CROPPING PATTERN AND WATER RELATIONS 359 



extent be derived from reasoning about fundamental relations than from 

 statistical evidence. 



If for all the crops the curves are fitted around the parabola by applying 

 a horizontal and a vertical shift, the partial curves comcide with the 

 parabola and the frequency observations appear as a scatter of dots around 



Frequency */o« 

 lOOr- V + 



CANARY GRASS (•) 



Fig. I. If the logarithm of the frequency of occurrence of a crop within the cropping 

 pattern is plotted against the silt content, this gives curves of different shape, which 

 may be considered as parts of parabolas. 



the parabola. The result is given in Fig. 2. The object of interest now is 

 not centred any more on the curve, but on the parameters .Vq and A which 

 appear in the graph as the distance of shifting of the co-ordinate axes. 

 Figure 2 represents the entire cropping pattern of this clay soil and the 

 effect of the clay content on this pattern. 



If it were desirable to distinguish between more than one production 

 parameter the formula to be used would be that of the normal multiple 

 correlation. It may be mentioned that the sum of a number of single or 



24 



