360 
The internal endodermis, as observed by Poirault (25) 
and other authors, in B. lunaria is entirely absent in B. simplex. 
The behaviour of the endodermis in the latter species is much 
simpler than that found by Lang for B. lunaria (22). The endo- 
dermis is absent in the leaf, but surrounds the stem and root 
stele with a complete ring, only interrupted by the entrance 
of other roots. Fig. 27 shows this condition. On this figure two 
root traces and a leaf trace join the stele. The endodermis 
begins already to close around the leaf-trace, but is interrup- 
ted at the entrance of the root traces. Outside the uninterrupted 
parts of the xylem a strong cambial activity occurs in the peri- 
cycle. The only irregularity observed in the endodermis of 
B. simplex was an occasional doubling (Fig. 28). This doubling 
has nothing to do with the insertion of a leaf or root trace, as 
found by Lang in B. lunaria. 
Occasionally independent tracheids were seen in the pith. 
Bower and Lang both hold the opinion that these free tra- 
cheids represent the remnants of the old primary xylem, per- 
haps comparable to the intra-medullary xylem of the Botry- 
opteridae as observed by Seward (30) in Ankyropteris and 
Zugopteris (see also Scott. (29). The pith of B. simplex being 
entirely of intrastelar origin, it seems that opinion of Bower 
and Lang does not hold for this form. Fig. 32 shows the junc- 
tion of two root traces. There is one tracheid in the medulla, 
which is apparently nothing else than a common root tracheid, 
but bent upwards inside the medulla instead of near the 
xylem. The figures of Lang on the intra-medullary xylem of 
B. lunaria can be explained in the same way. 
Secondary Growth, Cork Formation. Tannin Cells. 
Fig. 27 shows a large pericycle surrounding the xylem. This 
pericycle is the place where the secondary growth takes place 
in the other Botrychia and in some of the Marattiaceae (West 
42). It cannot be made out if in B. simplex, that the secondary 
