248 MISS MARIETTA PALLIS ON THE 
variety are fawn, of the East Anglian purplish brown. Caryopses occur 
quite usually in both. I have not tested their germinating capacity. 
The lower portion of the reed, the rhizome, is perennial and lives under 
water. Its aerial extremity, which bears the foliage leaves and flowers, lives 
in the air and is annual. Hence rhizomes are more numerous than living 
aerial extremities. The dead stems often remain standing for two years, and 
nearly always for one (see p. 257) 
The branch-system of the reed is sympodial (see Pl. 11). The rhizome 
advances for a short distance in the horizontal direction before turning 
upwards to the surface, and then another bud on the rhizome repeais the 
process. Behind the growing-point of the horizontal rhizome, rhizomes 
also rise to the surface immediately, and these are often branched. Thus the 
reed combines both the creeping and tussock mode of growth of the grasses 
(see Pls. 11, 12, & 18). 
The direction taken by the rhizome depends largely on its environment. 
In deep water it is much branched in the vertical direction, because the buds 
which occur at the nodes in the axils of the alternately-ranged sheathing- 
scales (see text-fig. 1), and also at the nodes of its continuation the aerial 
stem, develop if immersed. Hence, in deep water, where a great length of 
reed-stem is under water, huge stools *, consisting of numerous vertical 
branches, often compound to the sixth degree, are formed. The individual 
tussock is, in fact, a small isolated branch-system, the result of the branching 
of an ascending rhizome. The primary rhizome may, in fact, be regarded as 
the trunk, and the subsequent rhizomes arising from it as its branches (see 
р. 268). 
These branch-systems—z. e., the stools—are of course connected with each 
other. That the reed forms such branch-systems is evident all through its 
life, as the grouping of the shoots both on Play and on fen indicates. In 
shallow water, on the other hand, the reed branches chiefly in the horizontal 
plane : hence it appears highly probable that the reed arising from one seed 
(see р. 252) will cover larger areas in shallow than in deep water where 
it branches vigorously in the vertical plane. In either case the reed fills the 
space under water more or less completely. 
The aerial portion of the stem does not under normal circumstances 
branch ; to do so it must suffer injury f or immersion. 1+ is, as it were, the 
* [first noticed the stool form of the reed in 1908, in the Broads of Surlingham and 
Strumpshaw, in the river Yare, in Norfolk [Арр. С]. At the Phytogeographical Excursion 
in August 1911, I showed some of these stools to the assembled British and Foreign 
botanists, and was assured by all that they had not known of the existence of the reed- 
stools before (see Pl. 22. fig. 1). 
T If an aerial extremity is injured, in the delta of the Danube, one or two of the buds 
just below the injury develop into small ill-nurtured shoots which do not flower but which 
produce a few small leaves. It occasionally happens, also, that in the reed-swamps the 
