way down the stem; but there is one important difference. The upper
part of the sapling when thus moved remains straight; but with
twining plants every part of the revolving shoot has its own separate
and independent movement. This is easily proved; for when the lower
half or two-thirds of a long revolving shoot is tied to a stick, the
upper free part continues steadily revolving. Even if the whole
shoot, except an inch or two of the extremity, be tied up, this part,
as I have seen in the case of the Hop, Ceropegia, Convolvulus, &c.,
goes on revolving, but much more slowly; for the internodes, until
they have grown to some little length, always move slowly. If we
look to the one, two, or several internodes of a revolving shoot,
they will be all seen to be more or less bowed, either during the
whole or during a large part of each revolution. Now if a coloured
streak be painted (this was done with a large number of twining
plants) along, we will say, the convex surface, the streak will after
a time (depending on the rate of revolution) be found to be running
laterally along one side of the bow, then along the concave side,
then laterally on the opposite side, and, lastly, again on the
originally convex surface. This clearly proves that during the
revolving movement the internodes become bowed in every direction.
The movement is, in fact, a continuous self-bowing of the whole
shoot, successively directed to all points of the compass; and has
been well designated by Sachs as a revolving nutation.

As this movement is rather difficult to understand, it will be well
to give an illustration. Take a sapling and bend it to the south,
and paint a black line on the convex surface; let the sapling spring
up and bend it to the east, and the black line will be seen to run
along the lateral face fronting the north; bend it to the north, the
black line will be on the concave surface; bend it to the west, the