PRACTICAL INFERENCES. 121 
periments, although not in the manner that might at 
first have been supposed. Leguminous plants, such as 
clover, beans, vetches, though containing so much nitro- 
‘gen in their composition, are not only not specially bene- 
fited by nitrogenous manures, but they absolutely leave 
the land richer in nitrogen than it was before (Lawes 
and Gilbert), and thus prepare it for the growth of grain 
crops, which, though chiefly starch-producing, are yet 
specially benefited by nitrogenous manures. 
In growing beans and wheat alternately at Rotham- 
sted, it was found that eight crops of wheat grown al- 
ternately with beans supplied nearly as much produce 
(grain), and nearly as much nitrogen in that produce as 
were furnished by sixteen crops of wheat grown without 
manure. Here, then, the manure supplied to the beans 
not only favored those plants, but left a residue in an 
available form for the wheat. 
Botanically, the good effects of rotation are dependent 
on the variations in the mode of growth and in the in- 
ternal structure of roots, which allow of different layers 
of soil being utilized for plant-food, while the specially 
different requirements of different classes of plants obvi- 
ate the exhaustion of any one ingredient, and give time 
for the accumulation of fresh supplies. 
Improvement of Cultivated Plants.—This has already 
been alluded to, but its importance justifies repetition, 
the more so as to a considerable extent it is a matter that 
the farmer can do for himself. A series of small experi- 
mental plots might well be instituted on every farm. 
The first and perhaps most general use to which such 
trial grounds should be put, would be to test the quality 
of purchased seed, and ascertain what proportion might 
be expected to grow under different conditions. Other 
experiments should be devoted to the purpose of ascer- 
taining what particular varieties are likely to do best in 
6 
