72 



Rainfall, Natural Drainage, 



l^hese facts are important, and tell their own story so far as 

 averages can do. They seem to show a steady, and even con- 

 siderable decrease in the amount of rain, in whichever way we 

 take the periods. If we go a step further, and divide the period 



into ten of five years, the result 



T)ixGiixya\\-MeanrunnfalldurlmjPa^^^^^ js nearly similar, although we 



of Fire Years, from 181-i to 18Gi m- , -^ . ' ^• •, c 



clusice then perceive a peculiarity ot 



the distribution, which can be 

 best seen by the eye in a dia- 

 gram. In these short intervals 

 the diminution is not regular, 

 but there are alternate maxima 

 and minima ; the general result 

 being a decrease, as already 

 shown in the longer periods. 



This tendency to alternate 

 maxima and minima is, how- 

 26 ever, shown much more clearly 

 25 and decidedly if we compare 

 g , the rainfalls of the various years 

 placed before the eye as in the 

 preceding diagram (No. II.), 

 "' It will be evident that there is 

 a tendency to a succession of 

 maxima and minima, or rain- 

 falls alternately above and below 

 the average, and also of groups of maximum and minimum years. 

 Thus there is a group of maxima between 1814 and 1824, a group 

 of means succeeded by two maxima and one minimum thence 

 to 1841, and more recently small groups of maxima and minima. 

 It will also be evident that the tendency during the whole period 

 has been towards fewer and smaller maxima, and more, as well 

 as lower, minima. 



Both electric and magnetic storms or disturbances in the elec- 

 trical state of the earth, as exhibited in the atmosphere, affect the 

 rainfall in a very decided manner. It has been found of late 

 years that these have reference to the dark places on the surface 

 of the sun, which undergo a change in cycles of about eleven 

 years. During such periods, then, it might be expected that the 

 rainfall should be subject to useful comparison. Several cycles 

 require to be compared before any conclusion of value can be 

 arrived at ; but availing ourselves of the records of half a century 

 we shall at least have a starting-point. The result is shown in 

 Diagram No. V. It agrees precisely with all that has been 

 hitherto said. It is even more decided, and presents, as will be 

 seen, a very regular curve. 



21 



