Morpliology of the Female Flower of Gnettim. 
71 
normal micrppylar tube ; it bears a curious leaf -like lobe {lo). Envelope II 
is similar in structure to I except that it contains very slender vascu- 
lar traces. Nos. Ill and IV are much thicker, and in structure more 
differentiated than I and II ; they are well supplied with small vascular 
bundles. 
The authors, who are on the whole disinclined to attach importance to 
abnormal structures as evidence in morphological inquiry,* consider that 
this particular abnormality " offre justement toutes les apparences d'un 
retour ancestral." t 
Their conclusions are j-: : 
(1) The whole structure (nucellus + 4 envelopes + 3 axillary groups) 
represents a single node of the inflorescence axis, bearing a cupule and an 
axillary ovule. § 
(2) The cupule (envelope IV) is applied against the surface of the ovule § 
and has become a supplementary envelope. 
(3) The second envelope of the ovule § subtends an axillary group oi 
normal male flowers. 
(4) Probably the normal female flower of Gnetum (i. e. ovule with 
3 envelopes) really repriesents a simple, perhaps compound, axis provided 
with 2 nodal cupules (envelopes II and III). This axis is terminated by a 
plurilocular " ovary (envelope I) in the interior of which is found, as in 
Welwitschia, a single ovule basilar, orthotropous, erect, and naked. 
(5) This pseudo-ovule § of Grnetum seems therefore to be comparable 
with the male flower of Welwitschia, the ovary (envelope I) and its nucellus 
being almost identical in the two genera. 
The two external envelopes (II and III) of Gnetum correspond to the 
two lower whorls (" perianth ") of Welwitschia. 
The staminal whorl of Welwitschia (winged envelope of the female 
flower) has completely disappeared from Gnetum. 
(6) In spite of its position immediately outside the ovary (envelope I), 
it does not seem that the group of male flowers (Text-fig., M) can be 
considered as a reversion to an ancestral organisation from which the 
androecium of Welwitschia is derived. 
These conclusions demand careful consideration. We may note in the 
first place that the authors regard the female flower ("ovule" of §1) as 
being in the axil of the fourth envelope, the cupule. This seems to com- 
plicate the structure quite unnecessarily, for there is no direct evidence in 
favour of its axillary character. A normal female flower very frequently 
* Lignier and Tison, 1912, p. 165, footnote. ■ ., 
t Lignier and Tison, 1913, p. 70. 
X Log. cit., pp. 71, 72. 
§ This "ovule" is the " psendo-ovule " of §5; i.e. the female flower in the 
ordinary sense. In §4 the tonn " ovule " is applied to the naked nucellus. 
