Ford . — The Anatomy of Psilotum triquetrum . 601 
It seems, however, to be more likely that the leaves of Psilotum 
represent reduced, and not primitive, structures, for in Tmesipteris the 
leaves and leaf traces are well developed. In this case a suggestion, 
which has been given by Tansley and Chick 1 in their paper on Schizaea 
Malaccana , may be also made for Psilotum. The stele remains the same 
in size, but the demand for water conduction is less, and hence tracheids, 
which originally occupied the centre of the stem, are no longer developed 
and their place is taken by ordinary parenchyma. The great reduction 
of the leaves in Psilotum is a point in favour of this suggestion. 
The position of the Psilotaceae amongst the Pteridophyta is not very 
clear. The family was formerly placed as a sub-group of the Lycopodineae, 
but of late this affinity has been regarded as being somewhat remote, 
whilst a stronger relationship has been shown to exist with the fossil group of 
the Sphenophyllales. Anatomically, however, the Psilotaceae show certain 
Lycopodinean features. Professor Bower in 1893 2 drew attention to the 
resemblance between the central stele of the Psilotaceae and that found in 
the axis of Lepidostrobus Brownii. He enumerates the four chief points 
of resemblance, the crenulated margin of the xylem, the definite layer of 
endodermis as in Psilotum , the slight bulk of phloem tissue and absence of 
distinctive characters in it, and lastly, the presence of a parenchymatous 
pith in Lepidostrobus which resembles that ‘ well known to occur in 
Tmesipteris , though its place is taken in Psilotum by thick- walled 
sclerenchyma.’ The presence, therefore, of a central thin-walled pith in 
the stem of Psilotum helps to strengthen the possible relationship. 
There is again a strong resemblance between the aerial and intermediate 
regions of the stem of Psilotum and the fossil stems of Lepidodendron 
mundum 3 . Some sections of the latter in the Manchester Museum show 
a distinct medulla of varying size, the xylem forming a definite ring which 
may be broken at some levels. Other sections, again, show this parenchy- 
matous pith replaced by thick-walled tissue as in Psilotum. 
The ‘ secretory ’ zone 4 of some Lepidodendroid stems is absent in 
Psilotum , but the somewhat unsatisfactory nature of the phloem may 
possibly be another link, though a slight one, between the two forms. 
The affinity of the Psilotaceae with the Sphenophyllales is based not 
only on anatomical grounds, but also on the nature of the sporangial 
apparatus 5 . From the anatomical point of view, the characteristic triangular 
mass of primary wood, which is so noticeable in the stem of Sphenophyllum , 
recalls the structure of the smaller branches of an aerial stem of Psilotum , 
and the discovery of secondary tracheids serves to strengthen this 
1 Tansley and Chick (’03), p. 499. 2 Bower (’93), p. 337. 
3 Williamson (’89), p. 197, PI. 6, Figs. 7-14. 
4 Seward (’00), p. 155 ; (’02), p. 38. Weiss (’01), p. 3. 
5 It has already been stated in this paper (p. 599) that in the Psilotaceae the synangium and its 
axis are to be regarded as the product of a single fertile leaf. 
