274 



addition of phosphate and potash increased the yield to 22 lb., a 

 gain of 2l ft), per tree. Dried blood gave a yield of 24^ ft. a 

 gain of 4I pounds over no manure, or 2| ft), more than phosphate 

 and potash. The combination of dried blood with phosphate and 

 potash increased the yield to 28f fta. per tree a gain of 9 ft), per 

 tree over no manure, or 6f ft), over phosphate and potash, and of 

 4I ft), over dried blood. The combination of the two sets of 

 manures has given greater increments than the sums of the in- 

 crements from either singly, thus pointing the necessity for 

 general manuring, i.e., for manure which will supply potash, phos- 

 phate, and nitrogen. The changes which have taken place in the 

 soil during the previous three or four years probably account for 

 the relative effects of the manures in this third period. 



The plot mulched with grass and leaves, the sweepings of the 

 lawns at the Botanic Station, is a very interesting one. In the 

 first period this plot, though giving a greater yield than the no- 

 manure plot, fell far behind the plot receiving dried blood ; in 

 the second period it again exceeded the no-manure plot and its 

 yield was practically equal to, or a little better than, the yield of 

 the dried blood plot; while in the third period C1905) it has far 

 surpassed all the other plots and has given a yield 66 per cent. 

 greater than that obtained from the no-manure plot. The soil of 

 this plot is in better condition than the others, the surface soil is 

 moister and darker in colour, while the trees have a better surface 

 root development. 



" This method of manuring by means of mulches of grass and 

 bush is evidently the proper course to adopt in Dominica, where, 

 owing to the large supplies of the required material which are 

 available, the work of manuring can be carried out efficiently. 



" These experiments again emphasize the desirability in the 

 tropics of following agricultural methods which lead to the 

 conservation of humus or vegetable matter in the soil. In most 

 cases, if these methods are conscientiously adopted, sufficient 

 supplies of plant food will be conveyed to the soil to obviate the 

 necessity of bujang artificial manures, 



" These experiments should be carried on for a number of years, 

 when further interesting results may be expected. It is probable 

 that the plot mulched with grass and leaves will retain its vigour 

 and productiveness for a much longer period than the others. 



"As 100 ft), of wet cacao are found to yield 42 ft), of dry cacao, 

 and as the trees are planted about 18 feet apart, or at the rate of 

 134 trees per acre, approximate calculations may be made as 

 follows* : — 



* Some vacancies occur in the plot so that the calculations have preferably been made 

 on the yield per tree rather than per acre. The ca culations per acre based on these are 

 necessarily only approximations, but they are made for the sake of more ready appre- 

 ciation of the commercial bearing of the experiments. 



