The degree of curvature of the cotyledons of Phalaris within a given time, depends not merely on the amount of lateral light which they may then receive, but on that which they have previously received from above and on all sides.  Analogous facts have been given with respect to the nyctitropic and periodic movements of plants.  Of two pots containing seedlings of Phalaris which had germinated in darkness, one was still kept in the dark, and the other was exposed (Sept. 26th) to the light in a greenhouse during a cloudy day and on the following bright morning.  On this morning (27th), at 10.30 A.M., both pots were placed in a box, blackened within and open in front, before a north-east window, protected by a linen and muslin blind and by a towel, so that but little light was admitted, though the sky was bright.  Whenever the pots were looked at, this was done as quickly as possible, and the cotyledons were then held transversely with respect to the light, so that their curvature could not have been thus increased or diminished.  After 50 m. the seedlings which had previously been kept in darkness, were perhaps, and after 70 m. were certainly, curved, though very slightly, towards the window.  After 85 m. some of the seedlings, which had previously been illuminated, were perhaps a little affected, and after 100 m. some of the younger ones were certainly a little curved towards the light.  At this time (i.e. after 100 m.) there was a plain difference [page 460] in the curvature of the seedlings in the two pots.  After 2 h. 12 m. the chords of the arcs of four of the most strongly curved seedlings in each pot were measured, and the mean angle from the perpendicular of those which had previously been kept in darkness was 19o, and of those which had previously been illuminated only 7o.  Nor did this difference diminish during two additional hours.  As a check, the seedlings in both pots were then placed in complete darkness for two hours, in order that apogeotropism should act on them; and those in the one pot which were little curved became in this time almost completely upright, whilst the more curved ones in the other pot still remained plainly curved.

Two days afterwards the experiment was repeated, with the sole difference that even less light was admitted through the window, as it was protected by a linen and muslin blind and by two towels; the sky, moreover, was somewhat less bright.  The result was the same as before, excepting that everything occurred rather slower.  The seedlings which had been previously kept in darkness were not in the least curved after 54 m., but were so after 70 m.  Those which had previously been illuminated were not at all affected until 130 m. had elapsed, and then only slightly.  After 145 m. some of the seedlings in this latter pot were certainly curved towards the light; and there was now a plain difference between the two pots.  After 3 h. 45 m. the chords of the arcs of 3 seedlings in each pot were measured, and the mean angle from the perpendicular was 16o for those in the pot which had previously been kept in darkness, and only 5o for those which had previously been illuminated.