187 



in the number of canes continued into August, but at time of harvest 

 the amount of canes in each plat was almost identical, the plat from 

 rich seed having 875 canes against 870 in the other. Fourteen of the 

 canes in the rich seed plat died before reaching maturity, and seven in 

 the poor seed plat. In point of number of canes grown or those lost 

 before reaching maturity no preference can be given to either plat. 



I will now call attention to the relative amount of cane from the 

 two plats. It will be remembered that last year, while no actual weights 

 were made, it was remarked that the cane from the rich seed gave a 

 larger, healthier- looking stalk, this being so very pronounced that 

 there was no mistaking it. This present year all the samples brought 

 in were weighed, and as the same number of canes were taken from each 

 plat at every sampling, and these samples extended through the whole 

 length of the rows, a very good idea can be formed of the relative 

 quantity of cane This is best expressed by giving the average weight 

 per stalk. For the cane from poor seed this was 2 1 58 pounds, and for 

 the cane from the rich seed 2 . 42 pounds, a showing against the rich 

 cane seed. While it may be that each year we will have a return in 

 quantity similar to this, I am at present inclined to think that the rich 

 cane will in the end prove the larger one 



" It is true that with sorghum and beets the medium-sized plant is 

 the most satisfactory one to grow for sugar ; yet I believe that it could 

 not in the same way be said of the*e that the smaller or medium sized 

 seed are as satisfactory for planting as the large ones, containing, as 

 they would, a much greater amount of starch to be transformed into food 

 for the young plantlets. So, I believe, it will be with sugar cane, and 

 that the larger healthier stalks will, in a series of years, produce the 

 thriftiest canes, for I have continually noticed that in the selections the 

 rich canes are the larger and better stalks. In three of the samples 

 taken the weight of cane from the rich seed exceeded that from the poor, 

 the other four samples giving opposite results. Also it was noticeable 

 that at one end of the rows one plat contained the larger looking 

 cane, and at the other end the other plat did, and the samples taken 

 corresponded to this appearance. Certainly, from the limited trials made 

 here, it would not be the part of wisdom to assert positively 

 whether the rich cane seed will give a larger or a smaller cane, as the 

 two years' results have been contradictory in this particular. Such con- 

 tradictions, however, are to be expected in field agricultural experiments, 

 and it will take the average results of a number of years to 

 furnish ultimate proof. 



" We now come to the most important part of the work in judging 

 of its utility, viz., the analytical results. Seven sets of analyses were 

 made, and then it had become so late in the fall that it was deemed ex- 

 pedient to make the selections for planting, and as this took all the 

 canes it stopped further analyses. The last analyses were made on 

 November 12th. These samples were, with one exception, taken from 

 directly opposite parts of the two rows and contained the lame number 

 of canes. The one sample taken differently was during the time the 

 selections for further planting were being made, and consisted of every 

 thirtieth cane as the plat was being ground. This method of cutting 

 out sections of the row in sampling standing cane for comparison of 

 different plats I have found to be the most satisfactory tried. It is much 



