1915] BURLINGAME—ARAUCARIANS 109 
its use of very doubtful value. It seems very unlikely that its 
mere absence from a mesozoic form otherwise unlike an araucarian 
is a character of sufficient importance to justify its inclusion in the 
Araucarineae. 
The argument from abnormalities as applied to this theory is 
subject to attack on the historical side as well as on the ground of 
inherent probability. That a given abnormality represents always 
a reversion is an assumption that no one seriously maintains. 
For example, six toes are not uncommon in mankind. No one 
believes that this is an ancestral character any more than brachy- 
dactyly, where there are fewer parts than usual. An extra digit 
in hoofed animals is, on the contrary, usually looked on as a reap- 
pearance of an ancestral condition, because this condition is 
believed on other and trustworthy grounds actually to have occurred 
in this evolutionary line. In’respect to the known conditions of the 
cones that may be supposed to be ancestral to modern pines, there is 
not a scintilla of evidence that they were any nearer the brachy- 
blastic condition than their modern representatives. As I have 
already pointed out, the case is still more difficult in regard to the 
Cordaitales,.the supposed remote ancestors, where the condition 
should, theoretically, be well developed. The strength of this 
theory is that it explains the pine cone, and its weakness is that it 
makes it necessary to apply the same explanation to other cones 
and shoots where it is very much less satisfying. 
To sum up, the argument for the abietinean theory, therefore, is 
seen to be of a less convincing kind. Moreover, there are other 
arguments of the same kind that favor the cordaitean theory. 
It seems clear to the writer, also, that there are well defined 
rules of evidence in the investigation of phylogenetic problems, 
but that the conclusions attained through their application have 
far less certainty than a chemical analysis or a mathematical 
prediction of a comet’s course. With care the latter attains a 
high degree of certainty. The degree of probability in phylo- 
genetic inquiries is more nearly that pertaining to the verdict of 
juries in our law courts. They are both always subject to attack 
by the introduction of new evidence. 
* 
