2994 PROF. F. W. OLIVER ON THE STRUCTURE 
be that there was a certain local specialisation in connection with the discharge of the 
antherozoids. From a study of the structure of the pollen of Stephanospermum 
akenioides, Renault conjectures that a state of affairs may exist corresponding to 
_Eitheotesta. I have endeavoured to show (p.378) why such an interpretation of 
S. akenioides seems inadmissible. If the little tubular processes occasionally found 
projeeting from the periphery of these grains (as in the large grain, fig. 15) be indeed 
naturally formed, the inference I would draw would be that this pollen was of a more 
generalized type, and that any of its cells could produce these tubes. In other words, 
that short multiple pollen-tubes were formed through which the antherozoids were 
discharged. As compared with this type, Ætheotesta and Codonospermum, each in its 
own way, shows a slight advance without departing from the Stephanospermum-type of 
pollen. Coming to Stephanospermum caryoides, reduction and specialisation are more 
complete. A considerable part of the pollen-grain remains undivided, as in recent 
Gymnosperms. This outer cell, whilst it may have possessed nutritive functions in 
connection with the maturing of the antherozoids, was doubtless concerned in promoting 
their discharge into the pollen-chamber. Whether this result was achieved through the 
agency of an incipient tube, as suggested by our preparation, or whether the internal 
cells were liberated through the deliquescence or dehiscence of the outer cell, must 
remain an open question till more evidence accrues. 
The internal complexes of cells in the pollen of Stephanospermum akenioides, of 
Cordaites, and of Stephanospermum caryoides may thus represent different stages in the 
reduction of an archaic male gametophyte of a complexity not known in any Fern. 
It is true that the retieulations of the supposed “microspores” in Botryopteris and 
Zygopteris have been interpreted as an internal tissue *, thus offering an easy transition 
to the compound pollen of the early Gymnosperms. To this view of the reticulations 
there seem to be grave objections, and, after careful study of specimens, it is difficult to 
avoid sharing the doubt which has been expressed t whether these ** microspores oi 
justify the interpretation to which allusion has been made. 
The main obstacle to definite and certain conclusions as to the behaviour of the pollen 
in these seeds is, of course, the fact that the seeds fell from the tree at a stage in 
development prior to fertilisation. The conditions are somewhat analogous to those that 
would obtain in the case of Pinus sylvestris were only first-year cones available for 
investigation. "The mode of transport of the sexual cells is not yet fully declared. Nor, 
indeed, in Stephanospermum akenioides, do we know with certainty which are the 
sexual cells. The preparations that have come under observation, however, seem to 
indicate decisively that the “central body " is not of this nature (cf. p. 877). Of the 
twenty or more peripheral cells which compose the grain, it is impossible to select any 
as differing from the rest. They are all alike. Occasionally, short tubes are found— 
spout-like outgrowths from some of these cells. But their occurrence is too occasional as 
yet for acceptance as a normal stage in development. That polien-tubes were produced 
penetrating the nucellar tissues as in recent Cyeads and Conifers appears highly 
* < Flore fossile d'Autun et d’Epinac,’ pp. 44, 54, 57. 
t R. Zeller, * Éléments des Paléobotanique,’ 1900, p. 74. 
