Leaves, when they go to sleep, move either upwards or downwards, or in the case of the leaflets of com-

* Ch.  Royer must, however, be excepted; see ‘Annales des Sc.  Nat.’ (5th series), Bot. vol. ix. 1868, p. 378.

[page 282] pound leaves, forwards, that is, towards the apex of the leaf, or backwards, that is, towards its base; or, again, they may rotate on their own axes without moving either upwards or downwards.  But in almost every case the plane of the blade is so placed as to stand nearly or quite vertically at night.  Therefore the apex, or the base, or either lateral edge, may be directed towards the zenith.  Moreover, the upper surface of each leaf, and more especially of each leaflet, is often brought into close contact with that of the opposite one; and this is sometimes effected by singularly complicated movements.  This fact suggests that the upper surface requires more protection than the lower one.  For instance, the terminal leaflet in Trifolium, after turning up at night so as to stand vertically, often continues to bend over until the upper surface is directed downwards whilst the lower surface is fully exposed to the sky; and an arched roof is thus formed over the two lateral leaflets, which have their upper surfaces pressed closely together.  Here we have the unusual case of one of the leaflets not standing vertically, or almost vertically, at night.

Considering that leaves in assuming their nyctitropic positions often move through an angle of 90o; that the movement is rapid in the evening; that in some cases, as we shall see in the next chapter, it is extraordinarily complicated; that with certain seedlings, old enough to bear true leaves, the cotyledons move vertically upwards at night, whilst at the same time the leaflets move vertically downwards; and that in the same genus the leaves or cotyledons of some species move upwards, whilst those of other species move downwards;—­from these and other such facts, it is hardly possible to doubt that plants must derive some [page 283] great advantage from such remarkable powers of movement.

The nyctitropic movements of leaves and cotyledons are effected in two ways,* firstly, by means of pulvini which become, as Pfeffer has shown, alternately more turgescent on opposite sides; and secondly, by increased growth along one side of the petiole or midrib, and then on the opposite side, as was first proved by Batalin.** But as it has been shown by De Vries*** that in these latter cases increased growth is preceded by the increased turgescence of the cells, the difference between the above two means of movement is much diminished, and consists chiefly in the turgescence of the cells of a fully developed pulvinus, not being followed by growth.  When the movements of leaves or cotyledons, furnished with a pulvinus and destitute of one, are compared, they are seen to be closely similar, and are apparently effected for the same purpose.  Therefore, with our object in view, it does not appear