244 SECTIONAL ADDRESSES 



not made in soil studies of drain gauges or lysimeters, or instruments of a 

 similar kind. 



The first drain gauges, so far as I am aware, were made by Lawes and 

 Gilbert at Rothamsted over sixty years ago. They were designed to study 

 evaporation and percolation in relation to depth of drainage, and were 

 therefore of different depths, 20, 40 and 60 inches respectively. They 

 were also used to study the amount of nitrogen washed away from un- 

 cropped and unmanured soil. The blocks of soil enclosed in these drain 

 gauges were never broken up, they were built with as little disturbance as 

 possible into the water-tight structures which enable the drainage to be 

 measured. They consist therefore of real soils which have been formed 

 by a long course of natural soil-forming processes. Similarly the drain 

 gauges which I have had built at Craibstone, near Aberdeen, have been 

 formed by enclosing, without disturbance, in water-tight boxes of Caithness 

 slate, blocks of natural soil which have never been broken up. My drain 

 gauges are intended to study the changes which take place in cultivated 

 soil, and the losses which take place in the drainage water during ordinary 

 processes of cropping and manuring. 



Such drain gauges are not easy to construct. I suppose that is why this 

 method has been so little used in the study of soils. It is much easier, 

 and cheaper, to build a water-tight box and fill subsoil and soil into it, than 

 it is to enclose a block of natural soil, weighing several tons, in a water-tight 

 structure. If the easier method is adopted, as has been done to a large 

 extent in America and elsewhere, its limitations must be recognised. The ■ 

 soil, once it is broken up and filled into a lysimeter, is no longer a natural 

 soil and it is difficult to say how long it will take under the influence of the 

 soil- forming processes of the locality to become once more a real soil such 

 as is provided in nature. Had Lawes and Gilbert, for instance, filled the 

 soil into their drain gauges they would have defeated the object they had 

 in view, for such a broken up soil would have allowed water to run through 

 it quite differently from a natural soil for its structure would have been 

 destroyed, while the breaking up and aeration of the soil would have rendered 

 useless their studies of the loss of nitrate, for the nitrification in such an 

 artificial soil would have been quite abnormal. 



Artificially filled drain gauges have certain uses. I have used them 

 myself in studying the limits of fixation of manurial substances by the soil, 

 but we must always recognise that they are artificial and that results 

 obtained from them do not necessarily apply, or may only apply with 

 modifications to natural soils. They are in a similar position to pot 

 experiments as compared with field experiments. Pot experiments can 

 be very useful, and this method of experimentation has yielded most 

 valuable results, but we have always to recognise that it has its limitations 

 and that it is very difficult to find a formula which will enable us to apply 

 the results obtained by it with any degree of certainty to field conditions. 



The development of our knowledge of soil colloids and base exchange 

 during the present century is second in importance only to the advance 

 which has been made in the science of soil formation, structure and 

 distribution. As you know, the beginnings of our knowledge of this 

 subject can be traced back to the middle of last century when Way 



