014 Pearson. — On the Micro sporangium and Microspore of 
been seen, but the succession of these' is strictly acropetal. It appears 
that the basipetal development of lateral organs is very frequently, 
if not always, associated with arrested apical growth. 1 In the Gnetum 
male inflorescence the same association is found, for although the rings 
follow one another in acropetal succession, the growth of the axis is 
definite. The number of rings in each spike is almost constant within 
narrow limits, 14-17 in G. scandens , 8-10 in G . africanum , 7-9 in G. 
Biichholzianum ; exceptionally the axis becomes arrested early, and the 
rings produced are fewer — two in one example of G. Buchholziamnn. But 
whether the number of floral rings is normal or less, the axis at length 
terminates in a segment which differs from those beneath it. This last 
segment bears no flower-primordia, and the bundles which should supply the 
flowers are not represented even by procambial strands (Text-figs. 5 and 6). 
Such a segment as this puts an end to the normal apical development of 
the axis as effectively as if it were a flower. And in G. Brunonianum ‘ der 
Scheitel der Inflorescenzaxe gipfelt auch hier haufig, wie in den rein 
weiblichen Inflorescenzen, in einer einzigen weiblichen Bliithe’. 2 The axis 
then is of limited growth. This being the case, and since the inflorescence 
may consist of as few as two flowering nodes and a terminal segment (which 
in some cases is an ovule with appendages), it is conceivable that the 
ancestral type of inflorescence was an axis bearing a single lateral ring 
of male flowers, and a terminal female flower. In any case it seems 
at present impossible to refer the male spike of Gnetum and either the 
spike or flower of W elwitschia to a common type that is not very remote 
from both. 
The structural characters of the male spike in which G. scandens and 
other Indo-Malayan species differ from the African species may be sum- 
marized thus : 
Indo-Malayan species . 3 
Internodes suppressed. 
A ring of imperfect female flowers 
at the top of every node except one 
or two nearest the apex. 
No mucilage canal in the cupule. 
Vascular complex supplying the 
flowers derived entirely from a single 
series of strands ascending from the 
leaf- traces. 
1 Goebel (’00, ’05.) 
3 Hitherto described. 
African species. 
Internodes longer than nodes. 
In nearly all cases female flowers 
entirely absent. 4 
Large mucilage canal continuous 
round the cupule between upper 
epidermis and vascular bundle. 
Vascular complex supplying the 
flowers derived from a double set of 
strands, one as in G. scandens , the 
other descending from the top of the 
node or from the internode above. 
2 Strasburger (72), p. 158. 
4 A single exception seen in G. africanum. 
