32 T. W. KEELE. 



Although in some of the curves (Plate 2, fig. 1) notably 

 Horsham and Adelaide, a series of waves at nineteen years' 

 intervals arc clearly defined, it will be observed in the Nile's 

 curve— whose period appears to be one of 171 years, built up 

 of three periods of 57 years, which in their turn are composed 

 of three periods of 19 years— that the nineteen years waves 

 are obliterated in the first great sweep made by the 

 accumulated rise from 1736 to 1781, and also in the decline 

 made by the accumulated fall from 1782 to 1839. In the 

 second great sweep of 57 years from 1839 to 1896, the Nile's 

 wave, having lost the energy it possessed in the former 

 wave, permits the smaller waves of which it is composed 

 to be seen, but they are hardly discernible. 



It is fair to Mr. Russell to remember that he had very 

 slender data to work upon in formulating his theory, but it 

 is quite evident from the quotation previously referred to, 

 that he had more than a suspicion that his nineteen year 

 period was involved in a greater one, but he was unable to 

 demonstrate it. 



On a first glance at the diagram, which looks a little 

 complicated owing to the variety of curves there delineated, 

 the slow secular change in the Abyssinian rainfall, which 

 results in the annual inundation of the Nile, may not be 

 immediately discernible owing to the unusual system I have 

 adopted in depicting it, by the construction of the residual 

 mass curve; but on a careful study of the diagram, the 

 truth, I hope, will become revealed to you, as it has been 

 to me, and you will then be better able to appreciate the 

 work previously done by Mr. Russell, who had no informa- 

 tion concerning the Nile, and to realise with what rare 

 judgment, amounting almost to prescience, he was endowed. 



In my diagrams (Plates /, 2) you will observe that 1 have 

 made the fluctuations of the water in Lake George, New 

 South Wales, a conspicuous feature, and that the annual 



