430 Development of Roots of Agricultural Plants. 
have to do there with underground stems and not with true roots, 
the evil is of more remote origin. 
It is well known that both vegetable and mineral poisons are 
destructive to phaenogamic plants. How, then, do the poisons 
act? Roots physiologically sound do not absorb indifferently all 
substances which are soluble in water or which may be incor- 
porated with the soil in a minute state of division. The experi- 
ment may be easily tried with arsenious acid, and there is the 
more reason to repeat the experiment, as published accounts are 
very conflicting. The truth seems to be that if the dose is only 
small, it is not absorbed ; if large, the spongelets are destroyed, 
and a portion enters but only a short way up. The plant is then 
destroyed at once, or if some of the upper roots remain sound, 
the poison is eliminated partly from the leaves which fade pre- 
cisely according to their order of development, and partly together 
with the water of evaporation. In such cases the plant in the 
end is not affected, and ultimately no poison can be detected by 
the closest analysis. Dr. Daubeny made experiments also with 
strontian and barytes, which do not normally exist in plants ; 
and, as in the case of arsenic, he found on analysis no trace of 
these substances where plants arrived at perfection : showing 
clearly that either the roots had some power of refusing substances 
which were not necessary, or might prove injurious, or that if a 
portion was admitted, it was somehow or other eliminated. Some 
metallic salts, exclusive of such matters as potash and soda, 
however, do not seem so difficult of access. Iron, at least, seems 
essential to many plants, and it is occasionally found on analysis, 
as in beechwood or tree-lichens, where it does not normally form 
a constituent. 
" The power of roots," says Liebig, "to preclude the passing 
of certain substances from the soil into the plant is not absolute. 
Forchhammer (Poggend. Annal., xcv. 90) detected exceedingly 
minute traces of lead, zinc, and copper in the wood of the boech, 
birch, and fir, and tin, lead, and zinc in that of the oak ; but the 
fact that the outer rind or bark in particular is found to contain 
metals of this kind in perceptibly larger quantities than the 
wood, clearly points to the accidental nature of their presence, 
and to their taking no essential part in the vital processes of the 
plant." The copper which is often found in wheat on analysis 
is evidently derived from the sulphate of copper with which 
seed-wheat is dressed to prevent bunt. 
It now becomes a question how far any elimination can take 
place by the roots, a most important matter as regards the theory 
of cropping. M. Cauvet accordingly directed especial attention 
to this point, and the result of numerous experiments is entirely 
