CROPPING SYSTEMS FOR TOBACCO 7 



to year. Also, there was a variation in the ingredients nsed in the mixture during 

 the period of the experiment. For example, in the first year cottonseed meal 

 furnished 70 per cent of the total nitrogen, and in subsequent years about 50 per 

 cent. There were variations also in the kind and proportions of the carriers of 

 inorganic nitrogen and in the quantity of fertilizer, 3000 pounds per acre having 

 been used during the period 1924 - 1927 and 3500 pounds during 1928 - 1931. 

 These variations could cause dirterences only in the whole experiment from year 

 to year or period to period, but could not be considered as causes of differences 

 within the experiment for any one year, since a uniform fertilizer treatment 

 was given all plots with a few exceptions noted later. Detailed data of fertilizer 

 mixtures are given in Table 2. Figures for nitrogen and phosphoric acid .express 

 the total quantities of these elements. 



Table 2 — Total Nutrients Applied in the Different Cropping Systems. 



(Pounds per Acre) 

 Nitrogen (N) Phosphoric Acid (PjOj) Potash (KjO) 



1 = Continuous tobacco, with and without (checlc) cover crops; and cash crop rotation. 



2 = Animal-husbandry rotation. 



3 = Continuous tobacco with manure. 



The Plan of the Principal Experiment 



This experiment was laid out on the chessboard plan which permits the dis- 

 tribution of the replicated treatments in such a manner that a comparison can be 

 made between the results of any pair or among the entire set of treatments. In 

 using such a plan, considerable replication of plots of a given treatment is neces- 

 sary. As will be shown later, the experimental error of field experiments of this 

 nature is high at best, and every feasible device must be used to reduce the error to 

 a minimum. 



The dimensions of the original plots were 60.5 by 36 feet, or 1/20 acre, ex- 

 clusive of a border of 6 feet around each plot. The border was fertilized, 

 planted, and cultivated in the sair.e manner as the plot proper, but its yield was 

 not taken. In the original plan of the experiment, possible later subdivision of 

 the plots to meet needs which might develop in the course of the experiment was 

 contemplated, and in 1926, after the first two years, several of the plots were 

 subdivided into halves. Still others were subdivided in 1928. 



The tobacco-setting machine was set to space the plants 21 inches apart in the 

 row. The average number of plants per 1/40 acre plot at the end of the season 

 for eight years was 198.3, against a theoretical 207.4 for perfect spacing and per- 

 fect stand. Rows of tobacco were 3 feet apart. With the exception of the two 

 rotation series, all of the treatments were replicated at least four times and most of 

 them five or more. 



