INVESTIGATIONS ON THE PRODUCTION OF PLANT FOOD. 177 



becomes possible. In order to follow the process more closely we 

 have made at Rothamsted permanent beds of some of our clay sub- 

 soil, and have commenced systematic botanical, bacteriological, and 

 chemical observations of the changes taking place. 



The reverse process can also be seen at Rothamsted. A soil has 

 been persistently cultivated for seventy-one years and all of the crop 

 removed except the roots and the stubble : the organic matter and 

 the energy supplies are therefore running out, and we are gradually 

 approximating to the original condition. It is interesting that here 

 also colt's-foot and Equisetum flourish, although we hardly find them on 

 our other plots. 



The system we have to study is this mass of mineral particles 

 intermingled with plant residues and living organisms ; and our 

 object is to trace the stages by which the decomposition proceeds 

 and the energy runs down. 



The first obvious change is that the plant material loses its green 

 colour and goes black ; this can easily be observed when leaves are dug 

 into the soil or dragged in by earthworms. The old chemists were much 

 interested in this black substance, and in the early days of the last 

 century, when men of science were very prodigal with new names 

 for new forces and new substances, they supposed it to be made up of a 

 number of compounds which they called ulmic acid, crenic acid, 

 apocrenic acid, humic acid, &c. No one has ever succeeded in pre- 

 paring any of these compounds in any state that would satisfy a modern 

 chemist, and there is no evidence whatsoever that they exist ; but 

 their names have been piously handed down through long generations 

 of students, and they still occasionally turnup in popular articles and 

 in answers to examination papers. 



So far this black material has defied analysis. Modern organic 

 chemistry has been developed largely to deal with liquids and crystals ; 

 this black substance is neither, but is an amorphous glue-like body 

 of the class known as colloids. Physical chemists in several countries 

 are (or were) working at colloids, and we may yet hope to see some 

 method of resolving them. The way out of such a predicament is 

 to give the substance a non-committal name, and so we retain the old 

 designation " humus." 



Part of the humus is soluble in alkalis, and this was supposed 

 naturally enough to be necessarily of more value in plant nutrition 

 than the insoluble part ; analytical methods were therefore devised 

 for estimating its amount. It does not appear, however, that anyone 

 has ever tried the fundamental experiment of testing whether the 

 soluble part is really superior to the rest. A priori assumptions in 

 our subject are particularly treacherous, and at Rothamsted the 

 question has recently been under investigation. So far as present 

 experiments have gone, the removal of this soluble humus has failed 

 to make any appreciable difference to the growth of the plant, or to the 

 physical and chemical properties of the soil. Further experiments 

 with various types of soil would be necessary before pronouncing too 



