Circumnutation of Cotyledons.—With all the dicotyledonous seedlings described in the last chapter, the cotyledons were in constant movement, chiefly in a vertical plane, and commonly once up and once down in the course of the 24 hours. But there were many exceptions to such simplicity of movement; thus the cotyledons of Ipomœa caerulea moved 13 times either upwards or downwards in the course of 16 h.. 18 m. Those of Oxalis rosea moved in the same manner 7 times in the course of 24 h.; and those of Cassia tora described 5 irregular ellipses in 9 h. The cotyledons of some individuals of Mimosa pudica and of Lotus Jacobæus moved only once up and down in 24 h., whilst those of others performed within the same period an additional small oscillation. Thus with different species, and with different individuals of the same species, there were many gradations from a single diurnal movement to oscillations as complex as those of the Ipomœa and Cassia. The opposite cotyledons on the same seedling move to a certain extent independently of one another. This was conspicuous with those of Oxalis sensitiva, in which one cotyledon might be seen during the daytime rising up until it stood vertically, whilst the opposite one was sinking down. Although the movements of cotyledons were generally in nearly the same vertical plane, yet their upward and downward courses never exactly coincided; so that ellipses, more or less narrow, were described, and the cotyledons may safely be said to have circumnutated. Nor could this fact be accounted for by the mere increase in length of the cotyledons through growth, for this by itself would not induce any lateral movement. That there was lateral movement in some instances, as with the cotyledons of the cabbage, was evident; for these, besides moving up and down, changed their course from right to left 12 times in 14 h. 15 m. With Solanum lycopersicum the cotyledons, after falling in the forenoon, zigzagged from side to side between 12 and 4 P.M., and then commenced rising. The cotyledons of Lupinus luteus are so thick (about .08 of an inch) and fleshy,[22] that they seemed little likely to move, and were therefore observed with especial interest; they certainly moved largely up and down, and as the line traced was zigzag there was some lateral movement. The nine cotyledons of a seedling Pinus pinaster plainly circumnutated; and the figures described approached more nearly to irregular circles than to irregular ovals or ellipses. The sheath-like cotyledons of the Gramineæ circumnutate, that is, move to all sides, as plainly as do the hypocotyls or epicotyls of any dicotyledonous plants. Lastly, the very young fronds of a Fern and of a Selaginella circumnutated. [22] The cotyledons, though bright green, resemble to a certain extent hypogean ones; see the interesting discussion by Haberlandt (‘Die Schutzeinrichtungen,’ etc., 1877, p. 95), on the gradations in the Leguminosae between subaërial and subterranean cotyledons.