a. BOTANICAL GAZETTE [JANUARY 
may be left out of present consideration because there is little evi- 
dence that it lies in the line of direct descent to the conifers. Since 
the method employed in the cycads is an entirely successful one 
(more so than that of the conifers, in fact), there would appear to 
be no reason why variations from it in a direction that would be 
not only of no use to the plant but a positive hindrance to it would © 
be selected and preserved, even if they should occur. It seems 
very unlikely, therefore, that such a change of function did occur. 
If we start with the condition actually found among the Cor- 
daitales, where pollen grains without any tubes were deposited 
in a pollen chamber in the nucellus, can we see any sufficient reason 
for the giving up of the pollen chamber and the development of the 
tube? One type of ovule is just as easy to pollinate as the other, 
for if pollen can be gotten to the nucellar tip, there would appear 
to be no difficulty in getting it into the pollen chamber. [If it 
reached the pollen chamber safely and the pollen chamber broke 
through so as to give the swimming sperms direct access to the 
archegonia, it is difficult to see what would be gained by the giving 
up of the chamber and the formation of a pollen tube. That the 
conifer method is in fact inferior and would be selected against 
is strongly indicated by the fact that the proportional number of 
good seeds in their cones is decidedly less than that of cycad cones. 
The evidence would thus appear to be against such a derivation 
of the coniferous pollen tube. 
If it is difficult to see any adequate reason for the evolution of 
the ordinary coniferous pollen tube from the conditions found in 
the Cordaitales, it is vastly more difficult to imagine any adequate 
reason for its further evolution into the araucarian tube. We must 
imagine not only that the pollen chamber has been given up but 
that the place of pollen deposition has gradually retreated out 
through the micropyle and back along the scale from bad to worse. 
JEFFREY and CHRYSLER (11) would have us believe, not only that it 
did actually do this, but that to compensate itself for the dis- 
advantage it was compelled to form extensive lateral haustorial 
branches and to “‘proliferate” the two ‘“primitive’’ prothallial 
cells. I have not yet seen any reason advanced why the nucellus, 
having given up the habit of forming a pollen chamber, should have 
