measurements of rainfall tit the various places selected, 

 as well as the heights of the Nile River floods, have all been 

 converted into residual mass curves. The term is not strictly 

 correct where evaporation has not been allowed for; but the 

 fluctuations of the water in Lake George, when drawn to 

 scale as shown in the diagrams, make a curve representing 

 the true residual mass, for the reason that the lake may be 

 regarded as a large natural rain gauge, having no outlet. 

 The water stored in its basin is the residue of the total 

 amount which has fallen upon its catchment area, after 

 evaporation has done its work. If the levels of the water 

 surface, therefore, can be relied upon throughout the whole 

 period under review, the curve derived from it is a better 

 means of detecting periodicity, if it exists, than by any 

 artificial gauging not corrected for evaporation. It 

 appeared to me that if the measurements ol rainfall at the 

 various stations, and also the heights of the Nile floods, be 

 reduced in such a manner as to show the accumulated rise 

 and the accumulated fall, above or below the mean, the 

 curve thereby derived would be better for comparison with 

 the Lake George curve, than by any other method. The 

 results are now before you, and you will be able to judge 

 whether or not it is preferable to the other methods 

 ordinarily in use. I do not claim any originality for this 

 system. It was adopted by me after studying a report by 

 the Commissioners appointed to inquire into the New York 

 Water Supply, wherein I noticed that the method had been 

 used by Mr. Walter E. Spear in his investigation of the 

 Long Island Water Supply. The following is an extract 

 from his report:— 



"Since the height of the water table— on which depends the 

 delivery of the ground water on Long Island — represents the 

 cumulative effect of the rainfall for many years, it was appreciated 

 that the ordinary method of considering the maximum delivery 

 of a watershed to depend upon the rainfall during the driest 



C-May 4, 1910 



