1878. ] On the. Natural Succession of the Dicotyledons. 729 
among all the characters of a plant so that those which have 
similar organs of reproduction usually exhibit strong family rela- 
tionships. Were it not so familiar it would be a surprising fact 
that a Solanaceous plant can usually be detected as such without 
examining the flowers, merely by a certain undefinable appear- 
ance which belongs to the family. The same is true to a greater 
or less extent with all the large orders, Crucifere, Lorraginacee, 
Urticacee, etc. In some families still more subtle characteristics 
persist with great uniformity, of which the peculiar odor of the 
Orchidacee is a good example. 
This so-called “general aspect ” is in reality the exsemd/e of all - 
the characters which make a plant such as it is, and though any 
one character is as liable to vary as another or as more obvious 
ones, their number is so great that it requires an enormous period 
of time to so efface them all as to destroy all traces of resem- 
blance. Not so of any particular character which botanists may 
fix upon. This may vary in a manner comparatively rapid, and 
thus it doubtless often happens that species really related and 
bearing a general resemblance are divorced in the text books on 
special differences. 
This co-existence of so large a number of minor peculiarities 
as to give to two plants or groups of plants an obvious genéral 
resemblance should therefore be welcomed as a valuable accessory 
to the work of classification, not as a special guide to truth but as 
a general check upon error. A strong physiognomic resem- 
blance between two groups of plants should at least raise a sus- 
picion of their genetic relationship, and might frequently furnish 
a theory for the investigation of important questions. In a pre- 
vious paper it was shown how the physiognomy of the Cycadacee 
pointed to their natural position between the ferns and the palms, 
and how a closer inspection of the more reliable characters sus- 
tained this conclusion. If now we return to the case of the 
Ranunculacee and Rosaceæ above referred to, we shall find a fur- 
ther confirmation of this law. A careful comparison of all the 
genera of these two orders, which has been recently made, reveals 
the fact that there is to be traced “an easy transition from the 
wholly conical and much elongated receptacle of Ranunculus and 
Myosurus to that of Fragaria, flattened at the base and conical in 
the center, or of Rubus with its raised margin and convex center. 
From this we may pass to Sibbaldia, Potentilla, Horkelia and 
