COMPARATIVE ANATOMY 337 
essentials the same construction of the stele as in Lepidodendron, they 
illustrate’ steps towards the breaking up of the primary wood of the 
medullated stele into separate bundles. The details derived from various 
Sigillarian fossils have lately been put together in sfratigraphical sequence 
by Kidston,! and his conclusion has already been quoted above (Chapter 
XVIII, p. 230): he has shown a strong support for the view that the 
condition with primary xylem forming a closed ring surrounding the large 
medulla was the most primitive for Sigi//aria: such a structure is found 
in the more ancient specimens from the Lower Coal Measures (5. elongata, 
Brongn., and SS. e/egans, Brongn.): those from the lower Permian, however, 
(S. menardi, Brongn., and SS. spinulosa, Rost. sp.) show the primary xylem 
as a circle of separate bundles, though some of them may cohere laterally 
in the last-named species. This indicates an evolutionary progression from 
a concrete primary xylem to a condition where it is separated into strands. 
In such forms the pith, being of relatively very large size, the primary 
wood is reduced to a comparatively narrow investment round it, liable 
as we have seen to be broken up into distinct strands. The secondary 
tissues make their appearance, however, as in Lepidodendron; there being 
in Sigillarid a broad zone of secondary xylem, and a highly organised 
periderm. It is thus seen that the later Sigz//ardas have departed further 
in their structure from the simple protostele than other dendroid Lycopods, 
for they show not only medullation, and a secondary thickening, but 
breaking up of the primary xylem as well. 
It has been concluded above, on the basis of external comparison, 
that the plant of Jsoetes is like a partially differentiated -Zepzdostrobus 
seated upon a Lepidodendroid base. The question will now be how far 
its anatomy will countenance such an opinion. There has been some 
confusion in the descriptions given by various investigators, owing doubtless 
to the difficulty in decyphering a complex mass of tissues affected by 
the reduction which follows on an aquatic habit. But this has been in 
great measure cleared by Scott and Hill in their Memoir on Jsoedes 
Aystrix, one of the few land-growing species.? Nevertheless the terrestrial 
habit of this plant does not greatly affect its structure as compared with 
other species, a circumstance which is held to point to the conclusion 
that Zsoezes is a genus which has long hovered about the limits of terrestrial 
and aquatic life. The statement here given is based upon the Memoir 
of Scott and Hill. 
The stele of the mature plant is not composed merely of the united 
leaf-traces, but is best interpreted as a cauline structure, comparable to 
that of the simpler monostelic Lycopods, but much shorter than is ‘usual 
in them. The crowded leaf-traces are inserted upon it, the stelar wood 
serving to join up the xylem of the leaf-traces, but it does not belong to 
one trace more than another, and in structure it differs from them. The 
1 Trans. Roy. Soc., Edin., vol. xvi., Part iii., No. 23. 
2 Ann. of Bot., vol. xiv., 1900, p. 413. 
Y 
