296 THE BIOLOGY OF FLOWERING PLANTS 



or in the bark of trees, but are only moderately well fixed. 

 The security of the ivy is due rather to the great number of 

 roots than to their individual efficiency ; and, indeed, ivy 

 growing on a wall is not really secure, for it is very readily 

 torn off. Growing round a tree it is much better off, and 

 this is the more natural position. The anchoring roots 

 are not much use in water absorption, being soon cut off by 

 cork formation. 



Root climbers are not infrequent in tropical rain forests. 

 Schenck (1892) describes examples from some twenty 

 different families in Brazil. Many are epiphytic, and in 

 all the genera which include root climbers epiphytic species 

 are also found. The possibilities for the evolution of 

 epiphytic types from root climbers are obvious. Only 

 rarely are adventitious roots produced as a supplementary 

 mode of attachment by plants climbing by other means. 



Twining Plants.— A much higher type of organisation 

 is shown by the twining plants in which the apex of the stem 

 carries out a regular circling movement or circumnutatmi, 

 which enables it to twine round a suitable support. This 

 curious movement is induced and regulated by the action 

 of gravity on the inclined tip of the stem. On the clinostat 

 a twining stem, such as that of the scarlet runner, ceases 

 its regular circling and shows only irregular swaying move- 

 ments ; if the plant is inverted the tip bends up and circles, 

 in the opposite direction relative to the plant, but in the 

 same direction relative to the axis of the earth. The move- 

 ment is due to the more rapid growth of one flank of the 

 stem ; as the circling movement is necessarily accompanied 

 by a torsion this flank is constantly changing. If a particular 

 longitudinal strip of tissue hes below at a given instant, 

 it gradually twists round till it lies at the side ; the increased 

 growth rate now sets in and carries on the circling movement. 

 As this proceeds the strip we are following goes on twisting 

 till it lies above, then on the opposite side, and finally once 

 more below. A complete twist on the axis accompanies 

 each complete revolution of the tip. The stimulus to more 

 rapid growth occurs when a given strip lies below. Gradman 



