Apogeotropic Movements effected by the aid of joints or pulvini.—­Movements of this kind are well known to occur in the Gramineae, and are effected by means of the thickened bases of their sheathing leaves; the stem within being in this part thinner than elsewhere.* According to the analogy of all other pulvini, such joints ought to continue circumnutating for a long period, after the adjoining parts have ceased to grow.  We therefore wished to ascertain whether this was the case with the Gramineae; for if so, the upward curvature of their stems, when extended horizontally or laid prostrate, would be explained in accordance with our view—­namely, that apogeotropism results from modified circumnutation.  After these joints have curved upwards, they are fixed in their new position by increased growth along their lower sides.

Lolium perenne.—­A young stem, 7 inches in height, consisting of 3 internodes, with the flower-head not yet protruded, was selected for observation.  A long and very thin glass filament was cemented horizontally to the stem close above the second joint, 3 inches above the ground.  This joint was subsequently proved to be in an active condition, as its lower side swelled much through the action of apogeotropism (in the manner described by De Vries) after the haulm had been fastened down for 24 h. in a horizontal position.  The pot was

* This structure has been recently described by De Vries in an interesting article, ‘Ueber die Aufrichtung des gelagerten Getreides,’ in ‘Landwirthschaftliche Jahrbücher,’ 1880, p. 473. [page 503]

so placed that the end of the filament stood beneath the 2-inch object glass of a microscope with an eye-piece micrometer, each division of which equalled 1/500 of an inch.  The end of the filament was repeatedly observed during 6 h., and was seen to be in constant movement; and it crossed 5 divisions of the micrometer (1/100 inch) in 2 h.  Occasionally it moved forwards by jerks, some of which were 1/1000 inch in length, and then slowly retreated a little, afterwards again jerking forwards.  These oscillations were exactly like those described under Brassica and Dionaea, but they occurred only occasionally.  We may therefore conclude that this moderately old joint was continually circumnutating on a small scale.

Alopecurus pratensis.—­A young plant, 11 inches in height, with the flower-head protruded, but with the florets not yet expanded, had a glass filament fixed close above the second joint, at a height of only 2 inches above the ground.  The basal internode, 2 inches in length, was cemented to a stick to prevent any possibility of its circumnutating.  The extremity of the filament, which projected about 50o above the horizon, was often observed during 24 h. in the same manner as in the last case.  Whenever looked at, it was always in movement, and it crossed 30 divisions of the micrometer (3/50 inch) in 3 ½ h.; but it sometimes moved at a quicker rate, for