6o 
The Podostemaceae of India and Ceylon. 
this latter factor in itself would not appear to be the cause of the 
specialisation in external morphology, attained by the Indian 
members of the order. There seems, however, no reason for 
assuming that it has played no part in the evolution of the plants 
in question. 
In general the plant in the Podostemaceae may be said to 
consist of a horizontal portion, which is in more or less intimate 
contact with the substratum and from which numerous secondary 
shoots are endogenously formed. This horizontal part is apparently 
of root-nature (except in Lawla ), but it is preferable not to 
denominate it root, as most of the structures concerned would 
scarcely come under our ordinary conception of this organ. The 
term thallus, w'hich Mr. Willis proposes, seems quite useful in this 
connection. In most cases the growing-point of this thallus has 
a distinct collenchymatous cap ; the branching is either endogenous 
or exogenous. Starting with Tristicha ramosissima, which inhabits 
relatively slow-flowing water and has a purely filamentous thallus, 
a whole series of stages may be found ending in the broad, flattened 
liverwort-or lichen-like structures of the Hydrobiymiis of the rapids. 
The thallus buds out laterally (and generally endogenously) 
from the primary axis which is developed from the germinating 
seed. The embryo is always destitute of a radicle and the hypocotyl 
immediately bends towards the substratum, to w'hich it becomes 
attached by rhizoids and frequently also by the haptera, so charac¬ 
teristic of the order. In general the primary axis remains quite 
short, only four or five leaves beyond the cotyledons being developed ; 
In Hydrobryum olivaceum how r ever it may attain a height of as much 
as 5 cm., and bears a number of long leaves at its apex, whilst 
certain very much elongated shoots, found in connection with the 
adult thallus and secondary shootsof Willisia selagiuoides , have been 
interpreted by Mr. Willis as the primary shoots of this plant. Such 
cases lead to the assumption that the Podostemaceae have arisen 
from ancestors with w r ell-developed primary shoots, from which 
lateral adventitious roots arose. 
It is impossible in the limited space I have at my disposal, to 
give any conception of the variety of form that can he assumed by 
the mature thallus, and I will only mention that in some species of 
Dicraea and Hydrobryum considerable polymorphism of this organ 
of the plant prevails. The numerous secondary shoots, which arise 
In Tristicha and Podostemon they are well developed, whilst in the 
