1 60 Saunders.— The Leaf-skin Theory of the Stem : 
Tmesipteris and Psilotum. In Tmesipteris and Psilotum the appendages 
of the stem are by general consent regarded as leaves, and in both genera 
they have been described as decurrent. 1 In Tmesipteris they are obviously 
so, and the envelopment of the axis is as self-evident here as in many of 
the Conifers. In Psilotum the very minute subulate leaves are seated on 
the ridges of the angled stem and their insertion occupies an insignificant 
fraction of the circumference in the adult axis. They are composed of 
epidermal and cortical tissue and contain no vascular strand. In P. trique- 
trum they are spirally arranged, though the divergence varies, as may 
happen, as we know, in a Spermophyte. In slender shoots, which are 
generally markedly three-angled, the divergence is plainly §. In stouter 
axes, which are roughly five-sided, it may equally plainly be seen to be 
§, and transitions can be found from one arrangement to the other. The 
ridges appear clearly as downward extensions of the leaf-insertion. When 
the divergence is § they descend past the level of insertion of the next two 
leaves before they strike one (the third) vertically below, which gives rise 
to a fresh ridge section in line with the one just terminated. In the case 
of a f divergence, each ridge section extends past the insertion level of 
four successive leaves before striking one which terminates it. The ridges 
are paler in colour than the rest of the surface, and like the free portion 
of the leaf are destitute of stomata, which are scattered thickly and 
irregularly over the broad intervening tracts (the flat sides). In P . flaccidum, 
where the leaves are arranged alternately on the edges of the flattened axis, 
the same relation holds, stomata being absent from the marginal tracts 
descending from each leaf-insertion to the next. These anatomical relations 
raise a question of considerable interest from the present point of view. 
Does the width of the Psilotum leaf where it becomes free represent the 
real width of the extension at the level of origin, as we find to be the case 
in the higher plants, however small the free portion of the leaf may be 
(see, e. g., Fig. 31), or are the free structures much-contracted tips, the 
downward extensions of which are nevertheless broad enough to meet and 
unite laterally ? In other words, do we see here a surface which is only 
partially foliar in character, a condition which so far as appears does not 
occur in the Spermophyta, or have we to suppose that a contact line runs 
down the middle of the stomata-bearing tract, passing now between and 
now through stomata indiscriminately, and that the fusion is so complete 
(or deep) as to leave no visible trace ? Among Spermophytes it is rare to 
find the insertions so small in proportion to the whole circumference that 
those belonging to the several leaves in one cycle of the spiral, if aligned at 
one level, would obviously not meet round the axis, but the succulent 
stems of Senecio articulatus and S. anteuphorbium furnish cases in point. 
1 See A. P. W. Thomas, Proc. Roy. Soc., vol. lxix, p. 349, 1902, and G. M. Sykes, Ann. Bot., 
vol. xxii, p. 73, 1908. 
