30 



lenis and testing the result on the battery of chips of different sizes, 

 best for diffusion, and the balance of the time iu regular working. 



A lot of Kansas Orange seed was selected and distributed among 

 twenty different farmers, thus repeating the experiment described under 

 season 1881, except that Kansas Orange of the finest quality was used 

 instead of Amber. The result was high and low test canes and large 

 and small tonnage. 



It is but just to say that many of these farmers had no knowledge of 

 cane raising and followed their own notions. Those who had knowledge 

 of our work and some experience raised high-test canes and large ton- 

 nage. 



This season completes the circle of observations and records of crops 

 for nine years. The data can be summed up, which shows the action of 

 fertilizers on large masses of c-me as it has been received at the sugar- 

 house, and the proper and safe dates for planting each variety are de- 

 termined. This will explain and answer many of the criticisms which 

 have been published from year to year by parties who only saw this 

 work from one season's stand-point. The following deductions are made 

 from the analysis of more than 38,000 tons of cane, and cover a period 

 of nine years. This table will be found convenient for reference, under 

 the heading of season 1880 to 1888, inclusive. It must be borne in 

 mind that these facts will only strictly apply to this climate and this 

 soil 5 but until it can be proved that they will not apply elsewhere it 

 will serve as a guide, and should be interpreted by taking into considera- 

 tion the fertilizers used, the variations of the seasons, and the nature of 

 the plant. These conditions are fully described. 



Summary of record for nine years. 



