82 BOTANICAL GAZETTE [FEBRUARY 
the peripheral zone, with the addition of two strands from the 
second zone; and that the gap in the second zone is repaired by a 
strand from the third, and so on through the four or five concentric 
zones that compose the stele. Toward the top of the broad stem 
the strands united into broad bands. MeETTENIUS thought that 
this might be due to the unhealthy state of the stem he used. In 
1900 Miss SHOVE (26), reinvestigating a mature stem of A. evectd, 
confirmed his theory. She found a further difference in her material 
in the absence of strands from the second zone in the leaf traces. 
Otherwise her results confirmed those of METTENIUS. 
In 1877 DE Bary (6) accepted Metrentus’ description of the 
mature stem of A. evecta, but found in the young stem a typical 
fern cylinder. In 1889 Kin (17) extended the work on juvenile 
forms to Kaulfussia aeschylifolia and Marattia fraxinea. Their 
steles consisted of a cylinder of bundles surrounding a central strand, 
which fused with the peripheral cylinder and went out as part of a 
leaf trace. He supposed that the place of the central strand was 
taken by a peripheral bundle bending in to the center. In the 
older. plants of Marattia, instead of one there were two concentric 
series of bundles in whose center again was a medullary strand. 
In Kaulfussia the central strand was single, in Marattia it was made 
up of different strands at different levels. 
Two important contributions to the knowledge of the anatomy 
of the Marattiaceae were made in 1902. (1) FARMER and HILL (9) 
investigated the young sporeling of A. evecta, filling in gaps in their 
material with Marattia fraxinea and Kaulfussia aeschylifolia. 
They found in the sporeling a single solid protostele which passe 
by a medullated stage to dictyostely or solenostely as the gaps are 
repaired. They found that the skeleton of Marattia is simpler than 
that of Angiopteris, because the leaves are not so crowded, although 
the leaf gaps are larger. (2) BREBNER (3) found that the sporeling 
of Danaea alata also starts with a protostele which does not pass 
through a medullated stage, but becomes crescentic after the 
departure of a variable number of leaf traces, and passes directly 
into the solenostelic stage. The medullary strand arises as 4 
branch from the interior of the solenostele. ; 
JeFFReY (15) in 1903 examined young stems of several species 
