47 2 LECTURE XVIII. 



surfaces downwards, are ascribed by de Vries to the unequal 

 weight of the leaves on the two sides. He found that when 

 lateral branches ( Uhnus campestris, Celtis anstralis, Rhodotypus 

 kerrioideS) etc.) from which the leaves had been removed, were 

 fixed horizontally in the inverse position, whilst still in con- 

 nexion with the plant, they curved upwards without undergo- 

 ing any torsion ; whereas branches bearing leaves, when treated 

 in a similar manner, underwent torsion as they attempted to 

 curve upwards. De Vries explains the torsion in the latter 

 case by saying that the leaves, when in the inverse position, 

 curved upwards in consequence of negative geotropism and of 

 epinasty, and since the curvatures were not quite uniform, 

 the mechanical moment of the weight of the leaves on the 

 two sides of the branch was unequal, and tended to twist the 

 branch on its own axis. He points out that the normal 

 torsions which take place in the internodes of branches bear- 

 ing decussate leaves, which are such that the leaves come to 

 lie in two planes, are due, not as Frank suggested in the case 

 of Deutzia scabra, to diageotropism, but simply to the me- 

 chanical effect of the unequal weight of the leaves. The same 

 cause, de Vries believes, induces the torsion of leaves when 

 placed in the inverse position. When a mid-rib, freed from 

 the lamina, is placed in this position, it curves upwards with- 

 out torsion ; when, however, the leaf is entire torsion results 

 so that the morphologically superior surface curves to the 

 uppermost. This torsion de Vries ascribes to the fact that 

 the mid-rib does not curve upwards exactly in a vertical plane, 

 but tends somewhat to one side ; consequently the strain on 

 the two sides is unequal and torsion results. 



In support of the assumption of the diageotropism of dorsi- 

 ventral shoots it may be pointed out, as Frank does, that it is 

 not obvious in de Vries' explanation why the torsion should 

 just cease when the leaves come to lie horizontally again, for 

 it is precisely in this position that the inequality of the 

 mechanical moment on the two sides of the shoot would exert 

 its greatest twisting effect. It would be quite intelligible that 

 they should so twist the axis as to come to lie in a vertical 

 plane, but this is not the position which they assume. 



