368 Mr Wilson, On the Measurement of the Earth-Air 



outer tube and to the capacity of the air condenser for a given 

 position of the compensator. When the compensator was in its 

 standard position, i.e. fully drawn out, the condenser tube was then 

 some distance beyond the end of the rod, so that the potential of the 

 tube might be varied largely without that of the rod being sensibly 

 affected. With the compensator in its standard position there is 

 thus no danger of an apparent gain of charge by the test-plate, 

 merely due to defective insulation of the Leyden jar carrying the 

 compensator tube. 



Let us suppose the operations described above for the measure- 

 ment of the charge on and current through the test-plate to be 

 performed. When the cover has beeu removed and the compen- 

 sator has been moved to bring the potential back to zero, the 

 condition of all parts of the system which includes the gold-leaf 

 test-plate and connecting rod will be the same as initially, except 

 that there is a charge on the upper surface of the test-plate pro- 

 portional to the potential gradient and an equal and opposite 

 charge on the rod of the compensator, the total charge on the 

 system remaining the same. The charge on the compensator rod 

 can be deduced from the scale reading when the standardising 

 measurements described below have been made. To maintain the 

 gold-leaf at zero potential the compensator may have to be moved 

 owing to either or both of two causes: (1) the potential gradient 

 may change, and (2) electricity may flow from the air into the 

 test-plate. An increase in the field (it is assumed to be of normal 

 sign) will increase the negative charge on the surface of the test- 

 plate, this requiring the compensator to be pushed further in till 

 there is a further equal increase in the positive charge on the com- 

 pensator rod. Under the action of the field, again, positive ions 

 are likely to move up to the test-plate, and the positive charge 

 entering the conducting system is prevented from raising the 

 potential by again pushing in the compensator to increase by 

 the corresponding amount the capacity of the sliding condenser, 

 so that the whole increase of charge is held on the compensator 

 rod. On putting on the cover the negative charge is no longer 

 held on the upper surface of the test-plate, and to bring the 

 potential of the system back to zero the compensator has to be 

 drawn out till its capacity is such that the positive charge on the 

 compensator rod when the potential is zero is equal to that which 

 has been gained by the system from the atmosphere. This charge 

 can then be determined from the position of the compensator. 



We have to consider the question to what extent any failure of 

 the insulation of the compensator may introduce error into the 

 results. The charge measured is, for a given reading of the com- 

 pensator, proportional to the potential of the compensator tube or 

 of the inner coat of the Leyden jar to which it is attached. A 



