NOTE ON INFLORESCENCE AND FLOWER OF GNETUM 153 
other hand, the differences which are ascertained are very considerable. 
Apart from the interpretation of the inflorescence and flower which is 
perhaps open to question, a comparison of the gametophyte stages does not 
at first sight suggest any close degree of affinity. It is true that some 
authors consider that “too much weight has been attached to ‘embryological’ 
dissimilarities 1 ,” but the weight which they can bear can hardly be estimated 
until they are better understood. Great as are the apparent differences 
between them it is probable that the type found in Welwitschia is not 
distantly related to that of Gnetum, and possible that both may be derived 
in a simple and direct manner from that of Ephedra. On the other hand it 
is not impossible that these gametophytic structures indicate that Gnetum 
and Welwitschia have diverged widely from the line represented by Ephedra. 
In either case the reproductive organs of the sporophyte of the three genera 
may be directly and not distantly related. But until clear evidence upon these 
questions becomes available, the conclusion that the alliance between the 
three genera is close enough to warrant the derivation of the floral structures 
of Gnetum and Ephedra from a type still preserved in Welwitschia is hardly 
more than an assumption. 
Gnetum itself presents special difficulties. In Ephedra and Welwitschia 
the bracts and flowers are strictly decussate in arrangement and, in 
Welwitschia at least, the products of successive nodes appear in a well 
marked acropetal order. In Gnetum the flowers stand in densely crowded 
whorls or, occasionally, they and their subtending bracts are disposed in 
a long and continuous spiral (Fig. 1) ; otherwise the cupules are clearly the 
result of the concrescence of decussate pairs of bracts. Further peculiarities 
are seen in the very common reduction of the female flowers which, in most 
species, stand at the top of each node of the male inflorescence; and the 
basipetal development of a succession of male flowers. Also the products of 
successive nodes do not arise (in G. Gnemon and probably other species) in 
acropetal order (Fig. 2 A). 
These considerations make it the more necessary that the importance of 
certain characters of the Gnetum flower and inflorescence which do not appeal 
to support the current views of their origin, should not be underrated. 
The material available for examination represents G. Gnemon, G. a ft ica- 
num, and G. scandens fairly abundantly, and includes a few specimens of 
undetermined species from Singapore and Java and male infloiescences of 
G. Buchholzianum. In addition to the acknowledgements already made 2 , 
I am indebted for supplies to the Acting Director of the Buitenzorg Gardens; 
Major Gage, I.M.S., F.L.S., Superintendent of the Royal Botanic Gardens, 
Calcutta ; Mr I. H. Burkill, F.L.S., Director of the Royal Botanic Gardens, 
Singapore; Mr G. H. Cave, Curator of the Lloyd Botanic Gardens, Darjeeling, 
1 Arber and Parkin, l.c. p. 506. 2 Parson, 1912, 1915. 
