384 Stiles .— The Structure of the Aerial Shoots of 
the smaller branches of the first species is at any rate interesting. The 
feeble secondary thickening in both species of Psilotum also recalls the 
normal secondary growth of wood in Sphenophyllum . As regards stem 
anatomy the comparison of Tmesipteris has been rather with Cheirostrobas , 
where mesarch xylem sometimes occurred. In this respect Psilotum 
probably approaches Cheirostrobus more nearly than Tmesipteris . 
Scott’s view has been adopted by A. P. W. Thomas 1 and Bower, 2 who 
go further than Scott and place the Psilotaceae as an order of the Spheno- 
phyllales. The chief arguments against the view are to be found in the 
phyllotaxis and in the mode of branching. In the Sphenophyllales the 
leaves are borne in whorls, and from the constancy of the character 
throughout the group, Scott 3 is inclined to attach much importance to it. 
In the Sphenophyllales the branching was axillary, while in the Psilotaceae 
it is dichotomous. But as Bower 4 points out, branching is axillary in the 
Equisetales, yet dichotomy may occasionally occur, and it is easy to 
imagine that a group where both modes of branching were possible might 
give rise to one line of descendants where axillary branching prevailed, and 
to another where the branching was dichotomous. 
Moreover, it seems to be not impossible that the branching in Psilotum 
is either now axillary, as Prantl supposed, or has been derived from an 
axillary mode of branching. The constant relation of a leaf to the stem 
bifurcation seems to me to support this opinion. This would remove one 
of the chief differences between the Psilotaceae and the Sphenophyllales, 
and in any case the fairly near relationship of the two groups seems to be 
almost proved. On the other hand, the resemblances between the Psilotaceae 
and the Lycopodiales seem too great to warrant the present tendency to 
separate the two phyla widely, and to me the most logical position appears 
to be that of Scott in 1907, 5 in which he regards the Psilotales as having 
most in common with the Sphenophyllales, but as possessing also some of 
the Lycopodineous affinities that have been previously attributed to them. 
Lignier, 6 however, holds a very different view. To him the Psilotaceae 
are primitive types allied to the Lycopodiales, while the Sphenophyllales 
are related to the primitive Fern stock through Archaeopteris. Further 
back still the Lycopod phylum and Fern phylum are connected, the 
common ancestor consisting of a dichotomizing axis bearing small appendages 
or phylloids. These are represented to-day in the leaves of the Psilotaceae 
and Lycopodiales. In the Fern phylum branches of the axis were specialized 
for assimilatory purposes and became leaves. 7 Thus, on Lignier’s view, 
the leaves of the Psilotaceae and Sphenophyllales are not homologous, and 
1 Thomas, A. P. W. (’ 02 ), p. 350. 2 Bower, F. O. (’ 08 ), p. 423. 
:i Scott, D. H. (’ 07 ), p. 166 ; (’ 09 ), p. 626. 4 Bower, F. O. (’ 08 ), p. 424. 
5 Scott, D. H. (’ 07 ), p. 166. c Lignier, O. (’ 08 ), p. 95 ; (’ 08 ), p. 278. 
7 Cf. Tansley, A. G. (’ 08 ), p. 6. 
