64 MK. GEORGE C. SIMPSON ON THE 



Methods of Work* 



Potential Gradient. BENNDORF'S self-registering electrometer, with radium collector 

 attached, was set in action on September 28, 1903, and produced a nearly continuous 

 record of the potential gradient until October 1, 1904. Each day the curve for the 

 previous day was measured and the mean potential gradient for each hour obtained. 

 Tli is was done by first drawing a curve as smoothly as possible through the registered 

 curve, then five equidistant ordinates in each hour were measured, and the mean of 

 these five taken to represent the mean potential gradient during the hour. In 

 discussing the potential gradient for any place it is usual to use only observations 

 made during fine weather, neglecting all those which have been affected by any 

 atmospheric disturbance. This plan I also followed during the summer months (April 

 to end of September), for then the curves drawn by the instrument were exceedingly 

 regular unless there was actually precipitation taking place in the neighbourhood. 

 But during the winter the curves were so irregular, even on the finest days, that it 

 was quite impossible to decide whether a particular curve ought to be neglected or 

 not, so T used, during the winter, all the curves quite independently of the weather 

 This caused irregularities in the final curves, but has not, I think, affected the 



cl 



conclusion to be drawn from them. 



Iti^iitxtioii. The value of the dissipation, as measured by ELSTER and GEITKL'S 

 instrument, depends to a very great extent on the manner in which the instrument 

 is exposed to the wind. This is as it should be, for the actual dissipation from the 

 earth's surface (which the instrument is designed to measure) depends largely on the 

 \vind strength. In order that the instrument should measure the amount of dissipation 

 taking place from the earth's surface, it should be exposed to the same wind condition 

 as the general surface. This fact has not been fully realized by most observers. It 

 has been quite a common practice to shelter the instrument from the wind, either by 

 erecting screens or by observing close to a building, and in several cases the instrument 

 has been placed within a room close to an open window. Observations taken under 

 such conditions are of very little value : they are certainly of no use in comparing the 

 dissipation of one place with that of another, and at the best can only be used to 

 compare variations from time to time at the same place. In order that the dissipation 

 at one place may be compared with that of another, the instruments used should in 

 both cases be exposed to the full force of the wind, for wind strength is just as much 

 a factor in determining the dissipation as is the ionizatiou. For this reason my 

 instrument was only used in a freely exposed situation, where it was in no way 

 sheltered from the wind. This method also has its drawbacks, for with anything like 

 a high wind the leaves of the electroscope were so blown about that they continually 

 discharged the instrument by coming in contact with the case. Hence measurements 

 could not be made in very high winds, and so the mean values of the dissipation 



* For fuller particulars of methods of work and arrangement of apparatus see Appendix. 



