242 



PHYSIOLO&Y OF NUTRITION. 



touched the marble at a ; it then took the direction indicated by 

 the broad stripe in the drawing, while the remaining lines of 

 corrosion owe their origin to the lateral roots. It remains to be 

 noted that the bean plants, the corrosive action of whose roots we 

 are investigating, should not be allowed to grow for too long- 

 a time (only from ten to fourteen days), since if the experiment 

 lasts long very many roots come into contact with the marble, 

 and the corrosion lines due to the individual roots no longer 

 stand out distinctly. 1 



1 Corrosion phenomena were first thoroughly studied. by Sachs; see his 

 Handbuch dcr Experimentalphysiologie der Ptfanzen, 1865, p. 188. 



91. The Causes of Corrosion Phenomena. 



Corrosive action obviously can only result from ,the fact that 

 the roots, lying in close approximation to the stones or particles 

 of soil, give out certain substances capable of decomposing them. 

 Naturally the Carbon dioxide, formed in the cells of the root as a 

 product of respiration, will first be thought of ; but organic acids, 

 and even Hydrochloric acid also, claim consideration in this con- 

 nection. If the membranes of the root-cells are permeated with 

 dilute solutions of these substances, then action of the roots on the 

 stony and earthy constituents of the soil is at once rendered pos- 

 sible, and this is clearly brought out by the following experiment.* 

 We put together the apparatus represented in Fig. 90. The 

 bottle G contains dilute Hydrochloric acid. Through the hole of 

 the cork closing this bottle passes the 

 comparatively wide glass tube ii*, covered 

 at its lower end with a piece of pig's 

 bladder. In our apparatus the fluid 

 represents the cell contents, and the 

 bladder the membranes of the root-cells. 

 If we lay a bit of marble on the bladder 

 permeated with dilute Hydrochloric acid, 

 it is soon partially dissolved, and the solu- 

 tion of Calcium chloride formed passes by 

 diffusion into the Hydrochloric acid. We 

 can easily detect the presence of Calcium 

 in the fluid by means of Ammonium 

 oxalate. Similarly the solutions permeating 



FIG.' 9). Apparatus for 

 illustrating some features 

 of corrosion phenomena. 



* This experiment was first made by Zoller at the suggestion of Liebig. 



