542 Brebner . — On the Anatomy of 
so that the protoxylem becomes mesarch. Figs. 20, 21 
and 22 show the structure of the leaf-meristeles in Marattia 
attenuata , Kaulfussia aesculifolia and Danaea alata respec- 
tively. These figures show that the structure of the meri- 
steles is practically identical in the different genera. The 
difference in shape is not distinctive, because in any one 
species all these forms may be found, not to mention 
others. Fig. 20 was drawn from a secondary rachis of M. 
attenuata near its base, Fig. 21 from a leaflet-midrib of 
Kaulfussia , and Fig. 22 from the petiole of Danaea alata. 
The most interesting point in connexion with them is the 
extension of Miss Shove’s interesting discovery of the endarch 
position of the protophloem in the meristeles of the stem 
of Angiopteris to the leaf-meristeles of the Marattiaceae 
generally. If the leaf-meristeles (in transverse section) of an 
adult frond be examined, there is very little evidence of the 
protophloem, but in stained preparations an irregular line 
of darker walls, especially thick at the angles where certain 
cells appear to meet, are noticeable near the outer border 
of the xylem. This layer was early observed but its cause 
was left for future investigation. In the meantime Miss 
Shove published her discovery in the stem of Angiopteris , 
and it at once became clear that this line ( pph . in Figs. 20, 
21, 22) was due to the obliterated protophloem. That this 
was really the case was subsequently proved in Marattia 
alata, pph. Fig. 18. Sections of young fronds of Angiopteris 
evecta and Marattia cicutaefolia , made in connexion with the 
mucilage-canal investigation, show this endarch position of 
the protophloem very clearly, but it was overlooked at the 
time. 
The structure of the sieve-tubes themselves has been so 
exhaustively dealt with by Poirault 1 , that it is superfluous 
to do more than quote his results. He found that there are 
two types of sieve-tubes, the Cticurbita- type, with a single 
plate on the transverse wall, and the vine-type with very 
1 Georges Poirault, Recherches sur les cryptogames vasculaires, Ann. Sci. Nat., 
2 e ser., Bot., xvii-xviii, 1893, p. 138 et seq., and p. 192. 
