they moved in zigzag lines towards the light; and at night they again became almost upright through apogeotropism.  After about 11 A.M. they moved a little back from the light, often crossing and recrossing their former path in zigzag lines. the sky on this day varied much in brightness, and these observations merely proved that the hypocotyls were continually moving in a manner resembling circumnutation.  On a previous day which was uniformly cloudy, a hypocotyl was firmly secured to a little stick, and a filament was fixed to the larger of the two cotyledons, and its movement was traced on a vertical glass.  It fell greatly from 8.52 A.M., when the first dot was made, till 10.55 A.M.; it then rose greatly until 12.17 P.M.  Afterwards it fell a little and made a loop, but by 2.22 P.M. it had risen a little and continued rising till 9.23 P.M., when it made another loop, and at 10.30 P.M. was again rising.  These observations show that the cotyledons move [page 19] vertically up and down all day long, and as there was some slight lateral movement, they circumnutated.

Fig. 9.  Brassica oleracea:  circumnutation of hypocotyl, in darkness, traced on a horizontal glass, by means of a filament with a bead fixed across its summit, between 9.15 A.M. and 8.30 A.M. on the following morning.  Figure here reduced to one-half of original scale.

The cabbage was one of the first plants, the seedlings of which were observed by us, and we did not then know how far the circumnutation of the different parts was affected by light.  Young seedlings were therefore kept in complete darkness except for a minute or two during each observation, when they were illuminated by a small wax taper held almost vertically above them.  During the first day the hypocotyl of one changed its course 13 times (see Fig. 9); and it deserves notice that the longer axes of the figures described often cross one another at right or nearly right angles.  Another seedling was observed in the same manner, but it was much older, for it had formed a true leaf a quarter of an inch in length, and the hypocotyl was 1 3/8 inch in height.  The figure traced was a very complex one, though the movement was not so great in extent as in the last case.

The hypocotyl of another seedling of the same age was secured to a little stick, and a filament having been fixed to the midrib of one of the cotyledons, the movement of the bead was traced during 14 h. 15 m. (see Fig. 10) in darkness.  It should be noted that the chief movement of the cotyledons, namely, up and down, would be shown on a horizontal glass-plate only by the lines in the direction of the midrib (that is, [page 20] up and down, as Fig. 10 here stands) being a little lengthened or shortened; whereas any lateral movement would be well exhibited.  The present tracing shows that the cotyledon did thus move laterally (that is, from side to side in the tracing) 12 times in the 14 h. 15 m. of observation.  Therefore the cotyledons certainly circumnutated, though the chief movement was up and down in a vertical plane.