Lysimeter Experiments 



25 



An adequate discussion of the results of these experiments would require 

 an accurate knowledge of the meteorological data for the place and duration 

 of each experiment. This is in most cases difficult to obtain at this late 

 date, but if such data had been published in connection with the reports 

 cited above these might have made possible a more satisfactory inter- 

 pretation of the experimental data. For that reason there are published in 

 the appendix of this paper such meteorological observations as are available 

 for the period included by the investigation (tables 4-6, pages 96-99). 



Relation of flow to precipitation and season 

 In figm'es 4 and 5 are shown the monthly drainage flow from tanks 

 4 and 8, respectively, and the rainfall per month, calculated in liters 



320- 



sao- 



Ra/'nfa//- 

 Drainaqe 



Fig. 4. relation of drainage to rainfall by months for the unplanted tank 4 



per tank. As these two tanks have always been kept free of vegetation, 

 they afford a means of studying the relation of flow to precipitation and 

 to season. The graphs for rainfall begin with January, 1910, and record 

 the total precipitation for each month up to and including April, 1915. 

 The drainage begins with May, 1910, and extends to the same date as 

 does the precipitation. 



The line for precipitation is usually higher than that for drainage. It 

 will be remembered that about three-fourths of the rainfall for the five 



25 



