PERCOLATION, AXD EVAPORATION. 



43 



free from vegetation, in a gauge ouo-tliousandtli of an acre in area, 

 at Rothamsted, during the twenty-one years ending 31st JSEarch, 

 1892, and the following general results may be deduced from it : — 



These results do not represent anything which takes place in 

 nature throughout the year in this country, for nowhere have 

 we soil in a natural state which is constantly free from vegetation 

 of any kind ; but the experiments are instructive in showing 

 the maximum percolation possible through ordinary soil with the 

 greatest artificial aid to it in the frequent hoeing of the surface 

 in order to keep it quite free from vegetation. 



The table is also of value in this inquiry in another respect. 

 No other published results of experiments on percolation are 

 of a more recent date than 1884, while here I am able to give 

 the results up to the date at which all my rainfall tables terminate, 

 viz., the 31st of March, 1892. If the twenty-one years are divided 

 into three equal periods of seven years each, it will be found that 

 during the first seven years the percolation was 83f per cent, 

 of that of the whole period, during the next seven years 117 per 

 cent., and diuing the last seven years 99;| per cent., or practically 

 the average of the whole period. 



The mean annual rainfall during the twenty -one years covered 

 by this table having been 30-11 ins., and the mean at Eothamsted 

 during the thirty-nine years 1853-92, as shown in Table VI, 

 having been 28'46ins., the mean annual percolation during the 

 shorter period must have been much greater than it would have 

 been during the longer period. But there is a still greater diff'erence 

 between this annual rainfall of 30-11 ins. and the mean annual 

 rainfall in Hertfordshire for the last half-century, viz., 26 inches. 

 The problem is, therefore, to determine, if possible, what would 

 probably have been the percolation at Rothamsted if the rainfall 

 there had been the same as the county average. 



"We are not justified in assuming that all the rain which falls 

 in wet years in excess of that which falls in dry years percolates, 

 or that it all evaporates, nor that the percentage of percolation and 

 evaporation to rainfall is the same, with a varying amount of rain. 

 It therefore seems to me that the only way in which a satisfactory 

 conclusion can be arrived at is to analyze this table, computing the 



