79 



therefore, in the direction of the split surface, but, on the other hand, they 

 were reduced to a minimum on the side where the roots were pressed against 

 the rock and were finally irrecognizable. On the free side of the wood the 

 vascular bundles developed very abundantly, in some annual rings, in fact, 

 the wood was very broad and provided with a thick bark ; on the side of the 

 root pressed against the rock, the wood lacked all vascular formation, was 

 short-celled and formed from wood fibres inclined diagonally instead of 



Fig-. 1. Fig. 2. 



Roots of Quercus Pedunculata grown between rocks. (After Dobner-Nobbe.) 



running vertically. Finally, differentiation into annual rings could not be 

 observed and only a very slender cork layer is seen lying on the occasionally 

 formed short-celled parenchyma, without any recognizable differentiation 

 into medullary rays. 



Nevertheless, the cambial activity was not lost in the board-like part of 

 the root as was evident when the pressure ceased, for the flattened part grew 

 normally in its cylindrical form. Anatomical changes in the roots pressed 

 between the rocks approximate so strikingly the results obtained by artificial 



