358 
poor soil, then the plants, as though they knew when they were well off, 
instead of forming long, tapering, cylindrical roots, as in the former 
case, either branch out into two or more ramifications, or taper rapidly 
to a point at the surface of the poor soil. Andrew Knight, who tried 
these experiments with the carrot and parsnip, also placed seeds of 
almost all the common esculent plants of a garden, so that the young 
plants had an opportunity of selecting either rich or poor soil, which 
was disposed in almost every possible way within their reach, and he 
always found abundant fibrous roots in the rich soil, and comparatively 
few in the poor. If two distinct species of trees, one requiring much 
moisture, the other naturally preferring a dry soil, are so placed that 
they can have the choice of either, the roots of the tree requiring much 
moisture will grow most in the direction of the wet soil, while the roots 
of the tree requiring a dry soil seem, from their direction, to avoid the 
water. 
The seeds of trees sometimes vegetate on the walls of old ruins and 
on ledges of rock, and in such situations they have been observed to act 
so wisely that some have concluded that plants must be endowed with a 
kind of instinct analagous to that of animals. The plants make very 
little progress at first, the soil on the rock or w~all affording them but a 
scanty subsistence ; to make the best of their position and get out of the 
difficulty, they are seen to direct their energies to the formation of a 
single root down the side of the rock to the soil below, which is no sooner 
reached than the hitherto dwarfed plant begins to extend its branches 
and assume its true character. In the neighborhood of Tunbridge 
Wells, in England, there are several remarkable groups of rocks. At 
the show place, called the High Rocks, I noticed several trees which had 
sprung from seeds in the ledges and in the crevices of the rocks, some 
of which had sent one straight root, many feet long, down the side of 
the rock to the soil; and in one instance a seed had vegetated in the 
lowest corner of an inclined ledge of rock, the main root of this, instead 
of growing perpendicularly down the side of the rock, had proceeded up 
the ledge and turned sharply over the top of the rock into the soil 
above. When a root proceeds down the side of a rock in this manner, 
it will in the first instance be as a delicate fibre, following either a crevice 
or a water-furrow, covered probably with mosses and lichens, and afford- 
