the sensitive cotyledon remains horizontal, imprisoned in its tube; it
will therefore be continually stimulated and will continue to transmit
influences to the bending region, which should therefore curl up into a
helix or corkscrew-like form,--and this is precisely what occurred.

I have referred to this work principally because the same method was
applied to roots by Massart (Massart, "Mem. Couronnes Acad. R. Belg."
LXII. 1902.) and myself (F. Darwin, "Linnean Soc. Journ." XXXV. 1902,
page 266.) with a similar though less striking result. Although these
researches confirmed Darwin's work on roots, much stress cannot be laid
on them as there are several objections to them, and they are not easily
repeated.

The method which--as far as we can judge at present--seems likely
to solve the problem of the root-tip is most ingenious and is due to
Piccard. (Pringsheim's "Jahrb." XL. 1904, page 94.)

Andrew Knight's celebrated experiment showed that roots react to
centrifugal force precisely as they do to gravity. So that if a bean
root is fixed to a wheel revolving rapidly on a horizontal axis, it
tends to curve away from the centre in the line of a radius of the
wheel. In ordinary demonstrations of Knight's experiment the seed is
generally fixed so that the root is at right angles to a radius, and as
far as convenient from the centre of rotation. Piccard's experiment is
arranged differently. (A seed is depicted below a horizontal dotted
line AA, projecting a root upwards.) The root is oblique to the axis
of rotation, and the extreme tip projects beyond that axis. Line AA
represents the axis of rotation, T is the tip of the root just above the
line AA, and B is the region just below line AA in which curvature takes
place. If the motile region B is directly sensitive to gravitation (and
is the only part which is sensitive) the root will curve (down and away
from the vertical) away from the axis of rotation, just as in Knight's
experiment. But if the tip T is alone sensitive to gravitation the
result will be exactly reversed, the stimulus originating in T and
conveyed to B will produce curvature (up towards the vertical). We may
think of the line AA as a plane dividing two worlds. In the lower one
gravity is of the earthly type and is shown by bodies falling and roots
curving downwards: in the upper world bodies fall upwards and roots
curve in the same direction. The seedling is in the lower world, but its
tip containing the