which produced the curve similar to that experienced 

 onwards from 1782. 



That an accumulated fall has been taking place is true. 

 It will be seen however that during the nine years which 

 have elapsed since 1896, the rate of fall has been much 

 greater, nearly twice as great as during the same period 

 following 1782, the intensity of the drought therefore in 

 the former case has been far greater than in the latter, 

 and if the long periods of accumulated rise preceding the 

 two crests be examined, it will be seen that they are not 

 exactly similar. It is possible, therefore, that repetition 

 of similar weather may not be taking place in this order, 

 namely, of 114 years, but that the period may be one of 

 even longer duration, namely, of 171 years, or three times 

 57 years, in which case the rapid fall which has been 

 experienced since 189G may have terminated in 1907, and 

 the curve may now be on a constant upward rise, similar 

 to that shown from 1736 onwards. 



That there is a considerable amount of probability in this 

 assumption being correct may be realised on an inspection 

 of the curves derived from two of the longest rainfall 

 records in the world, namely, that of Britain and Padua, 

 and comparing these curves with that of the Nile floods. 



Note.— This was written at the time when the informa- 

 tion I possessed concerning the Nile extended only to the 

 year 1905. I have allowed this portion of my paper to 

 stand as 1 originally wrote it, as it shows the difficulty that 

 then presented itself in determining whether the true 

 period of the Nile was 114 or 171 years. The Director of 

 Egyptian Surveys has, however, since then supplied me 

 with the heights of the Nile floods for 1906, 1907, 1908 and 

 1909, which have enabled me to extend the curve up to the 

 latter year. It will now be seen that my deductions were 

 correct, as the year 1907 proved to be the critical point of 



