1915] BURLINGAME—ARAUCARIA BRASILIENSIS 25 
undertaken to follow up the pollen grains by protruding itself 
through the micropyle. A theory beset with such manifest 
difficulties can be accepted only if no more probable one can be 
found. 
ARAUCARINEAE.—The araucarians have been thought by some 
‘authors to be derived from the lycopods (20, 4a, 4b, 22a, 22b), and 
by many others to be derived directly from the Cordaitales. We 
have seen that the pollen tube structures do not lend any support 
to the derivation of the araucarians from the cordaiteans through 
the other families of conifers. Whether the pollen tube structures 
could be derived directly is a question that we can best attack after 
considering the bearing of these structures on the theory of a lycopod 
origin. 
Miadesmia (19) and Lepidocarpon pee two seed-bearing lyco- 
pods, seem to me to present the most suggestive analogies of 
the manner in which such a seed as that of Araucaria may have 
been evolved. I do not mean to imply that these analogies are 
sufficient or adequate evidence for deriving the araucarians alone 
or conifers as a whole from the lycopods, but merely that the 
araucarian pollination apparatus could be easily derived from such 
seeds as these plants possessed, whether they belonged to lycopods, 
cordaiteans, or what not. 
The seeds of Lepidocarpon were formed in cones and not exposed 
as in the Cycadofilicales. The same is true of Miadesmia. I am 
inclined to attribute to this fact considerable importance. Seeds 
that originated on a naked foliar structure would necessarily have 
to be pollinated on: the ovule to have any chance of success at all 
under any ordinary conditions of plant growth. Otherwise, the 
ciliated sperms would have encountered almost insuperable diffi- 
culties in reaching the archegonia and would have been limited to 
wet weather. It seems from such considerations that Cycadofili- 
cales and their allies have been from the first pollinated on the 
nucellus, but no such compulsion rests on plants which had acquired 
the cone habit first. The natural, easy, and probable place for the 
lodgment of the earliest pollen would be between the scales any- 
where. There would be far less danger of the pollen blowing 
away before it could become effective because of its protected 
