DEHISCENCE OF STAMEN 



343 



side this is restricted by the thickened wall, but outwards the 

 radial walls are drawn together in folds between the thickened 

 strips. This process occasions a relatively much greater 

 shrinkage of the outer surface of the anther wall than of the 

 inner surface ; the wall contracts and also, after rupture has 

 taken place, bends outwards and the opening gapes ; the 

 wall may ultimately be rolled quite back. This explanation 

 has been combated, e.g. by Schipps (1914), but the evidence 

 in its favour seems to be very strong. It will be noted that 

 the mechanism is in principle the same as that which causes 

 the opening of the fern sporangium by the annulus. The 

 well-known springing back of the fern sporangium wall, 

 which is responsible 

 for jerking out the 

 spores, does not oc- 

 cur in the anthers, 

 because, when air 

 ultimately enters the 

 drying cells and the 

 cohesion tension is 

 thereby released, the 

 cell does not resume 

 its original shape 

 and size. This is 

 due to the pleats of 

 the delicate wall 

 sticking together. 



In the Cycads the mode of thickening of the exothecial 

 cells is strongly reminiscent of the fern annulus. In the 

 other gymnosperms the thickening is less marked and regular. 

 The mechanism of opening is the same throughout. It has 

 been retained, an extremely conservative feature, through 

 the whole terrestrial flora, though in the angiosperms it is 

 carried out by a tissue which is not homologous with that in 

 the lower forms. 



One or two departures from the typical opening by slits 

 may be mentioned. We have already alluded to the 

 absence of an endothecium in submerged plants, where 



Fig. 48. — Endothecium of Lilium candidum: 

 I, from unopened anther ; 2, from opened 

 anther. (After Steinbrinck.) 



