438 TRANSFORMATION OF ENERGY 



ment of curving suggests further lines of research, but his statements are open 

 to criticism.] The period of duration of the stimulus affects very markedly 

 the beginning of the geotropic curving, for while roots of Lupinus, kept in the 

 horizontal position for thirty-five, forty, fifty, sixty minutes, bend rapidly one 

 after the other on the klinostat, so that after ninety minutes the reaction is visible 

 in every one of them, the curving does not begin for two or three hours when the 

 roots are exposed for only twenty minutes. We may suppose that a stimulus 

 applied for a still shorter period, e. g. less than the latent period, must still have 

 some slight effect on the plant, although it does not induce a visible curving. 

 Indeed, there are observations recorded which confirm this supposition, i. e. ex- 

 periments on intermittent stimuli. If the root of Linum be placed horizontally 

 for two minutes and vertically for six minutes alternately, geotropic curvature 

 takes place after a certain time, although each separate stimulus is far shorter in 

 duration than the latent period, and hence is unable alone to induce a curving. 

 If each individual stimulus made no permanent impression on the plant it would 

 not be possible for a summation of these stimuli to result in a geotropic reaction. 

 We must assume that every stimulus, however short it be in the period of applica- 

 tion, produces some internal change, which we may describe as an excitation. This 

 excitation lasts longer than the stimulus, and hence a summation of excitations 

 is possible, and when this summation reaches a certain amount then the liminal 

 intensity is exceeded and curvature begins. Detailed research is stiU required 

 before we can say how small the stimulus periods may be when these follow each 

 other at regular intervals and also how great the intervals may be between the 

 individual stimuli, for it stands to reason that there must be limits to both. 

 The matter is of importance, for the whole theory of the klinostat rests on the 

 results of such experiments. We are as yet quite ignorant, for instance, whether 

 plants are really geotropically stimulated at all when placed on a klinostat or 

 whether the individual stimuli neutralize each other. Czapek till recently (comp. 

 1902, 468) held the former view. We may imagine with him the uniform rotation 

 of the klinostat replaced by four successive jerks, so that the plant for a certain 

 time is allowed to rest in each of the four chief positions, viz. above, right, 

 below, left. The plant, according to Czapek, must remain in each of these 

 positions for a briefer time than the length of its latent period; thus if the latent 

 period be twenty minutes one complete rotation may be effected in sixty minutes, 

 so that the plant remains for fifteen minutes in each of the four positions, i. e. less 

 than the latent period and hence is not stimulated. According to Noll (1900), 

 however, we have in this case to deal with intermittent stimuli, for at forty- 

 five minutes intervals, a definite face is brought under the influence of gravity 

 for fifteen minutes, but bending cannot take place because each stimulus is again 

 neutralized by the corresponding stimulus applied when the plant is in the op- 

 posite position. In many organs it is quite immaterial, so far as the result is con- 

 cerned, whether on the khnostat the reaction only or the stimulus as well ceases ; 

 the nodes of grasses, however, behave quite differently in either case, they must 

 be able to discriminate between the two possibilities. If laid horizontally they 

 not only bend but also exhibit renewed growth. If they be placed in a klinostat 

 they begin to grow (Elfving, 1884), but their growth is uniform on all sides. 

 This behaviour of the grass nodes appears to prove (compare Pfeffer, Phys. II, 

 126, and Noll, 1902, p. 413) that the movement of the klinostat inhibits the 

 geotropic ctirvature but not the geotropic stimiihis. If this conclusion be correct, 

 then undoubtedly an alteration in the rate of growth on the klinostat may be 

 universally demonstrated where an organ laid horizontally grows axially more 

 slowly or more rapidly than when placed vertically ; it (e. g. Hippuris) must 

 also grow more rapidly on the klinostat than under normal conditions. Only 

 such organs as exhibit a retardation on the concave side equal to the acceleration 

 on the convex side may go on growing in an unaltered form on the klinostat. 



