T 77 
Anatomy of Acrostichum aureum. 
be referred to the altered circumstances than in the case of shoot 
structures, which come under the influence of far more complicated 
factors. 
This being so, it is not without interest in this connection, to 
consider the structure of the roots of fossil plants known to us. 
The markedly lacunar character of the middle cortex of Cala- 
mite roots instantly occurs to one, and indeed the young root of 
A.aureum (PI. IV.,fig. 8) hears a surprising resemblance to the young 
Calamite root'. Still more striking is the immense development of 
middle cortex in Stigmarian roots, where the tissue was evidently of a 
very delicate nature, for it is rarely preserved except at the edges, 
where it joins the inner or outer cortex. It is highly probable that 
is was lacunar in character. A lacunar middle cortex has been 
described by Bower 2 in the axis of Lepidostrobus Brownii, and the 
prevalence of three distinct zones in most Lepidodendra, and to a 
less extent in living Lycopsids 3 is commented upon. The middle 
zone is always composed of delicate parenchymatous cells, arranged 
loosely, and frequently forming lacunae. The inner and outer 
cortical zones are generally each composed of a few layers of thick- 
walled or sclerenchymatous cells. 
Turning now to the fossil Ferns and Gymnosperms we find 
here also plentiful instances of lacunar cortices. Among the Ferns 
proper the roots of Psaronius show a marked lacunar cortex which 
is also found in “ Kaloxylon ”—a root probably to be assigned to 
Lyginodendron, but at any rate belonging to a member of the 
Pteridosperms. 
Vascular Anatomy. 
Root. 
The roots, both large and small, are formed on a triarch plan, 
but the stouter roots have in addition three alternating protoxylem 
groups, quite different in character from the main groups. The 
xylem arms of the small triarch roots are very wide—almost as 
extended as the alternating phloem groups—and each blunt 
extremity is obviously composed of two protoxylems, reminding one 
of the structure of a Sphenophyllutn stem. I can find nothing 
quite comparable in any described Fern. 
1 Williamson & Scott. “ Further Observations, Etc.” Part II., 
The Roots of Calamites, Phil. Trans., Vol. 186, B. 1895. 
3 F. O. Bower. Structure of the axis of Lepidostrobus Brownii , 
Annals of Botany, 1893. 
3 The name “ Lycopsida ” is used throughout in the sense pro¬ 
posed by Professor Jeffrey to include all “Vascular Crypto¬ 
gams other than Ferns,” 
