In all the species of Oxalis observed by us, the cotyledons are provided with pulvini; but this organ has become more or less rudimentary in O. corniculata, and the amount of upward movement of its cotyledons at night is very variable, but is never enough to be called sleep.  We omitted to ascertain whether the cotyledons of Geranium rotundifolium possess pulvini.  In the Leguminosae all the cotyledons which sleep, as far as we have seen, are provided with pulvini.  But with Lotus Jacobaeus, these are not fully developed during the first few days of the life of the seedling, and the cotyledons do not then rise much at night.  With Trifolium strictum the blades of the cotyledons rise at night by the aid of their pulvini; whilst the petiole of one cotyledon twists half-round at the same time, independently of its pulvinus.

As a general rule, cotyledons which are provided with pulvini continue to rise or sink at night during a much longer period than those destitute of this organ.  In this latter case the movement no doubt depends on alternately greater growth on the upper and lower side of the petiole, or of the blade, or of both, preceded probably by the increased turgescence of the growing cells.  Such movements generally last for a very short period—­ for instance, with Brassica and Githago for 4 or 5 nights, with Beta for 2 or 3, and with [page 314] Raphanus for only a single night.  There are, however, some strong exceptions to this rule, as the cotyledons of Gossypium, Anoda and Ipomoea do not possess pulvini, yet continue to move and to grow for a long time.  We thought at first that when the movement lasted for only 2 or 3 nights, it could hardly be of any service to the plant, and hardly deserved to be called sleep; but as many quickly-growing leaves sleep for only a few nights, and as cotyledons are rapidly developed and soon complete their growth, this doubt now seems to us not well-founded, more especially as these movements are in many instances so strongly pronounced.  We may here mention another point of similarity between sleeping leaves and cotyledons, namely, that some of the latter (for instance, those of Cassia and Githago) are easily affected by the absence of light; and they then either close, or if closed do not open; whereas others (as with the cotyledons of Oxalis) are very little affected by light.  In the next chapter it will be shown that the nyctitropic movements both of cotyledons and leaves consist of a modified form of circumnutation.

As in the Leguminosae and Oxalidae, the leaves and the cotyledons of the same species generally sleep, the idea at first naturally occurred to us, that the sleep of the cotyledons was merely an early development of a habit proper to a more advanced stage of life.  But no such explanation can be admitted, although there seems to be some connection, as might have been expected, between the two sets of cases.  For the leaves of many plants sleep, whilst their cotyledons