empurpling all their green leaves even to the ground: a quite distinct
nature in the flower, observe, this possession of a power to kindle the
leaf and stem with its own passion, from that of the heaths, roses, or
lilies, where the determined bracts or calicos assert themselves in
opposition to the blossom, as little pine-leaves, or mosses, or brown paper
packages, and the like.

13. The Giulietta, however, is again entirely separate from the other
leaf-flushing blossoms, in that, after the two green leaves next the flower
have glowed with its blue, while it lived, they do not fade or waste with
it, but return to their own former green simplicity, and close over it to
protect the seed. I only know this to be the case with the Giulietta
Regina; but suppose it to be (with variety of course in the colours) a
condition in other species,--though of course nothing is ever said of it in
the botanical accounts of them. I gather, however, from Curtis's careful
drawings that the prevailing colour of the Cape species is purple, thus
justifying still further my placing them among the Cytherides; and I am
content to take the descriptive epithets at present given them, for the
following five of this southern group, hoping that they may be explained
for me afterwards by helpful friends.

14. Bracteolata, C. 345. Oppositifolia, C. 492. Speciosa, C. 1790. These
three all purple, and scarcely distinguishable from sweet pease-blossom,
only smaller.

Stipulacea, C. 1715. Small, and very beautiful, lilac and purple, with a
leaf and mode of growth like rosemary. The "Foxtail" milkwort, whose name I
don't accept, C. 1006, is intermediate between this and the next species.

15. Mixta, C. 1714. I don't see what mingling is meant, except that it is