290 THEORIES OF FERTILISER ACTION [chap. 



concentrated solution, to which the plant's roots will be 

 drawn by the ordinary chemiotactic actions, and that 

 these zones will extend but a little way into the 

 generally much less dilute mass of the soil water, 

 because of the slowness of the diffusion process. 



That some such state of things prevails in the soil 

 may be surmised from the common farming experience 

 of the benefits derived from sowing the fertiliser close 

 to the seed, as when superphosphate is sown with turnip 

 seed, because in that case the fertiliser is not injurious 

 to germination and the young plant is specially dependent 

 on being rapidly pushed into growth in the early stages. 

 Again, the intimate way in which the feeding fibrous 

 roots of a plant will surround and cling to a fragment of 

 fertiliser in the soil, such as a bone or a piece of shoddy, 

 shows that some other actions are at work in the soil 

 than the feeding of the plant upon the nutrients con- 

 tained in the general soil solution. 



Whitney and Cameron's theory also supposes that 

 the plant itself exerts no solvent action, whereas it has 

 often been supposed that the roots excrete substances 

 of an acid nature which exert a solvent action upon the 

 soil particles. In this direction an experiment of Sachs' 

 has become classical. He took a slab of polished marble 

 and set it vertically in a pot of soil in which beans or 

 some kindred plant were grown. After the plants had 

 been growing for some time the contents of the pot 

 were turned out and the slab of marble washed, where- 

 upon the polished surface was found to be etched 

 wherever the roots had been growing in contact with 

 it. A polished slab of gypsum similarly treated shows 

 a raised pattern wherever the roots have protected the 

 surface from the solvent action of the general mass of 

 water in the soil. Although Sachs himself attributed 

 the etching to the action of the carbon dioxide which is 



