The probability of its occurrence was inferred by Sachs,* from radicles placed vertically upwards being acted on by geotropism (which we likewise found to be the case), for if they had remained absolutely perpendicular, the attraction of gravity could not have caused them to bend to any one side. Circumnutation was observed in the above specified cases, either by means of extremely fine filaments of glass affixed to the radicles in the manner previously described, or by their being allowed to grow downwards over inclined smoked glass-plates, on which they left their tracks. In the latter cases the serpentine course (see Figs. 19, 21, 27, 41) showed unequivocally that the apex had continually moved from side to side. This lateral movement was small in extent, being in the case of Phaseolus at most about 1 mm. from a medial line to both sides. But there was also movement in a vertical plane at right angles to the inclined glass-plates. This was shown by the tracks often being alternately a little broader and narrower, due to the radicles having alternately pressed with greater and less force on the plates. Occasionally little bridges of soot were left across the tracks, showing that the apex had at these spots been lifted up. This latter fact was especially apt to occur * 'Ueber das Wachsthum der Wurzeln: Arbeiten des bot. Instituts in Würzburg,' Heft iii. 1873, p. 460. This memoir, besides its intrinsic and great interest, deserves to be studied as a model of careful investigation, and we shall have occasion to refer to it repeatedly. Dr. Frank had previously remarked ('Beiträge zur Pflanzenphysiologie, 1868, p. 81) on the fact of radicles placed vertically upwards being acted on by geotropism, and he explained it by the supposition that their growth was not equal on all sides.
[page 71] when the radicle instead of travelling straight down the glass made a semicircular bend; but Fig. 52 shows that this may occur when the track is rectilinear. The apex by thus rising, was in one instance able to surmount a bristle cemented across an inclined glass-plate; but slips of wood only 1/40 of an inch in thickness always caused the radicles to bend rectangularly to one side, so that the apex did not rise to this small height in opposition to geotropism.
In those cases in which radicles with attached filaments were placed so as to stand up almost vertically, they curved downwards through the action of geotropism, circumnutating at the same time, and their courses were consequently zigzag. Sometimes, however, they made great circular sweeps, the lines being likewise zigzag.
Radicles closely surrounded by earth, even when this is thoroughly soaked and softened, may perhaps be quite prevented from circumnutating. Yet we should remember that the circumnutating sheath-like cotyledons of Phalaris, the hypocotyls of Solanum, and the epicotyls of Asparagus formed round themselves little circular cracks or furrows in a superficial layer of damp argillaceous sand. They were also able, as well as the hypocotyls of Brassica, to form straight furrows in damp sand, whilst circumnutating and bending towards a lateral light. In a future chapter it will be shown that the rocking or circumnutating movement of the flower-heads of Trifolium subterraneum aids them in burying themselves. It is therefore probable that the circumnutation of the tip of the radicle aids it slightly in penetrating the ground; and it may be observed in several of the previously given diagrams, that the movement is more strongly pronounced in radicles when they first [page 72] protrude from the seed than at a rather later period; but whether this is an accidental or an adaptive coincidence we do not pretend to decide. Nevertheless, when young radicles of Phaseolus multiflorus were fixed vertically close over damp sand, in the expectation that as soon as they reached it they would form circular furrows, this did not occur,--a fact which may be accounted for, as we believe, by the furrow being filled up as soon as formed by the rapid increase of thickness in the apex of the radicle.