Manchester Meuwirs, ]\^L iviii. (^\()\d,). No. 0. II 



accordance with the composition of the soils. Further, the 

 rate of growth either in these soil solutions or in others 

 artificially made up of varying concentration had been 

 found to depend both on the composition and the concen- 

 tration, and no evidence had come to light of the existence 

 in soils in which barlej^ or wheat had been grown for many 

 years in succession of the presence of toxic substances in- 

 jurious to these plants. The last experiments, however, had 

 raised an interesting side issue. One might conceive that 

 even if the normal soil solution be of constant composition 

 the rate at which it will be renewed after depletion by the 

 growing plant will vary with the actual amount of solid 

 compounds of phosphoric acid and potash in the soil. The 

 root hairs of the plant will exhaust the solution with which 

 they are in contact ; then since the water in the soil exists 

 in a state of thin films coating the soil particles so much 

 time might elapse during the travel of the nutrients from 

 the points of solution along the extensive film to the root 

 hairs that in a soil deficient in phosphoric acid and potash 

 the plant might always be less well nourished than in a 

 richer soil in which the length of travel of the nutrients 

 was smaller. To test this point, the nutrient solutions of 

 varied concentration employed in the last series of ex- 

 periments in bottles were also diffused as a thin film over 

 a mass of sand, until the sand was sensibl}' damp, but no 

 liquid could be squeezed out of it. The jars of sand 

 contained exactly the same bulk of solution as the 

 bottles ; they were weighed regularly and pure water 

 added as required. The photograph, PL III. (exactly 

 comparable with Pi /.), shows the results obtained, which 

 were unexpected, in that all the cultures in sand were 

 much better than the cultures in solutions of the same 

 concentration and amount in bottles (see Table VI.), 

 though they showed the same variation of growth with 



