308 
PHYSIOLOGY. 
BOOK II. 
proof of its contributing, in some way or other, to the nutri- 
tion of the vegetable system," 
Fixed as plants are to the soil, deprived of volition, and 
incapable of removing their highly absorbent roots from what 
is hurtful to them, except with extreme slowness, it appears 
scarcely probable that they should have any power of select- 
ing their food ; on the contrary, the facility with which they 
are poisoned would seem to confirm the correctness of the 
usual supposition. But, if roots are made to grow in 
coloured infusions, it is said that they take up only the co- 
lourless parts, leaving the coloured behind ; and we know that 
if an apple tree is planted in a piece of ground in which an- 
other apple tree has been growing many years, the new plant 
will languish and become unhealthy, whatever quantity of 
manure, that is of new food, may be offered to its roots. This 
last fact is accounted for upon the supposition that the soil 
contains some peculiar principles which are necessary to the 
health of an apple tree, and that the old tree, having selected 
for its own consumption all that the soil contained, has left 
none behind it for the new comer ; but the probability is that 
this hypothesis is untenable, and that the fact is to be ex- 
plained upon very different principles (see Chap. X.). It has 
been, however, demonstrated by Daubeny that plants have, to 
a certain extent, a power of selection by their roots. He 
found that when barley was watered with distilled water, con- 
taining in every two gallons two ounces of nitrate of strontian, 
not a trace of that earth could be detected in the ashes of 
the plants ; and when Lotus tetragonolobus was treated in a 
similar manner, excepting that only two ounces of nitrate of 
strontian were dissolved in ten gallons of distilled water, 
although the whole of that quantity was expended upon them, 
a minute examination demonstrated that the stems contained 
no trace whatever of strontian, although a small portion ap- 
peared to be present in, or at least adherent to, the roots. By 
other experiments it was ascertained, that the strontian was 
not in these cases first received into the system, and afterwards 
rejected through the roots ; for when the roots of a Pelargonium 
were divided into two nearly equal bundles, one of which 
had its extremity immersed in a glass containing a weak solu- 
