250 MISS MARIETTA PALLIS ON THE 
relatively deep in most places, so that they are not likely to be carried 
to the bottom. 
In the delta, Legehalme are therefore not important, nor do I think that 
there is much chance of their reaching the bottom in a surficiently healthy 
condition for their buds to develop. Water-logged rhizomes are, in part at 
least, decayed, and decay spreads to the nodes, and hence to the buds; 
I incline therefore to believe that in the delta practically only dead 
Legehalme reach the bottom. Legehalme are certainly, even under the 
most favourable conditions, of but slight use as regards the advance of the 
reed, and thus can have no connection with the building-up of Plav which 
is so widely distributed in the delta (see p. 244) and is moreover built up 
chiefly of vertical rhizomes. 
The question arises here: what is the irreducible minimum of detached 
aerial shoot or of rhizome from whieh the reed is able to found a fresh 
colony ; viz., how many internodes must be partially decomposed, and hence 
water-logged, in order to sink a given number of whole ones? А priori it 
would appear that the piece cannot possibly contain less than two absolutely 
healthy internodes, in which case there would be one node, and therefore one 
bud, removed from the areas of decay. These considerations lead me to 
believe that the reed does not establish itself by detached pieces—except 
where it is carried down quickly by the fall of the water—or, at any rate, 
only extremely rarely. Seed, and the vegetative multiplieation of rooted 
reed-shoots, are, I believe, the two important methods by which it succeeds 
in colonizing fresh ground : obviously, experiments are desirable here. 
Each succeeding vertical reed-rhizome is, as а whole, somewhat thinner 
and shorter than its parent (p.267). The stems which are tho finest and 
longest, as measured from the surface of the water or of the Plav, are in 
general those which arise from the most basal, that is to say, the least 
compound, of the existing rhizomes (Pls. 15 & 16). The higher branches 
of the rhizomes are often closer together than the lower branches; hence the 
water circulates more freely below (see p. 259 & Pls. 12 & 13). 
Besides the scale and the bud, roots arise at the node. These are of 
two kinds: (1) mud-roots (see Pls. 11 & 12), and (2) water-roots (see 
Pls. 11-17). The mud-roots are situated on the portions of the rhizome 
buried in the mud, whereas the water-roots are situated higher up, where 
the rhizomes emerge from the mud-layer which, as a rule, covers the bottom 
of the lakes, and enter the water; and also on the stem portion in the 
water. Water-roots are more or less abundant practically at all the nodes 
in the water. Between the mud- and the water-root portion of the rhizome, 
is an intermediate portion where there are roots which are to some extent 
transitional between the two kinds (Pl. 11). The mud-roots are soft and 
white and unbranched, have a conspicuous fawn-coloured tip, and sometimes 
attain a length of 3:5 m. (about 11 feet 6 inches) and a thickness of 6 mm. 
