Fig. 4.  Brassica oleracea:  circumnutating movement of buried and arched hypocotyl, with the two legs of the arch tied together, traced on horizontal glass during 33 ½ hours.  Movement of the bead of filament magnified about 26 times, and here reduced to one-half original scale.

Before the above observations were made, some arched hypocotyls buried at the depth of a quarter of an inch were uncovered; and in order to prevent the two legs of the arch from beginning to separate at once, they were tied together with fine silk.  This was done partly because we wished to ascertain how long the hypocotyl, in its arched condition, would continue to move, and whether the movement when not masked and disturbed by the straightening process, indicated circumnutation.  Firstly a filament was fixed to the basal leg of an arched hypocotyl close above the summit of the radicle.  The cotyledons were still partially enclosed within the seed-coats.  The movement was traced (Fig. 4) from 9.20 A.M. on Dec. [page 15] 23rd to 6.45 A.M. on Dec. 25th.  No doubt the natural movement was much disturbed by the two legs having been tied together; but we see that it was distinctly zigzag, first in one direction and then in an almost opposite one.  After 3 P.M. on the 24th the arched hypocotyl sometimes remained stationary for a considerable time, and when moving, moved far slower than before.  Therefore, on the morning of the 25th, the glass filament was removed from the base of the basal leg, and was fixed horizontally on the summit of the arch, which, from the legs having been tied, had grown broad and almost flat.  The movement was now traced during 23 hours (Fig. 5), and we

Fig. 5.  Brassica oleracea:  circumnutating movement of the crown of a buried and arched hypocotyl, with the two legs tied together, traced on a horizontal glass during 23 hours.  Movement of the bead of the filament magnified about 58 times, and here reduced to one-half original scale.

see that the course was still zigzag, which indicates a tendency to circumnutation.  The base of the basal leg by this time had almost completely ceased to move.

As soon as the cotyledons have been naturally dragged from beneath the ground, and the hypocotyl has straightened itself by growth along the inner or concave surface, there is nothing to interfere with the free movements of the parts; and the circumnutation now becomes much more regular and clearly displayed, as shown in the following cases:—­A seedling was placed in front and near a north-east window with a line joining the [page 16] two cotyledons parallel to the window.  It was thus left the whole day so as to accommodate itself to the light.  On the following morning a filament was fixed to the midrib of the larger and taller cotyledon (which enfolds the other and smaller one, whilst still within the seed), and a mark being placed close behind, the movement of the whole plant, that is, of the hypocotyl and cotyledon, was traced greatly magnified on a vertical glass.  At first the plant bent so much towards the light that it was useless to attempt to trace the movement; but at 10 A.M. heliotropism almost wholly ceased and the first dot was