fact of ellipses being described, shows that the alternately increased turgescence of the cells does not occur on exactly opposite sides of the pulvinus, any more than the increased growth which causes the movements of leaves not furnished with pulvini.  When a pulvinus is present, the nyctitropic movements are continued for a very much longer period than when such do not exist.  This has been amply proved in the case of cotyledons, and Pfeffer has given observations to the same effect with respect [page 379[97]] to leaves.  We have seen that a leaf of Mimosa pudica continued to move in the ordinary manner, though somewhat more simply, until it withered and died.  It may be added that some leaflets of Trifolium pratense were pinned open during 10 days, and on the first evening after being released they rose up and slept in the usual manner.  Besides the long continuance of the movements when effected by the aid of a pulvinus (and this appears to be the final cause of its development), a twisting movement at night, as Pfeffer has remarked, is almost confined to leaves thus provided.

It is a very general rule that the first true leaf, though it may differ somewhat in shape from the leaves on the mature plant, yet sleeps like them; and this occurs quite independently of the fact whether or not the cotyledons themselves sleep, or whether they sleep in the same manner.  But with Phaseolus Roxburghii the first unifoliate leaves rise at night almost sufficiently to be said to sleep, whilst the leaflets of the secondary trifoliate leaves sink vertically at night.  On young plants of Sida rhombaefolia, only a few inches in height, the leaves did not sleep, though on rather older plants they rose up vertically at night.  On the other hand, the leaves on very young plants of Cytisus fragrans slept in a conspicuous manner, whilst on old and vigorous bushes kept in the greenhouse, the leaves did not exhibit any plain nyctitropic movement.  In the genus Lotus the basal stipule-like leaflets rise up vertically at night, and are provided with pulvini.

As already remarked, when leaves or leaflets change their position greatly at night and by complicated movements, it can hardly be doubted that these must be in some manner beneficial to the plant.  If so, we [page 398] must extend the same conclusion to a large number of sleeping plants; for the most complicated and the simplest nyctitropic movements are connected together by the finest gradations.  But owing to the causes specified in the beginning of this chapter, it is impossible in some few cases to determine whether or not certain movements should be called nyctitropic.  Generally, the position which the leaves occupy at night indicates with sufficient clearness, that the benefit thus derived, is the protection of their upper surfaces from radiation into the open sky, and in many cases the mutual protection of all the parts from cold by their being brought into close approximation.  It should be remembered that it was proved in the last chapter, that leaves compelled to remain extended horizontally at night, suffered much more from radiation than those which were allowed to assume their normal vertical position.