AND AFFINITIES OF STEPHANOSPERMUM. 381 
this view are possible. First, that a fragment of the matrix has splintered away in the 
grinding of the preparation. This interpretation is untenable, as the matrix is actually 
present at c. "The other alternative is that the excrescence is an accidental blister that 
originated prior to mineralisation *. As there is no means of determining from this 
isolated example whether the outgrowth was a natural or an accidental production, any 
conclusion as to its nature must be based on other considerations. Before this is possible 
the remaining stages have to be considered. Whatever may have been its nature, this 
tube was not very extensive. If we assume that the tube curved out of the plane of 
section, we should expect to find it in one of the two adjacent sections, if the appropriate 
regions be examined. No trace of it, however, is to be found. The tube consequently 
was of very limited extent. In the preparation it reaches to a distance of 22 » from the 
pollen-grain. ‘Traces of the triradiate slit are also present (Pl. 44. fig. 40, tr.s.). 
(b) Isolated internal Cells.—The pollen evidently underwent. some process of partial 
dissolution or rupture, for the isolated internal cells, agreeing approximately in dimen- 
sions with o and 5 of fig. 40, are common in the pollen-chamber. Those of importance 
to this account are shown at pl, ni ni (text-fig. 2). 
p.'—This is represented in fig. 41, and corresponds in dimensions (40 45x30 4) to the 
larger internal cell (a, fig. 40). It is still enclosed by the reticulated exospore of the 
convex part of the pollen-grain, and at w appears to show a small portion of the radiating 
sculpturing belonging to the flatter portion of the original pollen-grain. Within, a four- 
celled complex has been developed. "These secondary cells show no contents. They 
have an average diameter of 15 4. The containing (sculptured) membrane of the 
complex is incomplete above, but the walls of the secondary cells appear to be complete. 
p#—This pollen-fragment consists of the pair of internal cells shown in fig. 42, p.?. The 
cells are still joined together, but their original relations in the pollen-grain seem to have 
undergone a slight alteration by the outward rotation of 5 from « (fig. 42) in the 
direction of the adjacent pollen-grain p.? ‘The close association of o and b with this 
pollen-grain seems to be purely aecidental—p.? is intact, o and b are the remains of 
another pollen-grain. The two joined cells of p.* are laid open and present their 
concavities. Each contains the remains of a secondary cell-complex: that in 4, relatively 
well-preserved (5, fig. 42), and somewhat displaced in relation to its containing 
membrane; that in o only recognisable as a crushed cell-group at c (fig. 42). PI. 42. 
fig. 20 represents the cell-complex lying in the containing cell b, taken direct from 
the specimen. 
The dimensions are as follows :—a, 37 uX 25 n; b, 344X254. The larger cells of 
the reticulum contained in b have a diameter of about 15 n. | 
p.5— By itself, little importance would attach to this fragment. It consists ofa broken 
containing-membrane with the remains of a possible secondary tissue lying close 
beside it; it is not figured. | 
p? essentially resembles p.!, but is less well placed for observation. 
From the foregoing exposition of the facts concerning the pollen-grains and pollen- 
* A slight blistering has been observed in occasional instances in Cordaitean pollen-grains lying freely in the 
matrix. 
