554 
Bancroft.—A Contribution to our 
those of the a type—are modifications of structure which find their parallel 
among water-dwelling forms of recent plants, where they are too well known 
to require a detailed reference. 1 
It is therefore suggested that the differences between a and /3 types 
of R. cylindrica have been caused by the influence of water upon /3 
individuals. 2 
With regard to the general habit of R. cylindrica , the presence of an 
outer cortical zone of thin-walled cells, which may very possibly be com¬ 
parable with the assimilatory tissue of Psilotum , indicates that in the 
a type, which shows the characters of a land plant, the stems were exposed, 
and not of the nature of underground rhizomes. They may either have 
grown along the surface of the ground, rooting at the nodes ; or, as their 
radial organization, lax habit, and fairly large petioles would suggest, they 
may have been semi-erect, supporting themselves upon other vegetation. 
It must be remembered that any suggestions with regard to the 
ecological aspects of Coal Measure plants are necessarily very tentative ; 
there is, however, some support for the view that the apparently straggling 
plants of Rachiopteris cylindrica grew at the edge of swamps or still lagoons. 3 
The evidence drawn from the comparative structure of a and J3 types, and 
from their constant association with one another, points to the conclusion 
that the plasticity of the species-—which is not necessarily connected with 
primitiveness 4 —allowed different individuals to exist side by side, some 
above, and others below, the water level. These individuals respectively 
constitute the a and ft types, which may be regarded as habitat forms, or 
‘ ecads ’ of the species. 5 
1. The Relationships of R. cylindrica. 
Williamson, 6 in his original description of the organs referred to 
Rachiopteris cylindrica , mentioned the possibilities that they might be 
1 Cf. Schimper, A. F. W. : Pflanzengeographie auf physiologischer Grundlage, 1898. 
Land and water forms of Cardamine pratensis (p. 26, Fig. 29) and of Callitriche stagnalis 
(p. 29, Fig. 32) show the same modifications of structure as a and /3 forms of R. cylindrica. 
2 The modifications are those caused by fairly still, rather than by strongly-flowing water; cf. 
Schimper (’ 98 ), p. 27. 
If these conclusions be tenable, it is interesting to note the presence of hairs in both forms, for 
as Dombois (Einfluss der geringeren oder grosseren Feuchtigkeit der Standorte der Pflanzen auf 
deren Behaarung. Inaug.-Diss., Saarbriicken, 1887) has shown, these structures tend to be reduced 
or absent under the influence of moisture. A similar instance to that of R. cylindrica y however, is 
found in Hottonia paluslris, where glandular hairs are present on the surface of the aquatic rhizomes 
and leaves, as well as on the land plants and aerial shoots. 
8 Cf. p. 532; and foot-note 2. The excellent preservation of the specimens, and the type of 
structural modification of /3 specimens, indicate the presence of fairly still water. 
4 Cf. the case of Polygonum amphibium. 
5 Clements, F. E. : Research Methods in Ecology. 1905. See pp. 148 and 316. 
Blackman, F. F., and Tansley, A. G. : Ecology in its physiological and phytotopographical. 
Aspects. New Phyt., vol. 4, 1905, pp. 199 and 232, See p. 253. 
6 1. c., p. 351. 
