On the other hand, all roots which penetrate the ground (including the modified root-like petioles of Megarrhiza and Ipomoea leptophylla) are guided in their downward course by geotropism; and so are many aërial roots, whilst others, as those of the Ivy, appear to be indifferent to its action.  In our first chapter the movements of the radicles of several seedlings were described.  We may there see (Fig. 1) how a radicle of the cabbage, when pointing vertically upwards so as to be very little acted on by geotropism, circumnutated; and how another (Fig. 2) which was at first placed in an inclined position bowed itself downwards in a zigzag line, sometimes remaining stationary for a time.  Two other radicles of the cabbage travelled downwards in almost rectilinear courses.  A radicle of the bean placed upright (Fig. 20) made a great sweep and zigzagged; but as it sank downwards and was more strongly acted on by geotropism, it moved in an [page 513] almost straight course.  A radicle of Cucurbita, directed upwards (Fig. 26), also zigzagged at first, and described small loops; it then moved in a straight line.  Nearly the same result was observed with the radicles of Zea mays.  But the best evidence of the intimate connection between circumnutation and geotropism was afforded by the radicles of Phaseolus, Vicia, and Quercus, and in a less degree by those of Zea and Aesculus (see Figs. 18, 19, 21, 41, and 52); for when these were compelled to grow and slide down highly inclined surfaces of smoked glass, they left distinctly serpentine tracks.

[The Burying of Seed-capsules:  Trifolium subterraneum.—­The flower-heads of this plant are remarkable from producing only 3 or 4 perfect flowers, which are situated exteriorly.  All the other many flowers abort, and are modified into rigid points, with a bundle of vessels running up their centres.  After a time 5 long, elastic, claw-like projections, which represent the divisions of the calyx, are developed on their summits.  As soon as the perfect flowers wither they bend downwards, supposing the peduncle to stand upright, and they then closely surround its upper part.  This movement is due to epinasty, as is likewise the case with the flowers of T. repens.  The imperfect central flowers ultimately follow, one after the other, the same course.  Whilst the perfect flowers are thus bending down, the whole peduncle curves downwards and increases much in length, until the flower-head reaches the ground.  Vaucher* says that when the plant is so placed that the heads cannot soon reach the ground, the peduncles grow to the extraordinary length of from 6 to 9 inches.  In whatever position the branches may be placed, the upper part of the peduncle at first bends vertically upwards through heliotropism; but as soon as the flowers begin to wither the downward curvature of the whole peduncle commences.  As this latter movement occurred in complete darkness, and with peduncles arising from upright and from dependent branches, it cannot be due to apheliotropism or to epinasty, but must be attributed to geotropism.  Nineteen