HYGROSCOPIC MECHANISMS 553 



parenchymatous, but has somewhat oblique pits. Next the inner 

 surface of the bract which becomes the convex side on drying there 

 is a second prosenchymatous layer with very oblique or longitudinal 

 pits, a feature which shows it to be the component antagonistic to the 

 outermost layer. Evidently the axes of maximum contraction of the 

 two prosenchymatous layers intersect one another at right angles ; on 

 drying, the outermost undergoes the greatest, the innermost the 

 smallest, amount of contraction in the longitudinal direction, and the 

 whole bract consequently curls outwards. 



In the genus Campanula the valves of the capsule, which likewise 

 curl and uncurl longitudinally, have a structure very similar to that 

 of the bracts of Centaurea. Here there is an outermost zone of 

 transversely pitted parenchymatous elements, a middle zone of somewhat 

 more elongated cells with slightly oblique pits, and an innermost layer 

 of parenchymatous cells with pits which are longitudinal or very oblique. 



According to Steinbrinck, the branches of the familiar Rose of 

 Jericho (Anastatica hierochuntica), which curl inwards when dry and 

 outwards when wet, contain a mechanism of the same type. In the 

 adaxial half of each branch the fibrous motor-cells are transversely 

 pitted, while the corresponding elements in the abaxial half have very 

 oblique pits ; the antagonism is, therefore, of the same nature as in 

 Centaurea and Campanula. 



In the cases which have just been described, the cells of all the 

 component layers are elongated in the same direction, namely, in the 

 same sense as the whole organ. The requisite antagonism between 

 opposite sides is therefore attained, not by any special orientation of 

 the active cells, but mainly by differences in the micellar structure of 

 the walls in opposite layers. In other words, the axes of maximum 

 contraction of the cell- walls run in different directions on the two 

 antagonistic sides. It is evident that the same effect might be 

 produced in a different way, if cells all possessing the same micellar 

 structure were arranged in rows which ran in different directions on 

 the two sides of the organ. Let us, for example, consider the case of 

 an organ in which the cells are all elongated transversely on one side, 

 and longitudinally on the other. It is assumed that all the cell-walls 

 consist of longitudinal or very oblique micellar rows as shown by the 

 orientation of the pits so that they contract more strongly in the 

 transverse than in the longitudinal direction. Such an organ, on 

 drying, will tend to curve towards the side composed of transversely 

 elongated cells. In order, however, that the elements in question may 

 be able to overcome the resistance of the longitudinally extended 

 elements, they must either form a thicker layer or be endowed with a 

 greater power of imbibition. 



