412 TRANSFORMATION OF ENERGY 



lies parallel with the long axes of the pits, the average axis of contraction is made 

 the most of in the dynamical external cells, and it operates in opposition to the 

 longest axis in the innermost cells, forming the resistant layer, (b) A difference 

 in the striation of the wall of a single cell may also occur, as in Saponaria, where 

 in the dehiscence of the capsule we have merely to deal with the external epi- 

 dermis, whose greatly thickened outer walls act as the contractile layer, whilst 

 the radial and inner walls serve the purpose of the passive layer. But, contrary 

 to one's expectation, the pits on the outer walls are not transverse, nor are those on 

 the inner wall longitudinal ; the difference is of quite another 

 character. According to STEINBRINCK (1891) the inner wall 

 bears numerous clearly marked, narrow, elliptical pores, trans- 

 versely placed, while on the outer wall the pits gradually 

 become less distinct, less numerous and more elongated and, 

 finally, in this region of maximum curvature, fade away 

 into dark narrow streaks running transversely from one 

 radial wall to the other, alternating with clear striae. It 

 is easily seen how, by this arrangement, the differential con- 

 traction is effected, for the inner wall with its short trans- 

 verse pores contracts far less than the outer wall. It should 

 be noted that STEINBRINCK (1891) found an extreme case in 

 Dianthus prolifer ; here the most external layer of the outer 

 walls of the epidermal cells acts dynamically, while the inner- 



, _ * 1 11 -.~ / . ^ . , 



2, p. 773). most layers of the same cell- walls are the resistant ones, that 



is to say, the antagonistic units are parts of the same cell- wall. 

 The way in which we have treated this subject might lead to the supposition 

 that in each individual case of hygroscopic bending only one or other of these three 

 principles came into operation, i. e. qualitative differences between the imbibitory 

 capacities of the cell-wall, differences between different layers, or differences in 

 striation. That is, however, not the case ; as a rule, combinations of these pos- 

 sibilities occur in nature, and it is only for brevity's sake that we have avoided 

 treating of such in individual cases. 



Let us now turn from curvatures in one plane to the more complex pheno- 

 mena of twinings and torsions (compare NAGELI and SCHWENDENER, 1877). In 

 these cases also the same structural principles are applicable, but we must resist 

 the temptation to discuss these in as great detail as we have the 

 simple bending movements. We meet with very noticeable 

 twistings in the two segments into which the pods of Leguminosae 

 divide on ripening. The general nature of this twisting is illus- 

 trated in Fig. 121, where the inner surface of the fruit wall main- 

 tains its internal position during twisting. Anatomical investi- 

 gation shows that the inner epidermis abuts on a sclerotic layer, 

 which alone is the factor concerned in the twisting (ZIMMERMANN, 

 1881, p. 25). All the cells of this layer are elongated, but the 

 innermost have great powers of contraction (15 per cent.) while 



x 



X 



Fig. 122. the outermost have none at all. It is possible to determine 

 anatomical differences between these fibres, but probably it is 

 not to these differences but to certain chemical differences which have not as 

 yet been elucidated that the varied behaviour of the cells is to be attributed. 

 If these fibres lay parallel to the long axis of the pod both segments would 

 simply bend inwards in a concave manner, but, as a matter of fact, they lie at 

 a sharp angle with the long axis of the legume, and thus the curvature which is 

 transverse to the direction of the fibres is also oblique to the long axis of the pod. 

 Let us assume a long narrow piece of paper folded obliquely, as represented by 

 the dotted lines in Fig. 122, it will take the form of a spiral when pulled out. 

 Although, as ZIMMERMANN has shown, the sclerotic layer is alone sufficient to 



