416 TRANSFORMATION OF ENERGY 



this tension becomes so great that it overcomes the cohesive force of the water 

 particles, the water continuity in the interior of the cell is ruptured, and the 

 cell membranes assume once more their original form, and the annulus bends back 

 again with a sudden jerk into its original position of rest. Its cells now appear 

 dark, for they contain only a little water distributed on their walls, and for the 

 rest a space which may in general terms be said to be full of air. The entry of 

 air is not, however, essential to the execution of the elastic recoil, since this 

 recoil occurs when the sporangia are placed under low pressure in an air pump 

 (ScHRODT, 1897) ; in that case a vacuum appears in the centre of each cell of the 

 annulus after the recoil. 



The mode of opening of the anthers of Phanerogams corresponds in all 

 essential points, according to STEINBRINCK'S researches (1898, 1899 a), with the 

 mode of dehiscence of fern sporangia. Each of the four lobes of the anther, filled 

 with pollen-grains, consists of a wall which in the ripe condition is often only two 

 cells thick. The pollen-grains are released in consequence of the power the wall 

 has of curling backwards. In this process the outer cell layer of the anther wall 

 takes no part ; the inner, usually known as the fibrous layer, holds the dyna- 

 mical elements. The thread-like thickenings on the inner walls of these cells are 

 laid down in a very characteristic manner (Fig. 125). They run almost parallel 

 with each other at regular distances apart, over more or less of the surface of the 

 lateral walls, uniting on the inner wall like the rays of a star ; the outer wall is 



destitute of any thickening. The 

 comparison with the fern annulus 

 is perfectly obvious, but the fact 

 that the lateral walls are in this 

 case unequally thickened results 

 in a difference in behaviour be- 

 tween the anther and the fern spo- 



Fig. . Anthers of Lilium candidum. isolated rangium. The fibrOUS Cells also 

 fibrous cell in the wet state. 77, the same as seen from Undergo deformation during the 



x ST '" loss of their imbibition water on 



drying. The deformation consists, 



in the first instance, in the shortening of the diameter of the cell on 

 the external face of the anther, while that of the inner side remains rigid 

 owing to its secondary thickening. The change in form does not in this case 

 express itself in an inversion of the outer wall, but in a very noticeable contrac- 

 tion of the radial walls at right angles to the lines of thickening, so that these 

 bars approximate (Fig. 125, IV). This contraction, according to SCHWEN- 

 DENER (1899) and STEINBRINCK (1901) may amount to 50, 60, or 70 per cent. 

 of the original diameter. Were this the result of simple shrivelling the con- 

 traction would far and away exceed that shown by any other cell. But 

 STEINBRINCK has shown that the contraction begins while the cavity of the 

 cell is still full of water, and hence it is obvious that it cannot be due to mere 

 shrivelling. 



As a matter of fact, the process must be explained in an entirely different 

 manner. Under the influence of the tension exerted by the imbibition water 

 the thin parts of the radial wall lying between the thickening bars are thrown 

 into folds, and hence the volume of the cell is reduced. Apart from these foldings, 

 which may be best observed in good tangential sections through the anther, 

 there is another difference to be noted between these cells and those of the fern 

 annulus. At the moment when the elasticity of the bent fibres overcomes the 

 cohesion of the imbibition water, when a bubble of air appears in the interior 

 of the cell, no jerk takes place as in the annulus, but the anther wall remains 

 in the outwardly bent concave condition. The reason for this is probably 

 that the thin-folded portions of the wall adhere to each other, and only become 



