510 Arber and Parkin. — Studies on the 
of branching in Gnetum and Ephedra. The dicotyledonous embryo, the 
general morphology of the seedlings and leaf-trace system, and the presence 
of elements in the xylem, which more or less resemble true vessels, afford 
additional evidence. 
The question is, then, not as to the existence of an affinity, but the 
degree of that affinity. The possibility of including the Gnetales within 
the Angiosperms is, as is now generally agreed, ruled out of court by the 
increasing importance, which has continued to be attached, since the days 
of Robert Brown, to the distinction between naked and enclosed ovules. 
Nor could the present day Gnetales have been the ancestors of the 
Angiosperms, for not only are they contemporaneous with them in point 
of time, but their fructification remains a pro-anthostrobilus, as opposed 
to the eu-anthostrobilus or flower of the Angiosperms. 
On the other hand, all the evidence points to the conclusion that these 
two groups have sprung from a common ancestor, the fructification of which 
was also a pro-anthostrobilus, and that they have progressed in a great 
measure parallel to one another. This hypothetical ancestor we have 
already attempted to restore provisionally in our earlier paper, under the 
name of the Hemiangiosperm . 1 It possessed a typical pro-anthostrobilus, 
which was amphisporangiate, and also a perianth. In the male fructification 
of Welwitsckia, all the organs of this strobilus are still present, though in 
a much reduced form. The advance towards the monosporangiate con- 
dition has, however, been great, for the megasporangium, though present, 
has ceased to function. The female of the same plant, and both the male 
and female fructifications of Gnetum and Ephedra , are easily derived by the 
further reduction of such a strobilus. 
Though the male strobilus of Welwitschia may be regarded as having 
retained a larger number of primitive features than the other fructifications, 
it is no doubt itself extremely reduced, and thus differs very widely from 
the typical pro-anthostrobilus of the Hemiangiosperm. While we can 
still trace the relationship, so far as the essential characters of the antho- 
strobilus are concerned, it is not easy to restore the exact stages, which 
represent the path along which the reduction has been carried. Yet, as we 
have pointed out, the origin of the male synangia of the Gnetales does 
not present any great difficulty, except in details. The absence of the 
megasporophyll can be accounted for by the fact that, where great re- 
duction has been the rule, it is not surprising that the total suppression of 
an unimportant organ should have taken place, since the method of pollen- 
collection in both the Hemiangiosperms and the Gnetales was, and is, 
independent of it. 
In other respects the scanty evidence to be drawn from a few, such 
highly reduced strobili compels us to leave the picture unfinished, in the 
1 Arber and Parkin (’ 07 ), p. 62, and Fig. 4 on p. 63. 
