_ I9t0] YOUNG—PODOCARPINEAE gt 
LAND (6), are reducing the epaulette type of Taxus, while no other 
family, except araucarians, has a stamen with more than two pollen 
sacs. If the two types have a common origin, only the Taxus stamen 
can be the primitive one; and in this case Phyllocladus has gone a 
long way in the podocarp line of development. A different origin, of 
course, would remove Phyllocladus still farther from Taxineae. 
Winged pollen is entirely absent among the Taxineae, but is 
characteristic of the Podocarpineae, being absent only in the 
Saxegothaea. The irregularity in size and number of the wings in 
Microcachrys has given rise to THompson’s (16, 1'7) theory that the 
two-winged condition has developed within the group and shows no 
relation to Abietineae. This gives us another podocarp line along 
which Phyllocladus has advanced. 
THOMPSON (20) lays a good deal of stress on the megaspore mem- 
brane as a primitive character. It is entirely eliminated in Taxineae, 
but characterizes all the members of the other family except Podo- 
carpus itself. As the ovulate structures of the type genus are by far 
the most specialized of the family, this is not surprising. Miss 
RoBERTSON thought the megaspore membrane and spongy tissue 
Were in some way correlated with the presence of winged pollen, but 
their occurrence in Saxegothaea with its wingless microspores breaks 
down this supposition. 
CONCLUSION 
When we consider the question of relationships, it is evident on 
the whole that: (1) Phyllocladus has primitive characters of the 
Taxineae which are being eliminated in the Podocarpineae; (2) it 
has primitive characters of the Podocarpineae which have been 
entirely eliminated in the Taxineae; (3) it has some advanced char- 
acters of Podocarpineae; (4) the taxad resemblances are on the whole 
more. superficial and variable, and the podocarp features more 
fundamental; (5) the resemblances to Podocarpineae are too strong 
to justify the retention of the intermediate family. 
We conclude, therefore, that Phyllocladus is a relatively primitive 
member of the Podocarpineae, which branched off from them a 
comparatively short time after their separation from Taxineae. 
hether the two families’form a distinct line from other conifers is 
not at all involved in this conclusion. 
