FLGmA OF THE SUNDRIBUNS. 
24g 
ward growing roots from which they arise. Their structure is, how- 
ever, peculiar and is apparently adapted to serve a respiratory function.* 
The root-suckers of Baen (Avicenma) , though also small, are suffi- 
ciently conspicuous owing to the main roots passing horizontally for great 
distances from the tree to which they belong and giving off, from their 
upper side, lines of soft pith-like roots that rise well above the surface 
of the mud in which the true root is buried. These root-suckers of 
Avicennia are much too soft and flexible to serve either as mechanical 
supports to the tree or to any great extent as agents in arresting silt 
and debris. Their chief function appears to be respiratory. 
In the case of the remaining species a very decided mechanical effect 
is produced by these root-suckers. Keora {Sonneratia apetala)^ which 
is characteristic of river-banks, sends out very wide-spreading roots 
under the surface of the mud. These roots emit long close lines of 
root-suckers up to distances of 150 feet or more from the parent stem. 
The lines of vertical root-shoots act as spurs that deflect the imping- 
ing current, lead tp accretion of silt, and greatly aid the roots them- 
selves in holding on to the muddy substratum. The shoots that rise 
from the distal and deeper ends of roots that are nearest to the stream, 
rise higher and are larger, than those on the upper part of the slope 
and nearer the stem of the tree. In places where the set of a current 
has become altered, so that the silt thus accumulated is again being 
removed, it is noticeable that, as the erosion goes on, a deeper 
layer of roots and root -shoots than the one actually visible at the 
surface, but belonging to the same tree, becomes bared. The-e 
are roots with their suckers that had at some former period become 
completely buried, when their place was taken by a newer and more 
superficial system. Roots and suckers thus laid bare after previous 
complete burial, appear invariably to be dead* The roots of Keora are 
slender and conical from a rather thick base and are usually quite 
discrete. Thpse of Ora [SonneraHa acida) are in most respects 
like those of Keora but are often agglutinated at their bases. In both 
species they are tough and flexible but not very rigid. 
The suckers rising from the roots of Pussur and Amur, both of 
which species affect the more low-lying swampy localities in the interior 
of The islands, are sometimes as much as 3 feet long, the tips of the 
longest suckers appearing just above the surface of the water at the 
highest tides. The majority of the suckers hardly reach this level. 
In both Pussur and Amur the suckers are cylindric butTaper more 
* For an account of the structure of these roots a paper by Gage, On the 
Anatomy of the Roots of Phcenix paludosa^ in Sc. Mem., Med. Off., Army of 
India, XII., 103 (1901), should be consulted. 
C 2 
