56 MAmjRmG 



ment. The land should be flat and the soil similar in 



each experimental plot. The plots should have been 



similarly treated in regard to previous cultivation and 



manuring. 



The cocoa trees should be of the same variety in each 



plot, of like age, and be planted at a uniform distance 



apart. It is also of importance that the same system 



of pruning should have been applied to the trees in each 



plot. The plots should be uniform in area and each 



should contain the same number of trees. Nitrogen, 



phosphoric acid, and potash are the constituents in which 



the soil is most likely to be lacking. It may be deficient 



in one or even two, but most unlikely in all of these 



ingredients. To test for one constituent four plots will 



be necessary, including the control plot. The results 



in this case would not, however, be satisfactory if more 



than one constituent were lacking in the soU. It is 



therefore advisable to have eight experimental plots. 



Each should be at least one-fifth of an acre in area. 



Strips of land 88 yds. long by 11 yds. wide would be 



suitable, and, if the trees were planted at 15 ft. apart, 



each plot would contain 34 trees. At least two rows of 



trees should be left unmanured between each plot. In 



the subjoined table are given the weights of the various 



manures which it is suggested should be applied to each 



plot. 



No. Name and Weight of Manure 



of Plot. to apply per Plot. 



1. (Control plot) . No manure 



2. . . . 401b. Sulphate of potash 



3. . . . 80 ,, Basic phosphate 



4. . . . 40 ,, Nitrate of soda 



g /40 ,, Sulphate of potash and 



^80 ,, Basic phosphate 



g r4:0 ,, Sulphate of potash and 



\40 ,, Nitrate of soda 



7_ _ ^ r^O >> Basic phosphate and 



1 40 ,, Nitrate of soda 



{40 ,, Sulphate of potash, 



80 ,, Basic phosphate, and 



40 „ Nitrate of soda. 



The fertilisers should be mixed with dry soil and spread 



