388 MR. S. M. JACOB ON 



It is not contended that this will determine accurate values of the probable rain- 

 fall, but it would seem not impossible that it will give first approximations. 1 



The diagrams will now be understood without further comment. 



Here this paper must be brought to a close. 



It is hoped that the conclusions reached will not be applied without appreciation 

 of the limitations to which they are subject. 



In the future it is trusted that some of these limitations will be removed. 



In conclusion, I wish to express my thanks to Miss G. M. McLaren for her kind 

 help in Part III of this paper. 



1 The root-mean-square of the differences of the value of the product from the chosen harmonic curves is as follows, 

 the units being inches-squared. 



R. M. S. S. D. of points about mean straight line. 

 Kohat (curve I) 12-9 14-6 



Lahore .. 10-5 11*4 



Ajmere . . 8-8 9-0 



Beawar (curve 2) 5-8 6-6 



The improvement ir. the fit is small, but definite. It is hoped that harmonic analysis may be usefully employed in 

 getting better results. 



