Danaea and other Marattiaceae. 525 
is frequently not a well-marked tissue (cf. Fig. 9, PI. XXII). 
It may even be represented by an isolated cell or two 1 , the 
endodermis abutting directly on the phloem. If, as seems to 
be the case, the so-called pericycle consists of sister-cells of 
the endodermis, then it is not a pericycle by the definition 
that makes that layer the outermost one of the stele. Not 
infrequently the inner of the two sister-cells has cuticularized 
radial walls as well as the outer. In one or two cases un- 
doubted phloem-parenchyma cells likewise had cuticularized 
cell-walls. On that account it would seem that there is 
very little constancy in the position of the histologically 
differentiated endodermal cells. In the seedlings of Danaea 
simplicifolia , however, the endodermis is on the whole a well- 
marked layer, and, as a matter of convenience, it will be 
treated as the boundary-] ayer between the stele or meristeles 
and the rest of the extrastelar tissue. At the next stage to 
the constitution of the haplostele there seems to be a distinct 
difference between Danaea simplicifolia and Angiopteris. The 
observations of Leclerc du Sablon are confirmed by Farmer 
and Hill with regard to the appearance of parenchyma in the 
centre of the xylem of the haplostele, which increases in 
amount till what they call a pith is present. In Danaea 
there is no such distinct pith differentiated, and the very 
small amount of parenchyma which does appear is clearly 
to separate the xylem of the outgoing leaf-trace from that 
of the haplostele (Plate XXII, Figs. 1-3). The earlier leaf- 
traces are usually separated off before the first lateral root 
joins the stele, tracing from below upwards 2 , consequently 
1 Cf. Fig. 20, prothallus of Danaea , &c., loc. cit. 
2 In all cases the development is described from below upwards, and hence the 
roots will be spoken of as coming in to the stele, but at the same time it has to be 
clearly borne in mind that the roots arise from the endodermis (this has not yet 
been definitely proved) and pass outwards and downwards. It is also important 
to remember that there is a large area of junction between the stele of the root and 
that of the stem, and that the root-junction very considerably affects the structure 
of the xylem not only in the immediate neighbourhood of the anastomosis, but 
for a considerable distance upwards and downwards. The texture of the xylem is 
much looser on account of the presence of a considerably greater amount of 
parenchyma. 
