THE SOIL AND ITS EFFECT ON PLANT GROWTH. 23 



hold them together in somewhat the same way as flesh clothes 

 a skeleton. If we are to be able to modify the structure of 

 the soil, therefore, we must first learn how to modify its colloids, 

 and an important object of the present investigations at Rotham- 

 sted is to find out precisely what the colloids are and how they 

 can be controlled. One experiment which was recently conducted 

 may be mentioned, as it shows the great importance of colloids 

 in the soil. In the experiment, a careful comparison was made 

 of the rate of evaporation of the moisture contained in soils of 

 widely divergent type, namely, light sandy soil from Woburn and 

 loam clay from the Rothamsted farm. The results showed that 

 the effect of manuring in influencing the rate of evaporation was 

 less important than the proportion of clay particles (roughly 

 equivalent to colloids) present in the different soils. At present, 

 investigations into the nature and properties of colloids are being 

 rontinued with especial reference to their behaviour in contact 

 with different liquids, and liquids containing varying quantities 

 of salts. 



A reference may here be inserted to research on what is known 

 as " flocculation " in soils, which is being carried out at the 

 University of Leeds. By flocculation is meant the gathering 

 together of the small particles of the soil into " crumbs," thereby 

 opening the texture of the soil and improving its tilth and ease 

 of working. Hitherto it has been thought that the improvement 

 in the tilth of clay soil as the result of liming was not due to the 

 lime which was added, but to the soluble carbonate into which 

 the hme was converted. This view has been shown to be un- 

 founded, and the work at Leeds shows that the improved tilth 

 is due to the direct action of the lime. The mechanism of this 

 action has been investigated, and as a result of the work it is hoped 

 to develop methods for the improvement of soils which are 

 " heavy " on account of the presence of fine silt (not clay), and 

 whose texture is not improved by liming. Such soils occur very 

 • ommonly on Coal Measure soils in the north of England, the 

 Midlands, the Weald, and on the Lincolnshire warp-lands. 



To return to Rothamsted. It will be understood that labora- 

 tory research into soil properties must — unless largely increased 

 funds and staff are available — be limited to the investigation 

 of a small number of soil types, and that the effect of varying 

 climatic conditions can only rarely be compared. But before 

 practical conclusions can be drawn from results obtained in the 

 laboratory, it is essential that they should be checked against 

 known data regarding all main types of soil under different 

 conditions of climate. This difficulty, which is a serious one, has 

 for some time been recognised at Rothamsted, but with the 



