germinated on damp sand was fixed so that the slightly curved radicle, which was only .07 inch in length, stood almost vertically [page 39] upwards, in which position geotropism would act at first with little power.  A filament was attached near to its base, and projected at about an angle of 45o above the horizon.  The general course followed during the 11 hours of observation and during the following night is shown in the accompanying diagram (Fig. 26), and was plainly due to geotropism; but it was also clear that the radicle circumnutated.  By the next morning the tip had curved so much downwards that the filament, instead of projecting at 45o above the horizon, was nearly horizontal.  Another germinating seed was turned upside down and covered with damp sand; and a filament was fastened to the radicle so as to project at an angle of about 50o above the horizon; this radicle was .35 of an inch in length and a little curved.  The course pursued was mainly governed, as in the last case, by geotropism, but the line traced during 12 hours and magnified as before was more strongly zigzag, again showing circumnutation.

Four radicles were allowed to grow downwards over plates of smoked glass, inclined at 70o to the horizon, under the

Fig. 27.  Cucurbita ovifera:  tracks left by tips of radicles in growing downwards over smoked glass-plates, inclined at 70o to the horizon.

Fig. 28.  Cucurbita ovifera:  circumnutation of arched hypocotyl at a very early age, traced in darkness on a horizontal glass, from 8 A.M. to 10.20 A.M. on the following day.  The movement of the bead magnified 20 times, here reduced to one-half of original scale.

same conditions as in the cases of Aesculus, Phaseolus, and Vicia.  Facsimiles are here given (Fig. 27) of two of these tracks; and a third short one was almost as plainly serpentine as that at A. It was also manifest by a greater or less amount of soot having been swept off the glasses, that the tips had [page 40] pressed alternately with greater and less force on them.  There must, therefore, have been movement in at least two planes at right angles to one another.  These radicles were so delicate that they rarely had the power to sweep the glasses quite clean.  One of them had developed some lateral or secondary rootlets, which projected a few degrees beneath the horizon; and it is an important fact that three of them left distinctly serpentine tracks on the smoked surface, showing beyond doubt that they had circumnutated like the main or primary radicle.  But the tracks were so slight that they could not be traced and copied after the smoked surface had been varnished.

Fig. 29.  Cucurbita ovifera:  circumnutation of straight and vertical hypocotyl, with filament fastened transversely across its upper end, traced in darkness on a horizontal glass, from 8.30 A.M. to 8.30 P.M.  The movement of the terminal bead originally magnified about 18 times, here only 4 ½ times.