the apex drew itself backwards with comparative slowness for part of the distance which had just been gained; and then after a very short time made another jerk forwards.  Four conspicuous jerks forwards, with slower retreats, were seen on one occasion to occur in exactly one minute, besides some minor oscillations.  As far as we could judge, the advancing and retreating lines did not coincide, and if so, extremely minute ellipses were each time described.  Sometimes the apex remained quite motionless for a short period.  Its general course during the several hours of observation was in two opposite directions, so that the leaf was probably circumnutating.

An older leaf with the lobes fully expanded, and which was afterwards proved to be highly sensitive to contact, was next observed in a similar manner, except that the plant was exposed to a lower temperature in a room.  The apex oscillated forwards and backwards in the same manner as before; but the jerks forward were less in extent, viz. about 1/1000 inch; and there were longer motionless periods.  As it appeared possible that the movements might be due to currents of air, a wax taper was held close to the leaf during one of the motionless periods, but no oscillations were thus caused.  After 10 m., however, vigorous oscillations commenced, perhaps owing to the plant having been warmed and thus stimulated.  The candle was then removed and before long the oscillations ceased; nevertheless, when looked at again after an interval of 1 h. 30 m., it was again oscillating.  The plant was taken back into the hot-house, and on the following morning was seen to be oscillating, though not very vigorously.  Another old but healthy leaf, which was not in the least sensitive to a touch, was likewise observed during two days in the hot-house, and the attached filament made many little jerks forwards of about 2/1000 or only 1/1000 of an inch.

Finally, to ascertain whether the lobes independently of the petiole oscillated, the petiole of an old leaf was cemented close to the blade with shellac to the top of a little stick driven into the soil.  But before this was done the leaf was observed, and found to be vigorously oscillating or jerking; and after it had been cemented to the stick, the oscillations of about 2/1000 of an inch still continued.  On the following day a little infusion [page 244] of raw meat was placed on the leaf, which caused the lobes to close together very slowly in the course of two days; and the oscillations continued during this whole time and for the next two days.  After nine additional days the leaf began to open and the margins were a little everted, and now the apex of the glass filament remained for long periods motionless, and then moved backwards and forwards for a distance of about 1/1000 of an inch slowly, without any jerks.  Nevertheless, after warming the leaf with a taper held close to it, the jerking movement recommenced.