138 DR. C. CHREE: ATMOSPHERIC ELECTRICITY POTENTIAL GRADIENT 



between air collected at the usual level and air taken from immediately above the 

 ground. 



The Ebert apparatus determines the charges per cubic centimetre of the free 

 positive ions and the free negative ions separately. The difference between these 

 gives p. The results obtained at Kew in 1911 and 1912 have been published monthly 

 in the ' Geophysical Journal ' of the Meteorological Office. The 1911 tables give the 

 number of ions as calculated from Sir J. J. THOMSON'S original value of the ionic 

 charge, viz., 3'4 x 10~ 10 electrostatic unit. The 1912 tables give the charges per cubic 

 centimetre in electro-magnetic units. For the purpose of calculation it is simplest to 

 employ electrostatic units, i.e., to multiply the numbers in the 1911 tables by 3'4 x 10~ 10 , 

 and the charges in the 1912 tables by 3xl0 10 . I have similarly dealt with all the 

 available data down to the end of July, 1914. There were several gaps, so that only 

 25 months' observations were available. Allowing equal weight to the months of the 

 several years, the values obtained for />x 10 9 for the 12 months, January to December, 

 were in order 



31, 54, 30, 65, 70, 54, 60, 65, 65, 48, 39, 39, 



all being plus. 



The arithmetic mean of the 12 monthly values gives 



P = +52xlO- 9 . 

 To utilise this result it is perhaps simplest to regard (l) as equivalent to 



0, .......... (3) 



where M is the free charge enclosed by the surface S, of which n denotes the normal. 

 Apply this to a tube of force bounded by 1 sq. cm. of the earth's surface, with the 

 other end on a plane area 1 metre above the ground. Obviously the upper section 

 will be sufficiently approximately 1 sq. cm. We get 



(dVfdz), = - (dV/dz), = , metre = 47T x 52 x 1 0- !i x 1 2 



= 0'65 5 xlO- 4 E.U. 

 = 0'65 6 x 300 x 10- 4 volts per c.m. 

 = 2'0 volts per metre very approximately. 



This signifies a decline of 2 volts per metre in the potential gradient, or practically 

 that the potential gradient deduced from the first metre in the Kew observations 

 should exceed by 2 volts that derived from the second metre. 



As will appear presently, the average value of the potential gradient on rainless 

 days at Kew is about 300 volts per metre. Thus we should have as an average, 

 employing r 2 and i\ as in 4, 



rjr, = 0'99 3 . 



