circumnutated, for it moved during this day 3 times up and 3 times down, 4 times to the left and 4 to the right.  But the course was so complex that it could hardly be traced on the glass.  We can, however, see that the successively formed irregular ellipses rose higher and higher.  Apogeotropism continued to act on the fourth morning, as the stem was still rising, though it now stood only 23o from the perpendicular.  In this diagram the several stages may be followed by which an almost rectilinear, upward, apogeotropic course first becomes zigzag, and then changes into a circumnutating movement, with most of the successively formed, irregular ellipses directed upwards.

Fig 184:  Rubus idaeus (hybrid):  apogeotropic movement of stem, traced on a vertical glass during 3 days and 3 nights, from 10.40 A.M.  March 18th to 8 A.M. 21st.  Figure reduced to one-half of the original scale.

Lilium auratum.—­A plant 23 inches in height was placed [page 499] horizontally, and the upper part of the stem rose 58o in 46 h., in the manner shown in the accompanying diagram (Fig. 185).  We here see that during the whole of the second day of 15 ½ h., the stem plainly circumnutated whilst bending upwards through apogeotropism.  It had still to rise considerably, for when the last dot in the figure was made, it stood 32o from an upright position.

Fig. 185.  Lilium auratum:  apogeotropic movement of stem, traced on a vertical glass during 2 days and 2 nights, from 10.40 A.M.  March 18th to 8 A.M. 20th.  Figure reduced to one-half of the original scale.

Phalaris Canariensis.—­A cotyledon of this plant (1.3 inch in height) has already been described as rising in 4 h. 30 m. from 40o beneath the horizon into a vertical position, passing through an angle of 130o in a nearly straight line, and then abruptly beginning to circumnutate.  Another somewhat old cotyledon of the same height (but from which a true leaf had not yet protruded), was similarly placed at 40o beneath the horizon.  For the first 4 h. it rose in a nearly straight course (Fig. 186), so that by 1.10 P.M. it was highly inclined, and now apogeotropism acted on it with much less power than before, and it began to zigzag.  At 4.15 P.M. (i.e. in 7 h. from the commencement) it stood vertically, and afterwards continued to circumnutate in the usual manner about the same spot.  Here then we have a graduated change from a straight upward apogeotropic course into circumnutation, instead of an abrupt change, as in the former case.

Avena sativa.—­The sheath-like cotyledons, whilst young, are strongly apogeotropic; and some which were placed at 45o beneath the horizon rose 90o in 7 or 8 h. in lines almost absolutely straight.  An oldish cotyledon, from which the first leaf began to [page 500] protrude whilst the following observations were being made, was placed at 10o beneath the horizon, and it rose only 59o in 24h.  It behaved rather differently from any other plant, observed by us, for during the first 4 ½ h. it rose in a line not far from straight; during the next 6 ½ h. it circumnutated, that is, it descended and again ascended in a strongly marked zigzag course; it then resumed its upward movement in a moderately straight line, and, with time allowed, no doubt would have become upright.  In this case, after the first 4 ½ h., ordinary circumnutation almost completely conquered for a time apogeotropism.