 



iwo metres trom wall -^ Potential (volts) 2 7 17 48 



Tn pfiiitrp of fmirt /Height (metres) 5 10 15 20 



m centre ot court (Potential (volts) 5 11 32 68 



From measurements made with small balloons 

 filled with hydrogen gas, Exner obtained, for the 

 potential in free air, these values : — 



Height (met.) 17 18 20 21 22 34 25 27 30 34 40 48 

 Poten. (volts) 100 110 |J^^ 130 160 160 160 170 |g^^ 250 280 350 



from which 



a n metre 



These values were obtained with a burning match. 

 According to Pellat (Oomptes rendus, c. 1885). the 

 collecting efficiency of the match, compared with 

 water-dropper and flame, is in the ratio of 1 to 5 to 

 10 ; so that, for comparison with the observations 

 made here, where a water-dropping collector is em- 

 ployed, we have as a value for the electric force, dur- 

 ing calm fine weather, 



metre 



Another set of observations, made on an exposed 

 mountain-side, gave these results : — 

 Height (metres) 3 5 6 7 12 14 18 19 20 25 30 

 Potential (volts) 110 |S[^^ 210 ||^0 |^^0 480 ||^° 550 660 830 970 



or there is a linear potential gradient, but with a 

 higher value than in the preceding experiments. 

 Supposing a water-dropper to have been employed 

 instead of flame as the collecting agency, we have 

 the value 



d n metre 



It is evident, then, that this value of the electrical 

 force of the atmosphere is uncertain, and determined 

 largely by local surroundings. It is also further af- 

 fected by the conditions of temperature and relative 

 humidity, and, as intimated, by inconstancy of the 

 collecting agencj'. In working toward that ' electro- 

 geodesy' which Sir William Thomson has proposed, 

 we must determine and allow for these and doubtless 

 other influences. By taking the mean of many ob- 

 servations made at different times, the influences of 

 temperature and humidity are to some extent 

 avoided. As said above, the following observations 

 were made simultaneously, in 1886-87, at the top of 

 the monument, 500 feet above the ground, and at 

 the signal office, at an elevation of 50 feet. The 

 instruments used were modifiedMascart electrometers, 

 and large water-droppers with nozzles of the same 

 size. Similar methods and adjunct apparatus were 

 employed at both places. The values in the follow- 



ing table appear to be too small, judging from the 

 results quoted above. But it is to be remembered 

 that these observations are made in both cases from 

 buildings, and the points in air at which the collect- 

 ing stream breaks a way are not very distant from the 

 side of the building. 



Values of electric force of the attnosphere. 



(a) On this date some of the values at the lower station 

 were below the zero, i.e., negative : 69 observations gave 

 positive indications, averaging 38 volts, and 18 observations- 

 gave negative values, averaging 31 volts. The negative 

 values have been subtracted from the positive, and th& 

 remainder divided by the total number of observations. 



(&) JSfot simultaneous. 



(c) At both stations during rain the observations continued 

 for some little while negative. 



(d) As in (a). 



We have, therefore, from the above table, a mean 

 value of the potential for the top of the monument 

 of 637 volts, and a value of 



J7 = 4.33-^. 

 d n metre ' 



and at the lower station a mean value of the poten- 

 tial of 124 volts and a value of 



IT =8.43^. 

 a n metre 



Therefore it would seem that the mean value of the 



potential at the upper station is about five times that 



at the lower station. Among the observations, I 



find one striking confirmation of this ratio. On Nov. 



3, 1886, if we multiply the results obtained at the 



lower station by 5, we shall obtain approximately a 



duplicate of those at the upper elevation ; this for a 



series extending from 11 a.m. until 3 p.m. In some 



respects this date was most satisfactory, being a dry, 



somewhat hazy, autumn day, with light southerly 



winds, and sky about half covered with ill-defined 



cirro-stratus clouds. The electrification at the top 



of the monument was sufficient to give a spark a 



millimetre in length. 



These experiments were begun under the direction 



of Prof. T. C. Mendenhall, to whom, and to Col. 



T. L. Casey, of the Engineer corps, U.S.A., more 



than acknowledgment of kindness is due. 



Alexandeb MoAdie, M.A. 



