65 
Inter-relationships of the Phyla. 
Equisetum. The radial symmetry of these latter and the fact that 
not only do they bear more numerous sporangia than most Cala- 
marian sporangiophores, but that in certain Mesozoic Equisetales, 
closely resembling the recent Equisetum, there were sometimes over 
twenty sporangia on a sporangiophore (16), make it difficult to 
regard the sporangiophores of Equisetum, either as composed of 
two coalescent stalks bearing few sporangia, or as the ultimate 
sporangiferous pinnules of a fern-frond. Professor Lignier seems 
to feel this difficulty, for he says that the cones of Archceocalamites 
and Equisetum are both of a kind for which it would be rash to give 
an explanation ; he adds that the appendages of the cones do not 
seem to show a division into fertile and sterile lobes of the second 
order, but appear to have been transformed as wholes into 
sporangiophores, or that such a cone as that of Archceocalamites 
may have been derived from a form with alternating fertile and 
sterile whorls by the abortion of the latter. That this abortion 
should be so complete as to leave no trace of the sterile whorls, even 
in so old a type as Archceocalamites, seems very unlikely. The 
hypothesis of the primitiveness of the Sphenophyllum fertile- type 
of cone seems to offer a more natural interpretation ; if the dorsal 
and ventral lobes of the sporophyll of such a type became free from 
one another and came to be arranged in successive whorls on the 
axis (as appears to have been the case in Calamostachys, where* 
however, the dorsal lobes had been sterilized), we should get super¬ 
posed whorls of sporangiophores. Further arguments in favour of 
such a view have been given in the first and second articles of this 
series. 
The affinity between the Sphenophyllales and Equisetales 
seems clear. Anatomically the presence of some centripetal xylem 
in Protocalamites pettycurensis is important, for it suggests that the 
older Equisetales, of whom we have no record, may have possessed 
an exarch protostele (like that of the Sphenophyllales) in which, in 
the course of evolution, the centripetal wood was, as it often seems 
to be, replaced by pith, probably owing to the acquisition of centri¬ 
fugal wood. In both phyla the leaves seem primitively to have 
been large, dichotomous and arranged in superposed whorls (34). 
This superposition, constant so far as we know in the Spheno¬ 
phyllales, was characteristic also of Archceocalamites, of the older 
Calamites (21), and of some Mesozoic genera (36), but has been 
lost by most species of Calamites, by Equisetites and Equisetum . 
The strobili consist of whorls of sporophylls and there seems some 
