452 TRANSFORMATION OF ENERGY 



perception and reaction is, as we have already said (p. 441), quite arbi- 

 trary. 



The organs we have hitherto dealt with are radially constructed both as 

 regards their anatomical structure and branching, but from the physiological 

 point of view they only partially assume that position in space which we 

 primarily regard as the normal position of radial organs, viz. the erect ; a large 

 number behave plagiotropically, taking up an oblique or even horizontal posi- 

 tion. We should expect a priori that all organs with dorsiventral structure 

 would be plagiotropic, and, as a matter of fact, this assumption is correct in by 

 far the majority of cases, for only a few dorsiventral parts, e. g. the shoot 

 of Vicia faba, are orthotropic. Dorsiventrality, as previously mentioned, 

 is conditioned very often by external factors, more rarely (as in lateral members, 

 by relations to the chief axis. Among external factors light takes the first 

 place and gravity may combine with it in producing the result. Since many 

 causes co-operate in inducing dorsiventrality it is of course probable that 

 movements of orientation of dorsiventral organs are for the most part condi- 

 tioned by several factors, the analysis of which is often difficult. Dorsiventral 

 shoots of subterranean rhizomes would form very suitable subjects for the 

 study of movements of orientation, but as such shoots are not at present known 

 we are compelled to have recourse to aerial organs, in which, owing simply 

 to withdrawal of light, functional disturbances take place of so radical a nature 

 that it is impossible to carry out experiments on them. Still some examples 

 are known in which it is possible to exclude the light, when gravity alone makes 

 itself evident. These cases will be sufficient to give us answers to the funda- 

 mental questions on the subject. 



Let us study first of all the experiments made by FRANK (1870) on hori- 

 zontal branches of the yew or spruce. The dorsiventrality of these bodies 

 expresses itself externally in the difference in size of the needles on the upper 

 as opposed to the under side, and also in their mode of orientation. If such 

 branches be bent vertically upwards whilst emerging from the bud, or at least 

 during the period when the shoot is still growing, we find that, both in light 

 and in darkness, a curving takes place in them, by means of which they become 

 bent downwards into the normal horizontal position, and a corresponding upward 

 curvature when they are bent towards the ground. So far the shoots of Coni- 

 ferae conform entirely with the rhizomes of Heleocharis, and possibly they also 

 have labile rest positions directed perpendicularly upwards or downwards. 

 A difference from the radial plagiotropic organs makes itself obvious if we twist 

 round the branch on its long axis so that the under side becomes upper side, the 

 whole branch remaining meanwhile horizontal. It makes efforts at once to 

 invert itself once more so as to make the originally under side face downwards, 

 a result which may be achieved in one of two ways, either by a bending, which 

 will be complete when the branch has described a curve of 180 and turned its 

 apex towards the chief stem, or, more quickly, by retaining the same direction 

 of growth, i.e. radially outwards from the stem, and by inducing a torsion of 180. 

 The former method, by curving through 180 in one plane, has been observed 

 by SACHS in isolated lateral branches of Atropa belladonna ; the latter, torsion 

 through 180, is illustrated by FRANK'S experiments on branches of Coniferae 

 still attached to the tree. If a dorsiventral branch be fixed after being twisted 

 through only 90, so that one flank faces upwards, the other downwards, it is 

 obvious that curvature in one plane cannot effect a return to the normal posi- 

 tion in relation to gravity, and torsions must take place, torsions which have 

 been observed by FRANK (1870) in branches of the second rank after the 

 vertical fixing of branches of the first rank ; the twigs of the second rank then 

 take up a position pointing obliquely upwards, those on the other side pointing 

 obliquely downwards. 



Flowers also exhibit this phenomenon, especially dorsiventral or zygo- 



