597 
Life-history of Tetraclinis articulata . 
Appearances, difficult to interpret, seen in the megaspore membrane at 
this time, led to a very careful examination of that structure. According to 
Thomson (54) this membrane consists, in the CupressoTdeae, of an endospore 
and exospore, represented in his figures as of approximately equal thickness. 
The endospore is more or less homogeneous, while the exospore consists of 
numerous slender radiating rods. He also mentions that material derived 
from nucellar tissue may increase the apparent thickness of the exospore. 
As to the last point, the deeply staining deposit on the outside of the 
exospore does, in many cases, obscure its structure, but here and there parts 
may be found which are free from it. In order to see the structure clearly 
it is also necessary not only to select a part which shows no trace of 
‘ dragging ’ (as Thomson mentions), but also one which is cut almost exactly 
perpendicular to its own plane, so that a slight change of focus produces no 
section through the cotyledons 
of a tri-cotyledonous embryo. 
x 35- 
Text-fig. 8. — A very 
small part of the megaspore 
membrane of Tetraclinis in 
longitudinal section. The 
outer side to the right, 
x circa 3000. 
Text-fig. 7. — Sketch of an embryo 
with five cotyledons, x 16. 
lateral displacement of the image. Having satisfied these conditions, it is 
possible to distinguish in Tetraclinis a very thin homogeneous endospore, 
which I should estimate at only one-sixth of the total thickness of the 
membrane, and an exospore which consists of rods thickened in a dumb- 
bell-like manner at the apex, so that with a lens which does not resolve in 
an entirely satisfactory manner, the outer layer of the exospore appears 
continuous, while the inner and thicker layer appears to contain numerous 
radiating vacuoles. Text-fig. 8 represents a sketch of the structure as seen 
with a high magnification. It will be seen to be similar to that figured by 
Chamberlain (7 and 8) for Dioon and Ceratozamia , except that the terminal 
swelling of the rods is more abrupt. It would be interesting to know 
whether a re-examination of other Gymnosperms would show that a similar 
structure is more widely prevalent. It may be mentioned that a figure of 
the megaspore membrane of Juniperus given by Noren (37) shows a structure 
quite similar to that seen in Tetraclinis when the resolving power of the lens 
