Sykes . — The Anatomy and Morphology of Tmesipteris. 79 
into a ring, and this suggestion is supported by the breaks found occasion- 
ally by Miss Ford in the xylem ring in P silo turn} It seems also fairly 
easy to explain the loss of secondary thickening in Tmesipteris . In both 
of these slender-stemmed plants only a small amount of mechanical and 
conducting tissues is required. In some parts of the larger specimens of 
Psilotum the extra needs are supplied by secondary elements, but in 
Tmesipteris the centrifugal xylem, which here, as everywhere, forms the leaf- 
traces, is sufficient, in addition to the centripetal wood present in both 
plants, for all requirements. The lignification of the phloem in Tmesipteris 
may also be of mechanical importance. 
Again, one might, of course, consider the secondary thickening in 
Psilotum to be an adaptive and not a primitive character. When the 
centrifugal xylem became no longer necessary for the formation of leaf- 
traces, it was probably no longer formed. But insufficient conducting 
elements would then be present in such regions as the bases of branches 
or the transition region of large plants, and here a power of secondary 
growth may have arisen. If this last interpretation be the true one, it may 
be said that in Psilotum centrifugal xylem is no longer formed by primary 
growth, but in certain cases is produced secondarily by a cambium. 
The discovery of mesarch structure in the lower parts of some Psilotum 
stems 1 2 may be supposed to suggest this as an ancestral character. 
It has recently become usual to compare the Psilotaceae with Spheno- 
phyllales 3 and to draw conclusions as to relationship from such a com- 
parison. This comparison is based on the following characters : — 
1. The rudimentary secondary thickening in Psilotum, and the general 
resemblance of the exarch rayed xylem mass (especially in triarch steles) 
of that plant, and the similar xylem mass in Sphenophyllum. 
2. The mesarch xylem of Tmesipteris , separated into groups, recalls 
the structure of Cheirostrobus . 4 But anatomical characters such as 
mesarch protoxylem and a power of secondary growth in thickness 5 are 
so much a question of habit — the presence of mesarch xylem in particular 
being commonly correlated with the presence of leaf-traces — that it seems 
arbitrary to found theories of relationship on them. 
3. The real basis of such theories lies therefore on the third comparison 
made between the two families, a comparison based entirely on prevalent 
conceptions of the morphological nature of the * sporophyll ’ in the 
Psilotaceae. It will now be necessary to discuss the morphology of that 
organ. 
The interpretation of the structure, both of the whole fertile branch 
and of the synangium itself, is a problem of much difficulty. The first 
1 Ford, 1904, PI. XXXIX, Fig. 13. 2 Boodle, 1904. 
3 Scott, Studies, 1900. Scott, ‘Cheirostrobus/ Phil. Trans. 1897, p. 27; and Thomas, 1902. 
4 Scott, 1897, p. 417. 5 See above, pp. 78-9. 
