382 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



and a transverse crushing. Fig. 98 (after Keener) shows three 

 bundles or strands opposite other three ; if each pair were connected 

 by a vertical hand of stiff, woody fibres, then they might represent 

 three girders combined in the centre ; but there is nothing of the sort. 

 Moreover, there is no certainty of every strand having another exactly 

 opposite to it. In dicotyledonous stems, as of a primrose, it is usual 

 to find five, which ultimately supply the strands for the floral organs. 

 Of these no two can be opposite one another. 



Asking an engineer for his opinion on the point, he said it would 

 be as I have stated ; but he added, the stem being filled with pith 

 would more nearly resemble a hollow iron pillar or stanchion when 

 filled with concrete, only the pith, of course, offers no analogy with 

 such a rigid substance. 



If the strands are all coherent into a hollow cylinder without any 

 pith, then the latter is clearly not required for support ; and we get 

 the tubular form of a grass-stem and many other herbs. 



As the wood is formed in response to the strain of gravity, it is 

 interesting to find this " acquired character " is hereditary ; for 

 seedlings of trees and shrubs are already " miniature trees," having a 

 complete cyhnder of wood filled with pith, though they may be no 

 more than three or four inches in height. 



Nature now adopts engineering processes, as the tree grows to 

 maturity. If it has, hke the mahogany tree, a tall cylindrical trunk, 

 bearing a great weight of branches and foliage at the summit, aerial 

 roots take on the form of flat buttresses. This sometimes occurs with 

 short trunks when the superstructures are of great weight, as in 

 species of Stermlia. They sometimes resemble " flying buttresses," 

 when there is a hollow space near the base. 



A similar method of supporting wrought-iron columns is adopted 

 by placing iron plates, called ' ' gussets, ' ' vertically against the base of 

 the pillar. The pillar may be square or polygonal. The pillars 

 are finally riveted to an iron " bed-plate " on a foundation of brick- 

 work. So, while the " gussets " resemble the root-buttresses, the 

 " plate " might be compared to the spreading horizontal roots in 

 the soil. Moreover, sometimes the superficial roots get united by 

 fusion (as in Phytolacca). This ma}^ sometimes be seen in our elm 

 trees, when the roots are exposed on a bank. Such would presumably 

 strengthen them in supporting the trunk. Though it seems to be 

 an accidental occurrence, it may be compared with the " bed-plate " 

 for iron columns, mentioned above. 



Pile-roots. — In all Monocotyledons and aquatic Dicotyledons, the 

 original primary, axial, or tap-root is suppressed. Secondary 

 " adventitious " roots arise from the nodes of the stem. When 

 formed in the air, they somewhat resemble tent-ropes, as seen in 

 screw-pines {Pandanus). 



Aerial roots can descend from the branches of certain trees, as of 

 the banyan or Indian fig, in land, and from those of mangroves over 

 water. Similarly pine-trees growing in sandy soil send out long 



