the middle, so that the lower part answers to the main peduncle, [page 507] and the upper part to one of the sub-peduncles of O. carnosa.  The upper part bends downwards, after the flower has begun to wither, and the whole peduncle then forms a hook; that this bending is due to epinasty we may infer from the case of O. carnosa.  When the pod is nearly ripe, the upper part straightens itself and becomes erect; and this is due to hyponasty or apogeotropism, or both combined, and not to heliotropism, for it occurred in darkness.  The short, hooked part of the peduncle of a cleistogamic flower, bearing a pod nearly ripe, was observed in the dark during three days.  The apex of the pod at first pointed perpendicularly down, but in the course of three days rose 90o, so that it now projected horizontally.  The course during the two latter days is shown in Fig. 189; and it may be seen how greatly the peduncle, whilst rising, circumnutated.  The lines of chief movement were at right angles to the plane of the originally hooked part.  The tracing was not continued any longer; but after two additional days, the peduncle with its capsule had become straight and stood upright.]

Concluding Remarks on Apogeotropism.—­When apogeotropism is rendered by any means feeble, it acts, as shown in the several foregoing cases, by increasing the always present circumnutating movement in a direction opposed to gravity, and by diminishing that in the direction of gravity, as well as that to either side.  The upward movement thus becomes unequal in rate, and is sometimes interrupted by stationary periods.  Whenever irregular ellipses or loops are still formed, their longer axes are almost always directed in the line of gravity, in an analogous manner as occurred with heliotropic movements in reference to the light.  As apogeotropism acts more and more energetically, ellipses or loops cease to be formed, and the course becomes at first strongly, and then less and less zigzag, and finally rectilinear.  From this gradation in the nature of the movement, and more especially from all growing parts, which alone (except when pulvini are present) are acted on by apogeotropism, con-[page 508] tinually circumnutating, we may conclude that even a rectilinear course is merely an extremely modified form of circumnutation.  It is remarkable that a stem or other organ which is highly sensitive to apogeotropism, and which has bowed itself rapidly upwards in a straight line, is often carried beyond the vertical, as if by momentum.  It then bends a little backwards to a point round which it finally circumnutates.  Two instances of this were observed with the hypocotyls of Beta vulgaris, one of which is shown in Fig. 183, and two other instances with the hypocotyls of Brassica.  This momentum-like movement probably results from the accumulated effects of apogeotropism.  For the sake of observing how long such after-effects lasted, a pot with seedlings of Beta was laid on its side in the dark, and the hypocotyls in 3 h. 15 m. became highly inclined.  The pot, still in the dark, was then placed upright, and the movements of the two hypocotyls were traced; one continued to bend in its former direction, now in opposition to apogeotropism, for about 37 m., perhaps for 48 m.; but after 61 m. it moved in an opposite direction.  The other hypocotyl continued to move in its former course, after being placed upright, for at least 37 m.