The Evolution of Fruits. 13
the various apocarpous fruits, first spiral and then whorled,
of the Ranales, of which the Laurales are probably reduced
types, lead through the fruits of the water-lilies (Nymphaea)
and of Sarracenia to that of the poppy (Papaver); and in the
same tribe we have reversion to apocarpy in Platystemon and
a transition to the siliqua in Glaucium and Chelidonium.
Among Cruciferae the genus Tetratoma (an illustration of the
fact, well known to biologists, that exceptional forms teach us
more than those which are regular or typical) indicates the
close affinity to the other Parietales, the capsules of Resedaceae,
Violacea, &c.; and we have here probably a great "centre of
divergence" leading to the Disciflorae, Guttiferales, Malvales,
and Centrospermae. Of the latter mention has been made
already; among the Malvales I can only now refer to the
succulent Malvaviscus and the sand-box tree (Hum crepitans)
which indicates that relationship between the Malvaceae and
Euphorbiaceae which is paralleled by that between Tiliacea
and Urticaceae; among the Guttiferales the capsule with
central placentation leads from Camelliaceae to Ericaceae; and
among Disciflorae the superior schizocarp with a central
gynophore in Geranium leads to the fruit of Tropaeolum and
the stipitate capsule in the Rutaceae, which, being sometimes
fleshy, passes to the "hesperidia" of the orange tribe, whilst
the Celastrales, with the varying "nuculanes" of the holly,
Euonymus, Buckthorn, and Vine, lead to the still more
variable fruits of the Sapindales, which range from the horse-
chestnut and Litchi to the winged "samara" of maple and
sycamore.
The Perigynae, commencing with the numerous related
fruit-types of Rosaceae, among which the "etaerio" of achenes
in Potentilla perhaps represents the primitive type, lead
through the Drupacea: to the legume, perhaps through Pomcae
to the nut of the Cupuliferae, and through Spiraea to the
Saxifrages. Here we get a semi-inferior fruit with the
carpels reduced in number, near which the branch-phylum of
the Bicarpellatae (well represented by the Umbellales) may
have been given off; whilst it has certainly been a centre of
divergence leading through Hildebrandia to the Beyoniaceae