382 PROF. F. W. OLIVER ON THE STRUCTURE 
fragments occurring in this pollen-chamber, it appears evident (1) that the pollen-grains, 
originally containing two internal cells and an extensive undivided residue (as shown in 
fig. 40, and as may be gathered also from scattered grains in the matrix), underwent a 
process of dehiscence by which the internal cells were liberated; (2) that the two 
internal cells concurrently produced secondary cell-complexes which may be regarded 
provisionally as antherozoid mother-cells. So far as the stages that have been found 
are concerned, the manner of freeing of the internal cells is obscure. "That stage is 
wanting. 
We may now return to p., the pollen-grain represented in Pl. 44. fig. 40 with tube-like 
outgrowth, and consider whether it helps to bridge the lacuna in our series. Assuming 
for the moment that it is a naturally produced tube, it is easy to realise (from the point 
of view of the preservation of the pollen-fragments) that it may have constituted a weak 
spot from which decay may have started on the death of the seed, a decay which 
ultimately involved the whole peripheral part of the pollen-grain. Or, the changes in the 
outer cell of the pollen-grain which led to the production of a tube may have involved 
the whole of the peripheral part of the wall of the grain in decomposition. "This is one 
hypothesis. On the other hand, it must be borne in mind that a solution of this part 
of the wall might have been achieved without the production of any tube. Finally, 
there is a third possibility. The triradiate mark or crack on the pollen-grain p.*, 
represented also on p. (fig. 40, ¢.7.s.), and occasionally present on these grains in the 
matrix, may have offered the means of dehiscence, though the possibility seems a remote 
one. It is true that fern-spores with triradiate ridges use these lines for dehiscence at 
germination *, but there is no reason for supposing this method to have been perpetuated 
in seed-plants T. 
But however this may be, it does not touch the present case, for the question is how 
the pollen-grains come to lose their wing-like expansions. 
The detached internal cells 
still bear their covering of exospore on one side. 
The decision arrived at depends on 
the view taken of the tube-like outgrowth in fig. 40. 1f this is a genuine tube (however 
rudimentary), then the structure of the grain would be intelligible. The outer cell 
is concerned in the production of a short tube through which the antherozoids, which 
may be supposed to have arisen in the secondary complexes (shown in p.?, fig. 42), were 
discharged. On this view the grain is differentiated into a generative portion consisting 
of the two internal cells, and a directive portion which developed a tube and perhaps 
started the autherozoids in the right direction. 
Having traced events so far, we may pass on to the more conjectural question of the 
antherozoids themselves. In what is doubtless the fossilised mucilage of certain regions 
of this pollen-chamber, bodies are preserved which are susceptible of interpretation in 
this sense. The bodies in question are small ovoid vesicles having as average dimensions 
* See Sadebeck, in Schenk’s * Handbuch d. Botanik,’ vol. i. p. 154. 
T See Coulter & Chamberlain, * Morphology of Spermaphytes,’ p.91, fig. 69, c. The pollen-grains of Pinus 
Laricio are shown in the tetrad condition. From the position of the sac-like blisters of the extine, it would appear 
that the pollen-tube will arise on the side of the pollen-grain remote from the contact-angles of the members of the 
tetrad, 
