Even whilst the arched or doubled hypocotyl is still beneath the ground, it circumnutates as much as the pressure of the surrounding soil will permit; but this was difficult to observe, because as soon as the arch is freed from lateral pressure the two legs begin to separate, even at a very early age, before the arch would naturally have reached the surface. Seeds were allowed to germinate on the surface of damp earth, and after they had fixed themselves by their radicles, and after the, as yet, only slightly arched hypocotyl had become nearly vertical, a glass filament was affixed on two occasions near to the base of the basal leg (i.e. the one in connection with the radicle), and its movements were traced in darkness on a horizontal glass. The result was that long lines were formed running in nearly the plane of the vertical arch, due to the early separation of the two legs now freed from pressure; but as the lines were zigzag, showing lateral movement, the arch must have been circumnutating, whilst it was straightening itself by growth along its inner or concave surface. A somewhat different method of observation was next followed: as soon as the earth with seeds in a pot began to crack, the surface was removed in parts to the depth of .2 inch; and a filament was fixed to the basal leg of a buried and arched hypocotyl, just above the summit of the radicle. The cotyledons were still almost completely enclosed within the much-cracked seed-coats; and these were again covered up with damp adhesive soil pressed pretty firmly down. The movement of the filament was traced (Fig. 3) from 11 A.M. Feb. 5th till 8 A.M. Feb. 7th. By this latter period the cotyledons had been dragged from beneath the pressed-down earth, but the upper part of the hypocotyl still formed nearly a right angle with the lower part. The tracing shows that the arched hypocotyl tends at this early age to circumnutate irregularly. On the first day the greater movement (from right to left in the figure) was not in the plane of the vertical and arched hypocotyl, but at right angles to it, or in the plane of the two cotyledons, which were still in close contact. The basal leg of the arch at the time when the filament was affixed to it, was already bowed considerably backwards, or from the cotyledons; had the filament been affixed before this bowing occurred, the chief movement would have been at right angles to that shown in the figure. A filament was attached to another buried hypocotyl of the same age, and it moved in a similar general manner, but the line traced was not so complex. This hypocotyl became almost straight, and the cotyledons were dragged from beneath the ground on the evening of the second day. Fig. 3. Brassica oleracea: circumnutating movement of buried and arched hypocotyl (dimly illuminated from above), traced on horizontal glass during 45 hours. Movement of bead of filament magnified about 25 times, and here reduced to one-half of original scale.