Evolution of the A ngio sperms. 
491 
attempt to restore in imagination the essential features of the fructification 
of the immediate ancestors of the group. There appear to us to be three 
possible alternatives from which to choose. These may be stated as 
follows : — 
(1) The Gnetales were in past times, especially during the Tertiary 
period, a great and complex group, forming one of the dominant factors in 
the vegetation of the period. On this supposition the three surviving genera 
are probably very aberrant members. 
(2) The Gnetales are quite a new group, with little or no past history. 
(3) The Gnetales were in existence in Tertiary times, but have always 
been a small and little varied group, which has hitherto formed only one of 
the subsidiary elements in the flora. 
These three possibilities may be represented graphically by life-lines 
as in Text-fig. 1. 
Since the plant record of the Tertiary rocks does not furnish any real 
evidence in favour of one alternative 
1 *■* * 
rather than another, we must depend 
entirely on morphological evidence, 
and more especially on that afforded by 
the fructifications. The most plausible 
interpretation of the fructifications of 
these three genera, as we hope to show, 
is that they are reduced and highly 
modified forms of what we have called 1 
‘ pro-anthostrobili ’ (see p. 497). If this 
be the case, such evidence as there is 
strongly supports the first alternative; 
it is entirely incompatible with the 
second, and can only be regarded as 
in harmony with the third in a very 
slight degree. We assume, of course, 
one fundamental hypothesis, namely, that the Gnetales are monophyletic, 
a conclusion which, so far as we are aware, is not seriously challenged. 
Further, we believe it is only by first arriving at some clear idea of the 
status, so to speak, of the living genera, that we can begin the attack 
on the main problem of their relationships to the Angiosperms. The 
evidence that these two groups are distinctly related appears to us to 
be overwhelming and most convincing. Yet this by no means commits us 
to the view that one group gave rise directly to the other, but rather that 
both the Angiosperms and Gnetales sprang from a common ancestor. This 
ancestor, as yet unknown to us in the fossil state, we have already restored 
provisionally under the name of the Hemiangiosperm. 2 
1 Arber and Parkin (’ 07 ), p. 37. 2 Ibid., p. 62, &c. 
Ll 
Fig. 1. Diagrammatic representation of 
the three alternative theories with regard to 
the past history of the Gnetales. 
