160 PROF. HENFREY ON THE 
ix. p. 112), from the consideration of the structure of the genus 
Hydrocera. Through the kindness of Dr. Hooker, I have seen 
further observations by himself and Dr. Thomson, which not only 
confirm Roeper's view, in reference to Hydrocera, but illustrate it 
further by certain East Indian species of Impatiens, in which the 
number of sepals is either regularly or occasionally the same as in 
Hydrocera. In this genus the calyx has five sepals: first, the 
posterior spurred sepal; next, two lateral sepals, corresponding 
to the two small and usually green sepals of Zmpatiens Balsamina ; 
with two anterior sepals, which are mostly suppressed in Impa- 
tiens. Within this circle occurs a whorl of five petals, the anterior 
one being the so-called “ double sepal" of Kunth, which is inside 
the two anterior sepals just referred to. — . 
Payer (Traité d'Organogénie végétale) states that he finds the 
rudiments of this pair of anterior sepals on very young buds of 
Impatiens Royleana. Ihave not been able to find them in the 
Garden Balsam, nor in a developed state in the riumerous mon- 
strous specimens which I have examined. Оп the other hand, 
the monstrosity presently to be described favours the doctrine of 
Roeper far more than that of Kunth. In order to explain it more 
clearly, I have appended diagrams of the normal structure of 
Impatiens and Hydrocera (figs. 1 and 2). 
It is a well-known fact that the common Double Balsams of our 
gardens produce seed freely. Since only one cirele of stamens 
exists, we should scarcely have expected this; but the fact is that 
an extra corolline whorl is produced without the suppression of 
the stamens, and the metamorphosis takes place in a manner 
which bears an interesting relation to certain general questions of 
morphology, as well as to the theory of the flower of Impatiens. 
In the common Double Balsams of our gardens the flowers usually 
present a natural calyx, the small lateral sepals being often more 
or less coloured, and sometimes gibbous or slightly spurred; I 
bave never found the two anterior (suppressed) sepals developed. 
Within the calyx stand five petals (a broad anterior one and two 
pairs of lateral petals), which are, as usual, mostly more or less 
adherent by their limbs on each side, but with their claws free. 
Succeeding these are found five free petals resembling in appear- 
ance the lateral petals of the previous whorl, standing in the usual 
place of the stamens—that is, alternating with the normal petals, 
as shown in the diagram (fig. 8). Next comes a circle of five 
stamens, mostly all perfect, alternating with the preceding circle, 
and therefore in the ordinary place of the carpels. Lastly, the 
