ENGLISH AGRICULTURE. 91 



because it not only includes soils recently poured forth from 

 volcanoes during historic times, but also a large area of 

 ground which has been produced by lava streams from 

 extinct volcanoes. There are extensive areas in British 

 India, and even in Scotland, which owe their fertility to the 

 fact that they have been produced by the gradual crumbling 

 down and decay of lava. These soils have been already 

 noticed under the class of basaltic soils. 



This must bring to a conclusion the interesting questions 

 of soil formation and soil distribution. 



I have, in the next place, to notice certain methods at 

 our disposal for improving soils ; these measures are of two 

 descriptions, often spoken of as mechanical and as chemical. 

 The mechanical methods include the improvement of soils 

 by altering their mechanical texture. And there is then left 

 the methods of improving them by the direct addition of 

 fertilizing matter. I consider it more strictly true and a 

 better definition of these two methods to speak of them as 

 means of developing the resources of soils from within them- 

 selves; and in the second place, means of adding from without 

 to the wealth of plant food contained in the soil. 



With reference to the first class of improvements, land 

 drainage occupies by far the most important position. It has 

 been very properly considered that land drainage lies at the 

 foundation of all land improvement, because until wet land 

 is drained all other expenditures of capital are fruitless. 

 Neither deep tillage, clay burning, liberal treatment, or the 

 feeding of tons and tons of cake and corn can be profitably 

 undertaken. It is necessary, in the first place, that land 

 should be dry ; afterwards we may proceed as seems best to 

 carry out other improvements. The history of land drainage 

 is deeply interesting, but it is just one of those sections 

 which may be omitted at present, so that we may proceed to 

 the practical aspects of the question. I cannot pretend to 



I 



