38 TILE DRAINAGE. 



a tile drain can not "efcrato" much uphill. The area of very 

 decidedly greatest mortality lies northwest of the cellar- 

 drain, between it and the main drain A B ; that is, the part 

 where there are no drains at a??, except on the outer edges 



Still another fact : The row of trees northeast of the main 

 drain H I, in plat 1, has no lateral drain outside of itself , north- 

 east ; that is, it has only half" the benefits of drainage, and 

 that, too, with the main drain H I uphill from it. Well, in 

 this row 5 out of 12 trees have died, or over 41 per cent, while 

 in the rest of plat 1 (all thoroughly tiled, with drains on both 

 sides of each row), only 21 out of 163 trees died, or less than 

 13 per cent. 



Let me here restate these striking facts in more concise 

 form. 



PER CENT OF ORIGINAL TREES THAT HAVE DIED AND 

 BEEN REPLACED WITHIN 16 YEARS. 



Where tiled both sides of each row, 13 per cent. 



Where half tiled, that is, on one side of 



each row, - - - - - 36 to 41 per cent. 

 Where not tiled at all to speak of, - 63 per cent. 



It is fair to add once more, that the present thrift and 

 bearing capacity of the trees are about in the same ratio. 



EFFECTS OF TILE DRAINAGE UPON WHEAT. 



As to this point, the companion pictures, Figs. 9 and 10, 

 tell the story. In Fig. 9 (tiled) the wheat is nearly all headed 

 (June 5th of a very late season, 1891) ; is about four feet 

 high, or up to the waist of a six-foot man, and is thick and 

 thrifty. In Fig. 10 the wheat is a full week later, is just be- 

 ginning to head ; is about two feet high, and so thin you can 

 see the knees and almost the feet of the man who stands in 

 it. The wheat in Fig. 9 at harvest had 33 shocks per acre ; 

 and that in Fig. 10, 22 shocks per acre. I gave explicit orders 

 to have the two plats separated in stacking and threshing ; 

 but by a misunderstanding of the men it was not done, 

 greatly to my disappointment. The difference, I know, 



