CLIMBING PLANTS. 679 



outside the soft bast, such as enable the young branches of other trees to resist flexion 

 or to resume their position after bending by the wind. In the centre of the branch 

 is seen a woody cylinder, surrounded by strands of soft bast; and beyond this a 

 very voluminous parenchyma, but only a very few hard tenacious bast-fibres. It 

 is evident, therefore, why the branches break away so readily. And that they split 

 up most easily at their places of origin, i.e. where they arise from an older branch, 

 is explained by the fact that the woody cylinder is weakest at these points. The 

 method of growth of the branch is just as remarkable as its structure. When in 

 the spring leafy shoots proceed from the foliage-buds, they do not grow towards 

 the light, as in the greater number of plants, especially woody plants, but they turn 

 away from the light and seek the darkness, and even curve round projecting angles 

 into shady comers, growing into the dark crevices and clefts in the stone wall. 

 If the face is not cracked for a wide distance, but is smooth and even, the longer, 

 growing shoots always hug it closely, and take a straight course; but, as soon as a 

 fissure is reached, the shoot immediately bends round into it, much in the manner 

 characteristic of roots (cf. p. 88). While in other shrubs the young, growing shoots 

 arising from an old woody branch are directed upwards, it here frequently happens 

 that a downward course is followed. The burden of the foliage unfolding on the 

 shoots, and the consequent increase of weight cannot be regarded as the cause of this 

 bending, for not infrequently from one and the same branch, as it runs horizontally 

 over the rock wall, shoots arise side by side of similar shape, similarly leaved, and 

 of about equal weight, some of which grow downwards and others upwards. 



In this manner of growth it is unavoidable that the branches should sometimes 

 cross one another, forming a lattice-work which adheres to the rock. I have never 

 observed actual fusions of the intersecting branches in this buckthorn, but it often 

 happens that the younger branches which lie across the older are so firmly attached 

 to them that they still remain connected when large portions of the plant are re- 

 moved from the rocky wall. 



Such extensive lattice-branches have quite the appearance of a root-plexus which 

 has extended over a boulder, and we are reminded of tlie remarkable latticed root- 

 formation of certain tropical fig-trees, which will be discussed later on. There is 

 also a temptation to take the older stems of Rhamnus jjumila for roots, inasmuch 

 as they are frequently seen embedded in the clefts and crevices of rocks, a 

 phenomenon brought about in the following way. When the apex of the develop- 

 in<j, liffht-avoiding shoot reaches a dai'k cleft, it continues to grow in a manner 

 readily intelligible in the direction of the crevice, into which it nestles so far as its 

 foliage permits. Later on it becomes lignified and loses its foliage; in the following 

 year it sends up new shoots, but itself remains growing in diameter by the addition 

 of wood and bast, till sooner or later it becomes so thick that it is jammed tiglit in 

 the cleft, and resembles a root which has forced its way in. 



The lattice formation in tropical Clusiaceae, of which an illustration is given 

 on page 681, is effected in a manner quite different from that obtaining in the 

 buckthorn. The young stems of Clusiaceae grow erect, and prefer to make use of 



