1911] CHARLES—ANATOMY OF MARATTIA 87 
also the repetition of this process for the next leaf, when the phloem 
closes around the xylem strands (fig. 17) and then disappears on 
the side.-where the former leaf went off (fig. 18). At the departure 
of leaf 3, endodermis as well as phloem penetrates between the 
two xylem strands (fig. 19), and in fig. 20 the cylinder with internal 
phloem and endodermis is established. The 
transition from protostele to solenostele with 
leaf gaps is represented in text fig. 1. 
Angiopteris, Danaea, and Maraitia agree 
more closely in structure than in interpreta- 
tion of the tissues concerned in this transition 
from protostele to solenostele. The difference 
in interpretation involves the question whether 
the tissue not xylem within the endoderm is 
cortical or stelar.. FARMER and Hitt (Q) are 
inclined to regard only vascular tissues as 
stelar. This leads to the difficulty of deter- 
mining what tissues are vascular and what are 
cortical when there is no histological differ- 
entiation. At one level they regard a tissue 
as xylem parenchyma which in another closely 
related section they call pith (9, pl. 16, figs. 
S : : Fic. 1.—Diagram 
10,11). The confusion caused by setting aside sede Gt young ken th 
as non-vascular the pith in the Marattiaceae transition from proto- 
is the greater because of the widespread _ stele at the base to 
tendency to lignify only part of the xylem pao: 
parenchyma. The primary as well as later constantly deeper cres- 
roots may have a solid xylem core, or may be _ cents in the stele. 
lignified only at the poles. The cotyledonary 
node, usually solid, may contain scattered parenchyma, and the 
bundles of the dictyostele have extensive bands of parenchyma 
(fig. 22). Another difficulty in the way of regarding this pith as 
non-vascular in the Marattiaceae is that the parenchyma is rarely 
definitely localized in the center of the stele, so that similar stages 
in the development of the stele of two plants of the same species 
would demand as different interpretations as those given to 
Angiopteris and Danaea. On the other hand, the parenchyma 
