994, REPORT—1896, 
To determine the blood-relationships of organisms it is necessary to decipher 
their past history, and the best evidence we can have (when we can get it) is from 
the ancient organisms themselves. The problem of the morphologist is an 
historical one, and contemporary documentary evidence is necessarily the best. It 
is paleontology alone which can give us the real historical facts. 
ANATOMICAL CHARACTERS. 
In judging of the affinities of fossil plants we are often compelled to make 
great use of vegetative characters, and more particularly of characters drawn 
from anatomical structure. It is true that in many cases we do so because we 
cannot help ourselves, such anatomical features being the only characters available 
in many of the specimens as at present known. But the value of the method has 
been amply proved in other cases where the reproductive structures have also 
been discovered, and are found to fully confirm the conelusions based on anatomy. 
I need only mention the great groups of the Lepidodendrew and the Calamites, in 
- each of which the anatomical characters, when accurately known, put us at once 
on the right track, and lead to results which are only confirmed by the study of 
the reproductive organs 
In this matter fossil botany is likely to react in a beneficial way on the study 
of recent plants, calling attention to points of structure which have been passed 
over, and showing us the value of characters of a kind to which systematists had 
until recently paid but little attention, At present, owing to the work of 
Radlkofer, Vesque, and others, anatomical characters are gradually coming into 
use in the classification of the higher plants, and in some quarters thore may even 
be a tendency to over-estimate their importance. Such exaggeration, however, is 
only a temporary fault. incident to the introduction of a comparatively new 
method. In the long run nothing but good can result from the effort to place our 
classification on a broader basis. In most cases the employment of additional 
characters will doubtless serve only to further confirm the affinities already 
detected by the acumen of the older taxonomists. There are plenty of doubtful 
points, however, where new light is much needed ; and even where the classifica- 
tion is not affected it will be a great scientific gain to know that its divisions are 
based on a comparison of the whole structure, and not merely on that of particular 
organs. 
e The fact that anatomical characters are adaptive is undeniable, but this applies 
to all characters, such difference as there is being merely one of degree. Cases 
are not wanting where the vegetative tissues show greater constancy than the 
organs of reproduction, as, for example, in the Marattiacew, where there is a 
great uniformity in anatomical structure throughout the family, while the 
sporangia show the important differences on which the distinction of the genera is 
based. It is in fact a mistake to suppose that anatomical characters are neces- 
sarily the expression of recent adaptations. On the contrary, it is easy to cite 
examples of marked anatomical peculiarities which have become the common 
property of large groups of plants. 
For instance, to take a case in which I happen to have been specially interested, 
the presence of bast to the inside as well as to the outside of the woody zone is.a 
modification of dicotyledonous structure which is in many groups, at least of 
ordinal value. The peculiarity is constant throughout the orders Onagraces, Ly- 
thracez, Myrtacez, Solanacexz, Asclepiadacew, and Apocynacee, not to mention 
some less important groups. In other families, such as the Cucurbitaces and the 
Gentianee, it is nearly constant throughout the order, but subject to some exceptions. 
Among the Composite a similar, if not identical, peculiarity appears in some of 
the sub-order Cichoriaceze, but is here not of more than generic value. In Campa- 
nula the systematic importance of internal phloém is even less, for it appears in some 
species and not in others. Lastly, there are cases in which a similar character 
actually appears as an individual variation, as in Carum Carvi, and, under abnor- 
mal conditions, in Phaseolus multiflorus, 
These latter cases seem to me worthy of special study, for in them we cau 
