noticed that the lowest point in the whole curve occurred 

 in 1839, when it was evident that the Nile curve was in 

 exact agreement with that of Lake George, both curves 

 culminating in the maximum intensity of drought in the 

 same year (1839). 



The highest points in the Nile curve occur at two places, 

 one on each side of the lowest point. The distance in time 

 from it being closely approximate, in one case to 57 years 

 of accumulated fall, and in the other to exactly 57 years 

 of accumulated rise. Moreover, when it is seen that the 

 crests are each preceded by long accumulated rises over 

 similar features of the assumed Lake George curve, one 

 can hardly avoid the impression that such peculiar agree- 

 ment in the variation of the curves cannot but result from 

 the same mysterious influence whatever it may be. 



The fact that the lowest point in the curve in 1839 is not 

 exactly equidistant in time from the crests may, perhaps, 

 on further investigation of the Nile records, be found to 

 have occured through an error in converting the original 

 observations, which were made according to the Coptic 

 (Julian) Calendar, into the corresponding Mohammedan 

 years. Captain Lyons in his report mentions this difficulty 

 in a footnote, and explains the method adopted in endea- 

 vouring to obviate it. He does not place the same amount 

 of reliance on the accuracy of the first portion of the Nile 

 record from 1737 to 1800, as from 1825 to 1905. 



The interval of time between the two crests shown on 

 the curve of the Nile floods, being apparently 114 years, 

 made up of approximately equal periods of 57 years, one of 

 accumulated rise, and the other of accumulated fall, would 

 seem to indicate that secular change will occur in that 

 order, in which case the curve, having recently attained to 

 the maximum in 1896, there should subsequently be a 

 repetition of the weather on the Abyssinian table land 



