FLOWER OF GNETUM 
1G7 
clearly controlled by the meristem of the main axis, not by that of the 
secondary axis. Further it seems that the secondary axis has not attained 
to separate existence before at least the upper edge of the envelope is 
established. It is true that this upper edge has not been recognised 
earlier than the stage of Fig. 9 A, in which the primordium of the anthers 
is just visible as a slight convexity. But the whole structure is still a 
part of the general meristem, and it is at least an open question whether 
the secondary axis as such is yet in existence. It is of course possible that 
both the relatively late appearance of the axis and the basipetal development 
of the collar are consequences of the crowding of the products of the nodal 
meristem. At present however this can be no more than an assumption — 
an assumption which involves important conclusions regarding the morpho- 
logical value of the structures formed. 
If on the other hand the envelope is not composed of concrescent leaves 
and if it arises directly from the primary axis, as is not impossible on the 
evidence available, there are no longer any clear reasons for regarding the 
whole structure, consisting of the envelope, the antherophore, and the anthers 
as a male flower. All that can be recognised is a microsporophyll or a bundle 
of two or more monadelphous microsporophylls, surrounded by a “ basilar 
sheath.” In this case the axis of the primitive form of inflorescence 
suggested above (p. 162), bears stamens not male flowers. That such a 
view presents difficulties which are at the moment insurmountable is evident. 
Not the least of these is that of accounting for the basilar sheath if it be not 
a pair of concrescent leaves. It may however be pointed out that it is not 
the only structure in the genus of which a satisfactory explanation is not yet 
forthcoming. An example is seen in the fleshy “ basilar sheath ’’ or “ nodal 
cushion 1 ” which envelopes the young female flower and surrounds its lower 
half in the older stages in G. africanum. The origin of this structure is 
unknown ; vascular bundles enter its base. No one will attempt to derive 
this sheath from concrescent leaves ; and yet otherwise its nature is obscure. 
There is no reason to suppose that, in a morphological sense, there is 
anything in common between the envelope in Gnetum and the “ integu- 
mental ” covering of the microsporangium of Lepidostrobus Oldhamius 
(? Lepidocarpon Lurnaxi) 2 . Its existence may however be usefully re- 
membered until the nature of the Gnetum envelope is more clearly 
established. 
To formulate a new hypothesis of the morphology of the Gnetum flower 
and inflorescence is not one of the objects of this paper. It is rather an 
attempt to shew that the evidence at present available is not entiiely 
favourable to the only hypothesis which holds the field and is insufficient to 
1 Thoday, 1911, p. 1102, PI. lxxxvii, fig. 14; text-figs. 3, 13 A and C. 
2 Scott, 1901, p. 312 sqq., PI. 41, fig. 7. 
