These records of the behaviour of the water, entirely 

 derived from the rainfall, stored up in depressions having 

 no outlet, natural rain gauges in fact, corrected for evapora- 

 tion, are of the utmost value in an investigation to deter- 

 mine the question of periodicity, and as Lake Gracemere 

 has shown a similar fluctuation to Lake George, 1 thought 

 it would be of interest to refer to it as I have done. 



Turning again to the diagram (PUite 2, fig. l),and particu- 

 larly to the Adelaide curve, it is surprising to me that the 57 

 years' period, built upas it undoubtedly is of three periods of 

 nineteen years each, has not, up to the present time, been 

 detected in it. The reason perhaps is that by the ordinary 

 method of showing the rainfall by vertical lines to scale, it 

 is difficult to discern it ; but, when the rainfall returns are 

 converted into a residual mass curve, the truth becomes 

 immediately revealed. Take particular notice of that por- 

 tion of the Adelaide curve from 1840 to 1852, and compare 

 it with the portion from 1897 to 1909, and note the won- 

 derful agreement. The great drought ending in 1845, and 

 the subsequent rise for seven years to 1852, is faithfully 

 repeated 57 years afterwards in that portion of the curve 

 from 1897 to 1902 and on to 1909. With reference to the 

 1902 drought, and its counterpart in 1845, note the simi- 

 larity of the Adelaide and Nile River curves. 



At this stage of my investigation, I became curious to 

 ascertain whether the moon's south declination curve, or 

 the sun spot curve, would afford any clue as to causation, 

 if compared with the Lake George curve as extended, and 

 the other curves I was in possession of at the time. I 

 accordingly plotted on my diagram the moon's curve as far 

 as 1846, as shown by Mr. Russell, and then extended it 

 backward as far as the diagram permitted. Wolf's 

 proportional numbers denoting sun spot frequency were 

 obtained from Table xxvin, as published in the Encyc. Brit. 1 



» Ninth Edition, Vol".' xvi. p. 17*. 



