OBSERVATIONS ON DETERMINATIONS OF DROUGHT-INTENSITY. 83 



line, are the values dz 2 jdt, that is of the rate of absorption by the 

 soil, the curve would indicate that at first the whole fall — from 

 the commencement of the rain till the time ^ — was absorbed: 

 but that from that time on, only part was absorbed : and finally 

 that absorption was going on, even after the rain ceased at the 

 time t 2 , owing to the fact that the water lying on the surface still 

 contributed to the saturation of the soil right up to the time t z . 

 The quantity absorbed is determined by the area between curve 

 (2) and the axis of abscissae, and if there be no loss by evaporation 

 or otherwise, the difference of these areas would give the quantity 

 running away. These two curves serve merely to illustrate, in a 

 most elementary way, the nature of the invesigation. Curve (1) 

 is directly obtained by suitable observation of the rainfall : curve 

 {2) is deduced from all relevant facts. The ordinates of the first 

 are independent of one another : those of the second are not only 

 not independent but they are also continuously dependent on all 

 ordinates of the phenomena of which the instantaneous condition 

 is the resultant. 



22. Natural phenomena to be observed. — The following list of 

 phenomena to be observed, includes those usually registered at 

 meteorological stations and are all that are necessary for the com- 

 putation of the saturation of a soil, as the measure of 'drought- 

 intensity.' The abscissa in all cases is time : the ordinates should 

 be as specified : 



Rainfall Ordinates Rate of fall dz/dt. 



Air temperature „ Temperature r 



Velocity of wind „ Velocity w 



Humidity of air „ Humidity /x 



Solar radiation „ Temperature <r 



Evaporation of water „ Rate of Evap. de/dt. 



Deposition of dew etc. „ Quantity (if necessary as 



materially affecting result). 



23. General remarks. — The careful study of the phenomena as 

 affecting soils even under perfectly definite conditions would, as 

 I have already said, be elaborate, but would doubtless throw con- 



