4IO TRANSFORMATION OF ENERGY 



tudinally. Its behaviour during contraction naturally corresponds. If we think 

 of a point in the interior of the swollen cell- wall as on the surface of a sphere, 

 this surface changes during drying to an ellipsoid form ('contraction-ellipsoid ') 

 whose shortest axis is at right angles to the lamellation, while both the other axes 

 come to lie in a tangential direction. But it is not necessary, as has been pointed 

 out, that the longest axis should be coincident with the greatest length of the 

 cell, it may lie obliquely or transversely. In most cases it would be very difficult 

 to determine directly by measurement the lie of the axis of contraction during 

 desiccation, and hence it is of importance to be acquainted with indirect methods 

 of doing so. In the first place, we may draw attention to polariscopic research, 

 which lends itself to the determination of optical elasticity-ellipsoids. Experience 

 teaches us that this almost always corresponds with the lie of the axis of the 

 contraction-ellipsoid. In the second place, the direction of the thickening bands, 

 of the striations, and pits must be noted, since that corresponds to the position 

 of the longest axis of the ellipsoid, and conforms to the line of least contraction 

 in desiccation. The position of the longest axis may, however, be different in 

 the different walls of the same cell ; for example, it may run longitudinally on 

 the outer wall and transversely on the inner, and may alter in successive layers. 



It will not be necessary for us to discuss hygroscopic movements either 

 generally or in detail, for that would involve us in difficulties, due to the fact 

 that all authors do not agree as to the interpretation of the more complex cases. 

 Only a few examples need be cited here as illustrative of the chief types (for the 

 older literature see Kraus, 1866 ; Hildebrand, 1873). We begin with the 

 consideration of simple bendings such as we meet with in Anastatica hierochun- 

 tica, the Rose of Jericho, a member of the Cruciferae from the Steppes of the 

 south-eastern Mediterranean area. When the fruit is ripe the numerous divergent 

 branches dry up, and in doing so contract much more on the upper than on the 

 under sides, thus bending inwards and causing the plant to take on a spherical 

 form. When moistened they again open out, and this performance may be 

 repeated again and again. Movements of the fruits take place at the same time, 

 but into these we need not go. It is easy to show that the bending is due to 

 the wood only, and anatomical investigation of a twig shows (Volkens, 1884) 

 that this consists especially of excentric xylem fibres, much more thickened 

 and lignified on the under (convex) than the upper sides. The strongly ligni- 

 fied fibres are much less capable of absorbing water than are the feebly lignified 

 ones, and hence on desiccation the upper side of the branch contracts much 

 more markedly than the under side. The bending in this case is due to the 

 differential capacities for swelling of antagonistic tissues. In the same way 

 (Steinbrinck, 1878), each of the five mericarps in the fruit of the geranium 

 bends outwards after drying, releasing itself with a jerk from the central carpo- 

 phore, and so aiding in the dispersal of the seeds. 



The bursting of many forms of capsule is brought about in a similar manner. 

 Some part of the fruit-wall endeavours to bend outwards and the tensions set 

 up finally cause a rupture in the region of least resistance, frequently at places 

 where there are special anatomical structures, differentiating lines of dehiscence. 

 The cause of the tension, for the most part, does not lie in the different intensity of 

 imbibition of antagonistic zones but in the layering of the cells, or, in other 

 words, in the direction of the lamination or striation of the wall. The various 

 possible arrangements may be illustrated by a few examples. 



I. Differential contraction due to arrangement of cells. In the walls of the 

 segments of the capsule of Syringa we find a lignified layer which is the sole cause 

 of the bending, and which consists of six rows of elongated thick-walled 

 cells ; the innermost of these are arranged longitudinally, while the outer 

 layers are deposited obliquely and transversely. Since these cells are all alike, 

 in so far as their capacity for absorbing water is concerned, bending when 



