A ctinostrobus pyramidatis , Miq. ^ 327 
organized, each of which, before reaching very far towards the centre of the 
sac, cuts off a small cell at its base. The form of these small cells is clearly 
seen in Fig. 9. The alveoli are otherwise quite unsegmented, the apparent 
transverse walls appearing in the figure in some places, especially on the 
right, being due to other alveoli passing into the plane of the section. On 
reaching the centre of the prothallus, each alveolus is terminated by a wall, 
stages shortly after that represented in Fig. 9 showing, occasionally, the two 
end walls of opposite alveoli in contact. Apparently, however, they soon 
fuse together, as only a single wall can be distinguished later. 
The apex of the prothallus is narrow and pointed, not truncate or 
rounded as in most Conifers. 
The long inner cells of the alveoli in the upper half of the prothallus, 
except in the narrow part near the apex, are those from which, after cutting 
off some small cells at one or both ends, the archegonium initials are 
formed. Thus the archegonium initials are the largest cells in the pro- 
thallus at the time when they can first be distinguished, not amongst the 
smallest, as is often the case. 
It appears as though every alveolus in the upper half of the prothallus, 
with the exception of the narrow apex, gives rise in this way to a cell 
which is of the nature of an archegonium initial, and in cases where no 
pollen-tubes have grown down the prothallus, these initials remain for some 
time unsegmented and are rather conspicuous, but where one or more 
normal pollen-tubes have grown down, only the initials in contact with the 
pollen-tube become functional archegonia, all the rest becoming transversely 
divided into several cells. This is indicated in the transverse section of 
a prothallus shown in Fig. 10. Since, however, the pollen-tubes are already 
in position when cell formation takes place, it is likely that some modifica- 
tion of the method just described must occur in the case of functional 
archegonium initials. Exactly what this modification may be has not 
been determined, but enough has been seen to make it clear that the initials 
always arise in a similar manner to that just described. 
In a former paper ( 19 ) it was stated that in Widdringtonia the distri- 
bution of archegonia is determined largely by the position of the pollen-tube. 
It was subsequently shown, however, in Callitris (21) that archegonium 
initials might appear in an exactly similar position, in the absence of 
a pollen-tube in the prothallus. 
The present study makes the position more clear, and as the point is 
of some importance, the degree of correlation between pollen-tube, arche- 
gonia, and archegonium initials may be briefly recapitulated as follows : 
The formation of archegonium initials in their normal position, laterally 
and deep-seated in the prothallus, is not determined by the position of the 
pollen-tube, but this position does determine which of the very numerous 
archegonium initials become functional archegonia. 
