38 STUDIES IN INDIAN TOBACCOS. 



render the land as uniform as possible. The plots are all small, 

 carefullj' levelled and are well drained and cultivated, and for 

 most crops they present an ideal experimental basis. In the 

 present investigations, a certain amount of trouble has been 

 experienced even on these experimental plots. Slight local un- 

 evenness of the ground due to ploughing, a difference in the sub- 

 soil drainage, and the proximity of a hedge have all had an 

 effect. It is needless to say that all cultures which could 

 possibly have been affected have been rejected. The cultures 

 which have to be directly compared are grown on the same 

 plot in lines and the two parent types are grown at both ends 

 of the plot, and also in the centre. In this way, should there be 

 an}^ slight change in the conditions from one side of the plot to 

 the other, it would be indicated by the range in variation of the 

 parent forms. The impossibility of obtaining a large piece of 

 land with uniform drainage and soil must always limit the 

 number of cultures grown, even if the amount of work entailed 

 did not do so. One other important point must be mentioned, 

 namely, the time of transplanting. If many plants die after the 

 first transplanting and have to be reset, the replaced plants, 

 even though only a week later in planting, always remain behind 

 the earlier ones. By adopting a system of furrow irrigation, 

 and using great care in removing the seedlings from the nursery 

 boxes, the loss in transplanting in the experimental cultures 

 has been reduced to a minimum. Onl}^ one replacement is 

 carried out two or three days after the first transplanting. 

 Should others die, their places remain blank, but the system of 

 transplanting adopted has proved so successful that the number 

 of such blanks is very small indeed ; the number of deaths before 

 the first replacement is generally not more than one per cent. 



The methods of crossing and raising the self-pollinated seed 

 are those in ordinary use, and need not be specially mentioned. 

 Full details have been given in a former paper.' The only point 



^ Howard and Hcnvard, /.''. 



