"We have therefore to deal with a comparatively short period 

 variation in the meteorologi.-jd condition : hut this subject requires 

 further investigation." 



From these extracts from Captain Lyons' report it will 

 be seen that, under the system he adopted of showing the 

 variations from the mean values of the Nile floods "for a 

 period of 175 years during which the records have all the 

 appearance of being comparable mid reliable, no regular 

 alternation of high and low floods is to be found." 



By using exactly the same data, but treating it under a 

 different system, namely, by finding the exact variations 

 of the maximum readings of the annual floods above or 

 below the mean, and then ascertaining the cumulative 

 effect with reference to the mean, it will be observed that 

 a curve is produced which still defines, although not so 

 sharply, the alternations of high and low floods ; but the 

 cumulative effect of the variations from the mean is clearly 

 brought out, and slow persistent rises and falls become 

 observable, thus enabling not only the alternations of the 

 wet and dry periods to be determined, but also, what is of 

 the greatest importance, slow secular change, if such exists, 

 can be detected by this system better than by any other 

 that I am aware of. 



It has already been stated when dealing with the Lake 

 George curve, that there were good reasons for believing 

 that it represented a 57 years' period, made up of three 

 periods of nineteen years each, and on searching the Nile 

 curve to ascertain if such a period could be traced in it, I 



