Presidents Address. iii 



given instant, a definite position depending on the earth's field at 

 that place at that instant. If we consider the magnetic data for a 

 given place, as time goes on we find well-marked daily and annual 

 variations, just as in the case of the meteorological elements. These 

 diagrams show this for Cape Town. Further study of the magnetic 

 elements at one place shows, however, a new phenomenon. If we 

 consider the declination, for example, we see that there is another 

 and slow change going on. This is the secular variation, whose 

 period would appear to be about a thousand years. 



The study of atmospheric electricity may be approached from two 

 sides. A charged insulated conductor, for example, will not — even 

 if it is supported by a perfect insulator — retain its charge indefi- 

 nitely ; and the rate at which the charge is lost varies with time at 

 the same place. The systematic study of this loss in a free atmo- 

 sphere has only been carried on for a few years, but it is already 

 evident that here also there is a daily and an annual period. You 

 will see this very clearly in these diagrams, which give the result of 

 observations by Mr. Lyle at Bloemfontein. 



A second method of studying the electric state of the atmosphere 

 is to observe the difference in electric intensity between a point at a 

 fixed distance above the ground and the earth itself. In this case 

 also daily and annual variations are well-established facts. 



If instead of considering the change at one place as time goes on 

 we consider the value of the elements at the same instant at every 

 part of a given region, we would then have a survey of that region 

 for meteorology, magnetism, or electricity, as the case may be. 



The results of such surveys are embodied in maps. In a given 

 region, for example, all places with the same temperature may be 

 joined together by a line called an isotherm. A number of isotherms 

 corresponds to a temperature contour. The particular isotherms 

 which we obtain by taking the mean of the temperatures of a 

 sufficient number of years, and which therefore does not show the 

 effect of periods such as the daily or the annual, does not differ 

 jnuch, if at all, in position if it is derived from the mean of the tem- 

 peratures of another sufficient number of years. In other words, 

 there is no sign of a secular variation of temperature. It is impos- 

 sible as yet to say whether this is due to insufficient and non-homo- 

 geneous data or to the fact that the region under consideration is 

 actually not becoming slowly hotter or colder. A map of this nature 

 for South Africa would show irregularities due to the relative posi- 

 tion of sea and land and to the different heights of the latter if the 

 effect of the height were not eliminated by a special calculation. 

 Another way of showing the temperatures in a region is to calculate 



