228 Weiss . — A Stigmaria with Centripetal Wood. 
appear to be in the periderm and not in the middle cortex as in the 
Stigmaria under consideration. 
On the other hand, the course of the lateral bundles through the 
secondary wood is, in some respects, more like that of a leaf-trace than of a 
rootlet, though its connexion with the secondary tracheids through which 
it passes is in keeping with the Stigmarian rootlets (see Figs. 3 and 4 , 
PI. XV). The bundles, in their origin and course, agree, too, very nearly 
with those of the Stigmaria from the Champs des Borgis, Autun, described 
by Renault (’93). There remains to be taken into consideration the 
arrangement of the protoxylem with its obvious centripetal development. 
Seeing that Renault has described some undoubted instances of Stigmarian 
axes with centripetal wood, this arrangement, though more characteristic 
of Lepidodendroid stems than Stigmarian axes, must not weigh too much 
against the above-mentioned structure. In fact, its occurrence in conjunction 
with the peculiar centric lateral bundles and the reticulate tracheids goes 
far, in my mind, to establish, not an identity but a correspondence, between 
the axis under consideration and the Stigmaria Brardii of Renault. I 
cannot consider them as identical, for the amount of primary centripetal 
wood and the wavy outline of the latter in Stigmaria Brardii obviously dis- 
tinguish it clearly from our specimen. The same may be said of the other 
Stigmaria with centripetal wood described by Renault from the Champs 
des Borgis, and with which our Stigmaria does not entirely agree. 
On the other hand, the very close agreement of the primary wood of 
our Stigmaria with that of the Lepidodendron mundum (now identified 
with Bothrodendron) suggests the correlation of these two plant-remains. 
While believing the specimen to be of Stigmarian character, I would 
agree so far with Williamson in considering that it might very well belong 
to the same plant ( Bothrodendron ) to which the leaf-bearing axis described 
under the name of Lepidodendron mundum belongs. This latter is gene- 
rally of small dimensions, and all specimens, except for the single one 
described by Williamson, and which I consider identical with the Stigmaria 
now under consideration, have been devoid of secondary thickening. This 
need not, however, be considered an obstacle, for in the first place the 
fragments so far described may be secondary branches, while, even should 
they be main shoots, it is quite conceivable that secondary growth may 
have only taken place in the basal region of the same, and have therefore 
only been of limited occurrence. Such an occurrence has been described for 
Psilotum by Boodle (’04) and by Miss Ford (’04). Should the same apply to 
Bothrodendron , the secondary growth might occur either at the base of the 
aerial axis or in the Stigmarian region, or more probably in both. 
I incline to the opinion that the axis described in the preceding pages 
is, as stated above, the Stigmarian portion, both from the very wide nature 
of the cortex, and from the peculiar lateral bundles, which differ materially 
