1915] RITRLINGAMER—ARAIUCARIANS 105 
that this structure is a retained one and not a newly acquired one. 
If we could avoid the first assumption (that is, if we knew the 
structure of the ancient cone), we should not need this principle, 
but could apply the principle of resemblancé. Neither is it always 
easy to decide the correctness of the second assumption. Its limi- 
tations, therefore, are clear. It is useful in enabling us to infer a 
likeness which we do not know actually to have existed. It is, con- 
sequently, of much lower value than a direct comparison of known 
structures. Its highest possible value would equal that of a direct 
comparison between the homologous parts of the two plants, while 
its lowest value is actually zero. 
4. Ontogenetic recapitulations —This principle assumes that there 
will be formed in the juvenile stages of a plant or animal organs 
or structures that were characteristic of the adult ancestral forms. 
In the form I have stated it, this principle is almost certainly 
invalid. This is the form in which it is commonly applied. What 
is probably true is that related animals and plants resemble one 
another and their common ancestor at all stages from the egg to the 
adult, inclusive, in all those organs and structures which have 
neither been lost since the separation from the parent stock nor 
added to either of the descendants. This principle is not. infre- 
quently applied in such a manner as to deprive the conclusions of 
any real validity whatever. In so far as it possesses validity at 
all, it owes it to a direct comparison of homologous structures in 
the same stage of development. 
5. Traumatic reactions —When a plant or animal is wounded 
it not infrequently reacts by forming organs or structures that 
differ from those usually formed. In some cases these structures 
are such as are thought to be identical with those of its ancestors. 
There is no a priori reason why they should be reversions, so far as I 
can see, unless the original structures were introduced into the 
sum of the hereditary qualities through wounding in the first place. 
Pruning a grape vine or a fruit tree usually induces a yield of larger 
fruit or even a greater total quantity. Increased physiological 
activity is a very common result of wounding, but it does not there- 
fore follow that this is an ancestral quality of the stock. Some of 
these responses may represent ancestral conditions, but it seems to 
