Filicales. 
231 
In others, e.g., T. spicatum , the circular stele contains irregularly 
scattered small elements. In several species of Trichomanes and 
Hymenophyllum the stele is sub-collateral, i.e., consists of a band of 
xylem surrounded by a layer of phloem occasionally interrupted on 
the lower side; in yet others the stele is truly collateral and the 
phloem is confined to one side of the xylem, the latter being often 
much reduced (3). 
Mr. Tansley regards both the Botryopterideae and the Hymeno- 
phyllaceas as primarily endarch (31) (32); and Mr. Boodle, in 
discussing the anatomy of the Hymenophyllaceae, selects as the 
primitive type the endarch stele of Trichomanes reniforme (3). 
Curiously enough Mr. Tansley selects as primitive in the 
Botryopterideae the two genera reported to be exarch. He states 
that the solid cylindrical strand of tracheides devoid of parenchyma 
found in Grammatopteris and Tubicaulis is probably the most 
primitive type of vascular strand found in vascular plants (26). He 
claims that special protoxylem has not yet been detected in these 
types ; but although no true spiral protoxylem has so far been found 
in them, there seems little reason to doubt Miss Stopes’ report that 
the stele of Tubicazdis is nearly, if not quite, exarch (29). In 
Grammatopteris, although a re-investigation of the fossil is desirable, 
it would seem from the position of the protoxylem at the ends of 
the narrow band-shaped leaf-trace, that, at least at the insertion of 
the latter, the protoxylem of the stem must be nearly, if not quite, 
peripheral. Mr. Tansley points out that Botryopteris also has a 
solid stele, but that it contains fairly constantly a central strand of 
small tracheides probably representing protoxylem. In B. ( Rachio - 
pteris) cylindrica this central small-celled tissue, supposed to 
represent protoxylem, is usually remarkably clear, but preparatory 
to branching it divides into two groups. Mr. Hick records the fact 
that when four or five such groups occur, “ as is often the case,” 
they are arranged symmetrically around the centre as though they 
had originated by the division of a smaller number (19). This 
appearance cannot well be due to branching, as this species is known 
to have branched dichotomously. If Mr. Hick’s description is 
exact the supposed protoxylem cannot, in these cases, have occupied 
a truly endarch position ; its position would appear to be somewhat 
mesarch. But even after making allowances for all such possible 
exceptions the fact remains that Botryopteris cylindrica is typically 
endarch, and this is a strong point in favour of Mr. Tansley’s view. 
But when he proceeds to argue, on grounds of wide anatomical 
