Lycopodiales. 
157 
Lepidostrobus and Sigillariostrobus were derived from a type with 
such approximately spherical sporangia, either by independent 
radial elongation of their sporongia, or by such elongation in the 
sporangium of their common ancestor, which was carried further 
in the evolution of Lepidostrobus than in that Sigillarisotrobus. If 
Spencerites is a Lepidodendraceous fructification it may well be the 
most primitive type of all. The reasons for this are : 
(a.) The distal attachment of the sporangia appears to be a 
primitive character (13), (17); it is unknown among recent 
Lycopods, but was characteristic of the Palaeozoic Sphenophyllales, 
the phylum to which the Lycopods, though very much isolated, are 
probably closest; and if a character occurs among the primitive 
forms of allied phyla, its presence may be expected in the primitive 
members of a given phylum. It is, however, quite possible that the 
ventral sporangiferous projection so reminiscent of the Spheno- 
phyllaceous sporangiophore is not homologous with that organ, for 
it contains no vascular tissue. In this connection it should be noted 
that Miss Sykes has recorded the presence of lignified cells, not, 
however, in contact with those of the sporophyll-trace, in the 
sporangial stalk in other species of the phylum (23). She even 
contends that such cells, the existence of which is denied by Dr. 
Scott (14) and by Miss Berridge (2), occur in the latter’s pre¬ 
parations of Spencerites. This argument, founded on the primi¬ 
tiveness of the distal insertion of the sporangium, and therefore 
on a theoretical assumption not universally admitted, is certainly a 
weak one; but it is to a certain extent reinforced by the fact that 
the sporangium of Sigillariostrobus Crepiui is also attached to the 
sporophyll by its distal end; the appearance of such a character in 
another genus of the order favours the view that the common 
ancestor of the genera of the order may have possessed this 
feature. This cone also approaches Spencerites and Bothrodendrou 
in the fact that it is not radially elongated. Unfortunately its 
attribution to Sigillana, is not considered to be perfectly established 
( 12 ). 
(b). The axis of Spencerites was in some cases destitute of a 
pith, and a solid stele is generally admitted to be more primitive 
than a medullated one. In cases where a pith is present it appears 
to be a small one (14), whereas the pith of Lepidostrobus 
seems to have been well developed, except at the apex of the cone 
(3), (11). This difference between Lepidostrobus and Spencerites is 
presumably partly due to the difference in the size of the two genera. 
