converted these into a residual mass curve, and plotted it 

 on my diagram for comparison with the other curves. It 

 will be seen that the cumulative effect of the rainfall from 

 which the floods are derived, with reference to the mean, 

 produces a curve which clearly shows the wet and dry 

 periods of the Nile to be co-terminous with that of Lake 

 George. The change from dry to wet which took place in 

 1888, and the remarkable rainfall of 1887, which immediately 

 preceded that change, is very clearly defined in all the 

 curves, and in many others which I have been unable to 

 include in the diagram for want of space. 



I was so much impressed with the results obtained from 

 that portion of the Nile record in my possession, that I 

 immediately wrote to Captain Lyons, the Director General 

 of the Egyptian Survey Department, at Cairo, requesting 

 him to furnish me, if possible, with information which would 

 enable me to extend the curve backwards into the past. 

 He very kindly sent me his report, dated 1906, " On the 

 Physiography of the Nile River and its Basin," a most 

 valuable and exhaustive systematic exposil ion of the whole 

 subject. That portion of his report dealing with the 

 "Variation of the Nile Floods " is of very special interest 

 in this connection. I found that Captain Lyons had been 

 engaged on a similar investigation of the Nile River flood 

 records for traces of periodicity, and among numerous 

 diagrams, he prepared one 1 showing the variation from the 

 mean values of the Nile floods from 1737 to 1800, and from 

 1825 to 1903, in which the curve derived from the yearly 

 values, and that computed from the means of successive 

 five yearly periods, are compared with the sun spot curve. 

 Unfortunately there is no information available to fill in 

 the gap between 1800 and 1825 ; but I afterwards found 

 that it occurs in a portion of the whole record, where it is 

 felt less than it would have been if the gap had occurred 

 1 See p. 42. 



