AND AFFINITIES OF STEPHANOSPERMUM. 387 
apparently discharged in quite another way by Hirase’s “ tentpole " in the case of Ginkgo, 
and by the central cone of nucellar tissue in Lagenostoma. R 
During the later stages of development the pollen has been maturing and has undergone 
considerable enlargement. The internal cells of the pollen underwent a subdivision into 
antherozoid mother-cells in Stephanospermum caryoides, perhaps in S. akenioides also. 
In the former the pollen appears to have undergone a process of dehiscence or fragmenta- 
tion about this time, whilst in S. akenioides it became perforated. Ultimately the pollen- 
fragments of Stephanospermum caryoides were ruptured by the mother-cell reticula. 
From these, antherozoids escaped, leaving traces of the reticula behind. Bodies which 
might conceivably be the antherozoids have been observed in S. earyoides only. "They 
occur associated with the pollen-fragments or on the floor of the pollen-chamber on the 
way to the archegonia. 
Adequate supplies of moisture would be drawn from the tracheal sheath. If the 
mucilage began to dry up at any time, its greater concentration would lead to further 
supplies of water being absorbed from the tracheides. In the absence of pollen-tubes, 
fertilisation would. depend on the pollen-chamber being nearly full of watery mucilage. 
From the researches of Webber * we know that the pollen-chamber is dry, or, rather, 
saturated with vapour, not full of water, during thé maturation of the antherozoids in 
recent Cycads. The same may have obtained in Stephanospermum during the earlier 
stages of development. As the antherozoids came to maturity, however, it would become 
necessary for the chamber to be filled, so that they might swim across the considerable 
interval. Meanwhile the pollen-chamber had been extending downward to the archegonia, 
and a quantity of damp mucilage would thus be always present. Finally, the tracheal 
sheath which overlies the prothallium itself underwent a breakdown, probably of the 
same kind, and disappeared from this region by fertilisation-time (cf. p. 369). This would 
explain the differences in fissibility in this region between younger and “ normal” stages 
in S. akenioides noted at p. 370. In the younger seeds the tracheides formed a compact 
sheath, and the splitting in post-mortem contraction of the maerospore occurred in the 
least resistant plane, which was below the tracheides (see p. 370 and PI. 42. fig. 13); in 
the older ones, which seem to have been preserved just prior to fertilisation, the tracheides 
below the chamber were already weakened, and so the split leading to the production of 
the supra-archegonial gap occurred in the tracheal layer itself (Pl. 42. figs. 8 & 12). 
Further, in S. caryoides, where the evidence points to ours being the actual fertilisation- 
stage, tracheides are wholly wanting from the floor of the pollen-chamber, though they 
begin just where the corner below the shoulder is turned. It may be conjectured they have 
disappeared from this region. By realising that the tracheides degenerated in this way 
as fertilisation-time approached two minor facts of observation are rendered intelligible : 
(1) the relative fissibility of different layers according to the age of the seed; (2) the 
almost total absence of tracheides from the supra-archegonial gap in S. caryoides. That 
the breakdown of the tracheides in this region should have impeded the outflow of 
water is hardly to be supposed. That the top of the macrospore should remain littered 
with tracheides and partly unrolled spirals is hardly consistent with zoidiogamy. Such 
* H. J. Webber, * Spermatogenesis and Fecundation of Zamia; 1901. 
3x2 
