340 Johnstone: The Ancestors of the Angtosperms. 
spermic flower, and that being so, a short summary of their 
structure may be of interest. 
The habit of growth and stem-form agree very closely with 
those of the typical modern cycads; but the first great difference 
is apparent in the mode of attachment of the fructification, which 
consists of a strobiloid axis borne on the main stem and wedged 
in amongst the persistent leaf-bases. This short axis terminates 
in a convex receptacle bearing /a@) at its lowest level a series of 
spirally arranged bracts, (4) above these a whorl of frond-like 
sporophylls, fused proximally, bipinnate, and having numerous 
pollen sacs of a synangial type, (c) a central conical region, 
arranged on which are numerous stalked ovules interspersed 
*with elongated scales partially united at their distal ends to 
form a kind of ovary wall, but still allowing the micropyles of 
the ovules to project a little beyond them. 
The marked general agreement in the arrangement of the 
various parts of the flower, the perianth-like lower members, 
the hypogynous stamens, and the central ovary are all most 
suggestive of the angiospermous flower. This very close 
approximation was at once recognised by Dr. Wieland, and 
has met with very general acceptance. Minute examination 
of the structure strengthens the case by showing that not only 
were the seeds almost exalbuminous, but the embryo was in 
all respects that of a highly organised dicotyledon. 
The Bennettites of Mr. Carruthers and the earliest specimens 
examined by Dr. Wieland showed only the female organs ; this 
probably simply meant that the stamens had withered, and was 
quite in accord with the fact that the seeds were in a fully ripe 
condition as shown by the stage of development of the embryo. 
The basal portion left after the stamens had dropped probably 
represented the ‘disc’ of W7z/lzamsonia gigas. 
Such a striking assemblage of widely contrasting characters 
as we have in these surprising fossils has of course given rise 
to much fresh speculation. The fern-like sporophylls with 
their synangia of a Marattiaceous type indicate a connection 
with the oldest types of ferns. Side by side with these we find 
the highly evolved seeds and the ovary wall suggestive of the 
angiosperms, though the plant is yet essentially a Gymnosperm, 
inasmuch as the pollen is collected not by a modified carpel, 
but by the ovule itself. The spiral of bracts, which possibly 
may foreshadow the modern perianth, and the dicotyledonous 
character of the seed complete a most astonishing combination 
of great type features. 
Naturalist 
