224 Weiss. — A Stigmaria with Centripetal Wood. 
This region was, of course, traversed by the rootlet bundles on their 
outward passage ; but, owing to the defective condition of the parenchyma- 
tous tissue, the direction taken by the rootlet bundles cannot be determined 
with any degree of certainty. They appear in this specimen as irregular 
in position as those figured by Williamson (’ 87 ) in Stigmaria ficoides (Plate 
XII, Figs. 39-41). In all probability they ran somewhat obliquely in this 
zone, while close to the vascular cylinder their course, as is usual in Stig- 
maria, was almost parallel to the Stigmarian axis. In this region, there- 
fore, the rootlet bundles are cut almost transversely (see Fig, 5, PL XV), 
and they will be seen to consist of a group of medium-sized tracheids 
radiating very regularly from a centre of smaller elements. The very 
regular arrangement of the tracheids is suggestive of secondary growth 
having taken place in these rootlet bundles, a condition which is common 
in Stigmaria , and the fact that the bundles have attained different dimen- 
sions (see Fig. 5, PL XV) is in favour of this view. But, while in Stigmaria 
ficoides the secondary growth is only on the outside of the monarch rootlet 
bundle, here the secondary growth has taken place evenly all round the 
primary elements. Vascular bundles showing similar centric arrangement 
have been described and figured by Renault (’ 93 ) (PL XXXIX, Fig. 1 1 ) in the 
case of Stigmaria Brardii , with which our Stigmaria agrees also in the pos- 
session of the above-described reticulate tracheids. In Stigmaria Brardii , 
however, these centrically arranged bundles were the exception rather than 
the rule, and were taken by him to be arrested branches of the Stigmarian 
axis rather than the rootlet bundles, which he considered to have secon- 
dary wood, arranged in fan-shaped manner only on the outside of the primary 
wood. In our Stigmaria , on the other hand, the cyclic arrangement is the 
normal one and is evidently that of the rootlet bundles. The primary ele- 
ments are so small and so few in number that it is impossible to distinguish 
proto- and metaxylem, but the general impression gained is that of a 
mesarch bundle, which view is supported by the appearance of the rootlet 
bundles in their course through the secondary wood (see Fig. 4, Pl. XV). 
Owing to the state of preservation there is very little remaining of the 
phloem and cambium regions, In radial longitudinal sections a few rows 
of elongated parenchymatous cells are found on the outer boundary of the 
secondary wood, but they afford very little clue as to the nature of these 
tissues. 
The secondary wood (see Fig. 2, PL XV), on the other hand, is fairly well 
preserved, and of normal type. Its nature and extent agree with that of 
the specimen described by Williamson (’ 89 ) as Lepidodendron mundum (see 
his Fig. 15, PL V). It consists of medium-sized tracheids, *04 to -05 mm. in 
diameter, arranged in regular radiating rows, and showing only slight 
irregularity, where new rays of wood elements become added as secondary 
growth proceeds. The tracheids are square in transverse section, and are of 
