1914] BURLINGAME—ARAUCARIA BRASILIENSIS 505 
araucarians is “proto-angiospermic.” It seems to me, on the con- 
trary, that we have been altogether too ready to accept the type of 
ovule which has a specialized pollen chamber securely hidden away 
at the base of the scale, and which can be reached by pollen only 
by means of special devices, as a primitive type. It is scarcely 
credible that the first step in the evolution of the ovule and seed 
should have been so complex. If, however, the gymnosperms 
possessing this complex type evolved a seed before a cone, as in 
fact there is good reason to think they did, then this might have 
been at Jeast one of the earlier successful types. If, on the contrary, 
the cone was evolved before the seed, or simultaneously with it, as it 
may very well have been in an apparently simple cone, what would 
be more natural than that the pollen should lodge in any convenient 
place among the scales (sporophylls perhaps) of the cone? Such 
an ovule would have a much better chance of survival in such a cone 
than if exposed on the lobe of a fernlike leaf such as those possessed 
by the known Cycadofilicales. I am aware that I am thus attempt- 
ing to introduce an apparently ancestorless conifer, but fail to see 
that a search for fitting ancestors is likely to be more difficult than 
deriving it from unsuitable ones. 
Imbedded archegonia have been reported (6) for Araucaria, 
and a comparison made with Sequoia in which somewhat similar 
conditions are said to be present (4). My own observations do 
not bear out the presence of such imbedded archegonia in Arau- 
caria. They are superficial in origin and become overgrown by the 
neighboring cells. Srvnotr has recently reported (7) practically 
identical conditions in the podocarps. The necks are not specially 
noteworthy, though they show a rather closer resemblance to those 
reported for Podocarpus (7) than to those of most other conifers. 
The failure to find a ventral canal nucleus is somewhat sur- 
prising in so large a gametophyte, not having advanced beyond 
the evolutionary stage in other respects that has been attained by 
conifers generally. It seems more probable that it will yet be 
found. Ventral nuclei have not been found as yet in one species 
of Torreya (2). 
The gametophyte, therefore, appears to be neither highly 
specialized nor exceptionally primitive in its structure. Its large 
