every few minutes the lines would have been more or less curvilinear, and irregular ellipses or ovals, or perhaps occasionally circles, would have been formed.  The direction of the longer axes of the ellipses made during the same day or on successive days generally changed completely, so as to stand at right angles to one another.  The number of irregular ellipses or circles made within a given time differs much with different species.  Thus with Brassica oleracea, Cerinthe major, and Cucurbita ovifera about four such figures were completed in 12 h.; whereas with Solanum palinacanthum and Opuntia basilaris, scarcely more than one.  The figures likewise differ greatly in size; thus they were very small and in some degree doubtful in Stapelia, and large in Brassica, etc.  The ellipses described by Lathyrus nissolia and Brassica were narrow, whilst those made by the Oak were broad.  The figures are often complicated by small loops and zigzag lines.

As most seedling plants before the development of true leaves are of low, sometimes very low stature, [page 108] the extreme amount of movement from side to side of their circumnutating stems was small; that of the hypocotyl of Githago segetum was about .2 of an inch, and that of Cucurbita ovifera about .28.  A very young shoot of Lathyrus nissolia moved about .14, that of an American oak .2, that of the common nut only .04, and a rather tall shoot of the Asparagus .11 of an inch.  The extreme amount of movement of the sheath-like cotyledon of Phalaris Canariensis was .3 of an inch; but it did not move very quickly, the tip crossing on one occasion five divisions of the micrometer, that is, 1/100th of an inch, in 22 m. 5 s.  A seedling Nolana prostrata travelled the same distance in 10 m. 38 s.  Seedling cabbages circumnutate much more quickly, for the tip of a cotyledon crossed 1/100th of an inch on the micrometer in 3 m. 20 s.; and this rapid movement, accompanied by incessant oscillations, was a wonderful spectacle when beheld under the microscope.

The absence of light, for at least a day, does not interfere in the least with the circumnutation of the hypocotyls, epicotyls, or young shoots of the various dicotyledonous seedlings observed by us; nor with that of the young shoots of some monocotyledons.  The circumnutation was indeed much plainer in darkness than in light, for if the light was at all lateral the stem bent towards it in a more or less zigzag course.

Finally, the hypocotyls of many seedlings are drawn during the winter into the ground, or even beneath it so that they disappear.  This remarkable process, which apparently serves for their protection, has been fully described by De Vries.* He shows that

* ‘Bot.  Zeitung,’ 1879, p. 649.  See also Winkler in ’Verhandl. des Bot.  Vereins der P. Brandenburg,’ Jahrg. xvi. p. 16, as quoted by Haberlandt, ‘Schutzeinrichungen der Keimpflanze,’ 1877, p. 52. [page 109]

it is effected by the contraction of the parenchyma-cells of the root.  But the hypocotyl itself in some cases contracts greatly, and although at first smooth becomes covered with zigzag ridges, as we observed with Githago segetum.  How much of the drawing down and burying of the hypocotyl of Opuntia basilaris was due to the contraction of this part and how much to that of the radicle, we did not observe.