FOREWOED. 



The great advances made in agricultural science during recent years, by 

 the application of well-arranged series of experimental plots, has thrown some- 

 what into the background the importance of careful observations on the 

 ordinary crops grown, and the lessons which can be drawn from their varia- 

 tions. The laying down of a series of experimental plots and tabulating their 

 results at crop time is a comparatively simple matter, and the equally important 

 regular observation of the plots at intervals is more laborious and is not always 

 attended to. A well-known experimentalist has, indeed, asserted to the Miiter, 

 thit such observations are altogether ultra vires, and that the plots should be 

 only visited when sowing is taking place and the reaping done. AVhile some- 

 thing may be said for this point of view in permanent plots, when the results 

 are recorded over a long series of years, we cannot regard it as justified for 

 general work on our Agricultural Stations, when it is desirable to obtain 

 definite results within as short a time as possible. It has been found that 

 the periodic notes taken of experimental plots during growth, as was regularly 

 done at the Samalkota Sugarcane Station by the author, were often quite 

 sufficient to explain apparent anomalies in the results, thus saving years of 

 repatition, and it was usually found that these results could be forecasted 

 with tolerable accuracy some months before harvest. The present paper, 

 prepared in 1917 for the Lahore meeting of the Indian Science Congress, empha- 

 sizes this side of agricultural research. As will be seen in the context, it is not 

 intended in any way to discount the value of the experimental method, but 

 to explain the fact that there are cases when it is inapplicable and that, in 

 such cases, series of careful observations, although more difficult and laborious, 

 may be attended with useful results. To quote a remark in Nature 

 (p. 203, May 16th, 1918), "Dr. Balls' comments on the short article on 

 ' Cotton-growing Statistics' in the issue of this journal of April 11th, opens 

 up a wide and interesting feature in scientific research, namely, the value of 

 observed data and their interpretation." Owing to the lack of space, and the 

 general character of the work on the Cane-breeding Station at Coimbatore, 

 we have not been able to introduce much experimental work in the plots. 

 Observations have, however, been regularly carried out for the past six 

 years, and several Memoirs have already been issued, giving some of the 

 interpreted results of such observations. The present paper gives yet another 

 instance of such work, and it has been decided to publish it practically as 

 it stands, although obviously incomplete in some respects, in that there seems 

 to be little prospect of the longer Memoir projected being completed. It is 

 considered that the details given sufficiently demonstrate the method 

 proposed. 



Coimbatore, 

 October, 1918. 



