AT KEW OBSEEVATORY, 1898 TO 1912. 137 



of observations made under favourable conditions, and not redetermined until some 

 known change is made in the apparatus. Whether this is satisfactory or not depends 

 on the special circumstances of each installation. It would not be wholly satisfactory 

 at Kew, where the discharge tube is long and may develop a sag or have to be 

 replaced. An incidental advantage of regular absolute observations is the check they 

 afford on the working of the apparatus. They help to disclose defects such as poor 

 insulation, and secure more careful attention to the electrograph. One of the reasons 

 for observing potential at 2- as well as 1 -metre height, was the fact that the discharge 

 tube of the water-dropper is fully 3 metres above ground level. There is little 

 experimental information as to the differences that may exist in the potential gradient 

 at different small heights above ground level, and there is no general agreement as to 

 the height interval from which the standard potential gradient should be obtained. 



5. If V denotes the potential, and z be measured vertically upwards, the potential 

 gradient (dV/dz) must satisfy the two equations 



d?V/<.tf + 4 ! irp = 0, .......... (1) 



and 



= 0, ......... (2) 



where p is the volume density, and or the surface density of the earth's charge. 



Unless p is zero the potential gradient will vary with the height, and if p be known 

 the extent of that variation can be calculated. Various forms of apparatus exist 

 which profess to measure p, and one of these the Ebert apparatus has been in 

 operation at Kew, with interruptions, since May, 1911. There are, however, reasons 

 for accepting the results obtained with some reserve. It is now generally recognised 

 that the Ebert apparatus takes no account, or only very slight account, of the slow 

 moving heavy ions discovered by LANGEVIN, though it seems to catch the light mobile 

 ions on the whole satisfactorily. McCLELLAND has found large numbers of Langevin 

 ions near Dublin, and Kew, from its proximity to London, may also not unlikely have 

 large numbers of them, at least with easterly winds. Thus the results derived from 

 the Ebert apparatus as to p will be incorrect unless the positive and negative heavy 

 ions are at least approximately equal in number. 



In spite of this uncertainty, it seems worth while calculating how the potential 

 gradient may be expected to vary with height at Kew, assuming no free charge in the 

 atmosphere except that resulting from the excess of the positive or negative mobile 

 ions caught by the Ebert apparatus. It will have to be assumed that the distribution 

 of ions is sensibly uniform within 2 metres of the ground because in general the 

 ions observed at Kew are from air at a fixed height, about 2 metres above the ground. 

 The only direct evidence in favour of the hypothesis consists of some experiments 

 made by Mr. GORDON DOBSON* which showed no certain difference in ionic contents 



* ' Proceedings Physical Society of London,' vol. 26, 1914, p. 334. 

 VOL. CCXV. A. T 



