Journal of Proceedings.
lxiii
petals. In the case represented in Fig. 1 the calyx has developed only
three sepals (instead of the normal four), and the corolla has also but
three petals; the stamens are six in number (as against 8, the usual
number with quadruple parts). The blossom illustrated by fig. 2 has
the petals (4) in a transitional stage, showing in a very interesting
manner the order in the whorl of the development, or degeneration—
whichever we may consider it—some being stalked and some rising
direct from the disc as usual. Thus, on my severing the petals, I found
that the first one proceeded from its base in the ordinary manner, as
shown at a (Fig. 3); the second was slightly stalked as at b; the third
spatula-shaped as at c; while the fourth was mounted on a distinct
filament or stalk, and the cup-like character of the blade was fully
attained.
Whether this abnormity is to be considered a further stage in the
evolution of the flower, or whether it is a retrogression, is a matter not
easy to form a judgment upon, but, as a new departure, it would almost
appear to be the former.
Since the reading of this note Mr. Herbert Brooks has kindly for-
warded to me another example of the same type, together with some
other variations in the inflorescence of Fuchsias from plants growing in
the same greenhouse. Other similar abnormities in respect of the
number of parts of this extremely variable blossom have also lately
come under my notice.
The following are details of these, commencing with the two already
described:—
Thus it will be seen that, with one remarkable exception (No. 6),
while the number of petals is either the same as, or a multiple of, the
number of sepals, there is no rule with regard to the stamens, and that,
of all the examples, only two have the same total number of parts.*
*Sprengel, in his discussion of the facts relative to the variability of abortive organs
of flowers, and the numerical proportion of parts, remarks, "It is very difficult to
establish any fixed principle respecting the value of numerical proportion. On the one
hand Nature shows a regularity which cannot be mistaken in the numerical proportion
of parts ; .... on the other hand she seems sometimes to sport in such a manner
with numbers, that in many genera we scarcely find the same number of filaments and
pistils in all the species. Astonishing changes often take place in this respect, from
abortion and union of parts." (' Elements of the Philosophy of Plants,' &c., by A. P.
DeCandolle and J. Sprengel, English Edition, Edin. 1821, p. 132). He does not refer,
however, to any particular cases similar to the foregoing, and I have been unable to find