MOTILE MECHANISMS IN HIGHER PLANTS 467 



appear not to depend on a water-cohesion arrangement but on 

 the hygroscopic character of the single band which is present 

 in each of the elater cells. In this and a few other genera of 

 liverworts the dispersion of the spores is chiefly accomplished 

 bv means of a kind of elastic mechanism. The elaters form 

 cords which are attached by their ends to the wall and to the 

 base of the sporangium, so that when the walls of the capsule 

 split into the four diverging valves, the elaters are forcibly 

 extended. They are then torn away from the walls and in 

 springing back to their former length (or approximately so) they 

 cause the spores to be jerked from the capsule. 



Finally a glance may be cast at the mechanism by which the 

 stamens of a flower shed their pollen. There are, as might be 

 expected, considerable differences in detail but in the majority 

 of flowers the general result is fairly similar. The pollen is 

 commonly contained in four more or less tubular chambers 

 situated in pairs on each side of the axis of the stamen. 

 Dehiscence occurs down a line between the two chambers of 

 each pan and the pollen is exposed by the curling back of the 

 flap-like walls. The question arises as to the agency by which 

 the dehiscence and curling back is effected. 



When one watches an anther that is beginning to dehisce, it 

 is eas} 7 to see that the formation of the slit is accompanied by 

 the bulging out of a line of pollen grains ; the starting of the 

 process was stated by Schneider to be due to the internal 

 pressure of the pollen upon the walls, which when sufficiently 

 large caused a passive rupture. This explanation does not, 

 however, fit a very large number of examples nor does it take 

 into account a very general peculiarity of structure in the 

 anther. Just beneath the epidermis there is present a so-called 

 " fibrous layer " of cells. In transverse section this appears to 

 be formed of a large number of bars which extend from back to 

 front on the radial walls of the cells of the sub-epidermal layer. 

 On the back of the cell {i.e. the side abutting upon the anther 

 cavity) the fibrous thickening is continuous, whilst on the outer 

 (sub-epidermal) side of the cell it is either discontinuous or only 

 feebly developed. It must be clearly understood that each of 

 these " fibrous " cells contains nucleated protoplasm and this is 

 bounded by a thin wall, the fibrous structure forming the 

 locally thickened parts of this wall. In such a plant as the 

 Christmas Rose (Hellebore), the thickening is more or less 



