1 916. J The Third Indian Science Congress. ciii 



We could not be expected at first to progress with the same degree 

 of rapidity as in America, because we have to do a large amount of 

 research and experiment before we can demostrate improved methods on 

 a large scale. In America the advanced stage in the agricultural develop 

 ment of the Northern States supplied ready at hand the stock-in-trade 

 required for at once setting in motion the demonstration movement in the 

 backward Southern States. We are not so forward. Still we have 

 achieved enough with our small band of workers to show that the same 

 kind of work can be done out here and that all we require is expansion. 

 Given the means for this and a recognition of demonstration as an integral 

 part of a general scheme of education, we shall, by such a policy, lay the 

 best and securest foundations for the advancement of education as well a- 

 of the prosperity of the people. 



The Correlation between Rainfall and the Succeeding 



Crops 1 .— By S. M. Jacob, I.C.S. 



The aim of the paper is to establish formulae by which not only the 

 area sown with each class of crop can be forecasted, but also what will 

 be the yield of each crop at harvest per unit of area. In both cases the 

 attempt is to determine what are the quantitative relations of crops and 

 rainfall, and to make more definite what we already know as to the 

 dependence of plants on moisture. 



The first problem, the prediction of sowings, is very largely an 

 economic problem, and the effect of changes in prices, cost of produc- 

 tion and population must be taken into account. In the case taken in thi 

 paper, that of well-irrigated wheat, the first correction to be applied is 

 one depending on the number of wells. When this has been done and 

 the extent of sowings is calculated from the rainfall in August, September 

 and October, it is found, as was to be expected, that sowings fall off with 

 increased rainfall, and vice versa, and the extent of this falling off was 

 determined. The correlation co-efficient obtained is # 89, which is high 

 enough for very accurate predictions. A diagram was exhibited showing 

 for the last 30 years the concordance between the actual areas sown and 

 the amount given by calculations. 



The next step is the determination of yield, and in examining the 

 effect of rainfall in February, for example, on unirrigated wheat, it is 

 clear that the benefit the crop will derive in that month will depend on 

 whether the rainfall in September, October, November, December and 

 January has been good or not. If there has been an excess of rain in 

 January, for instance, it is obvious that an excess of rain in February 

 can do little good and may do harm. A numerical scheme was adopted, 

 based on a method applied with success to the yield of cotton in Texa* 

 by Kincer. A good agreement of actual and calculated values for the 

 failed areas of unirrigated wheat in the Punjab is obtained. The corres- 

 pondence was exhibited in diagrams. The method is thus valuable in 

 dealing with a very important practical problem. 



Agr 



P. H. Viok. 



l -By 



The Relation between Soil Bacteria and Fertility. — By C. M. 



Hutchinson. 



1 This paper will be publishe 

 the Agricultural Journal of India. 



