30 Rhodora [FEBRUARY 
of structures have arisen bearing a close resemblance to structures 
previously present. It is to be noted that the secondary features 
have arisen in situations similar to those in which the antecedent 
features were already to be found. Stipules arise on the axis of 
the plant where a leaf is given off; stipels arise on the axis of the 
branched leaf where a leaflet is given off. Petiolar absciss-layers 
come at the junction of leaf-stalk and stem; petiolular absciss-layers 
come at the junction of leaflet-stalk and rhachis. 
Furthermore, the secondary features seem to be in the main use- 
less. Certain stipels —as the spinous or glandular kinds — are 
evidently advantageous, or at least functional, But most stipels are 
weak affairs for which there is no apparent office. With absciss- 
layers in petiolules, the probability of inutility is very strong indeed. 
The dismemberment of the exhausted and now valueless leaf, from 
which plastic materials have been withdrawn, no longer functionally 
a leaf but merely the form or frame of a leaf, seems superfluous. 
While uselessness of apparently unimportant members can perhaps 
in no case be absolutely proved, yet here the probability of it approx- 
imates certainty to such a degree that it may properly become a 
ground of reasoning. It is plainly within the limits of cogency which 
must be set in all such arguments as the present, based as they are 
not upon experiment but only upon observation. I shall assume, 
then, that for the most part the secondary features in question are 
useless. 
Why have these structures arisen so often, that is repeatedly and 
independently, in Dicotyledons? And especially why do we find 
one or both of them present so persistently in spite of inutility ? 
These questions find no answer in the modes of evolution assumed 
to be the common modes: the formations do not seem to be well 
accounted for as products of evolution by slow steps under guidance 
of Natural Selection. 
The true account of the structures I believe is that they are repe- 
titions, echoes as it were, of the previously existent parts which 
they so closely resemble, and that they first made their advent fully 
formed when the conditions for their production had been matured 
in the course of the evolution of the leaf. The conditions were the 
development of a distinct rhachis and petiolule, and the establish- 
ment, at their junction, of an allocation of parts very much like the 
allocation obtaining at the insertion of the petiole upon the stem. 
