II.] ACTION OF NITRATE OF SODA ON CLAY SOILS 55 



idea that there is any direct attraction of water by 

 nitrate of soda remaining in the soil. The explanation 

 appears to be more complex. When a plant is feeding 

 upon a neutral salt like nitrate of soda, it takes up 

 rather more of the nitric acid than of the soda, leaving 

 behind in the soil some of the soda combined with 

 carbonic acid excreted from the root. Water cultures 

 in which plants are grown with nitrate of soda will 

 actually become alkaline to test-paper from this cause. 

 Now, a very small quantity of a free alkali, like carbonate 

 of soda, has an altogether disproportionate effect upon 

 clay ; the clay is deflocculated — i.e., the little aggregates 

 of very fine particles which cause the clay to crumble 

 down when dry and to allow water to drain through 

 it, are immediately resolved into their finest state of 

 division, and all the characteristic properties of clay are 

 accentuated. Deflocculation is effected mechanically 

 whenever clay is puddled or worked in a wet con- 

 dition, and all the features of puddled clay, which 

 is both retentive of water and impermeable by it, 

 which shrinks greatly in drying and then holds together 

 with extreme tenacity, are found in these soils when the 

 deflocculation has been brought about by a little dissolved 

 alkali. The fact that such deflocculation has taken 

 place may be illustrated by a very simple experiment. 

 Fig. 2 shows two large jars, each containing 3 

 litres of distilled water, in which has been shaken up 

 I gramme of the Rothamsted clay loam, in the one 

 case from a plot manured with nitrate of soda, in 

 the other, from the adjoining plot receiving ammonium 

 salts. It is obvious how much greater is the amount 

 of material remaining suspended in the jar con- 

 taining soil manured with nitrate of soda, which 

 means that this latter soil had been previously 

 brought into a more fine-grained and less flocculated 



