triangular, and it has usually three stigmas or sensitive surfaces, thus dimly pointing back to the three distinct cells of its lily-like ancestors, and the three separate ovaries of its still earlier alisma-like progenitors. In many species, however, even this last souvenir of the trinary type has been utterly obliterated, the ovary having only two stigmas, and assuming a flattened, two-sided shape. In all the carices the flowers are loosely arranged in compact spikes and spikelets, with their mobile stamens hanging out freely to the breeze, and their feathery stigmas prepared to catch the slightest grain of pollen which may happen to be wafted their way by any passing breath of air. The varieties in their arrangement, however, are almost as infinite among the different species as those of the grasses themselves; sometimes the male and female flowers are produced on separate plants; sometimes they grow in separate spikes on the same plant; sometimes the same spike has male flowers at the top and female at the bottom; sometimes the various flowers are mixed up with one another at top and bottom, a regular hotch-potch of higgledy-piggledy confusion. But all the sedges alike are very grass-like in their aspect, with thin blades by way of leaves, and blossoms on tall heads, as in the grasses. In fact, the two families are never accurately distinguished by any except technical botanists; to the ordinary observer, they are all grasses together, without petty distinctions of genus and species. Like the grasses, too, the sedges are mostly plants of the open, wind-swept plains or marshy levels, where the facilities for wind-fertilization are greatest and most constantly present.
And now, from this illustrative digression, let us hark back again to the junction-point of the rushes, whence alike the sedges and the grasses appear to diverge. In order to understand the nature of the steps by which the cereals have been developed from rush-like ancestors, it will be necessary to look shortly at the actual composition of the flower in grasses, which is the only part of their organism differing appreciably from the ordinary lily type. The blossoms of grasses, in their simplest form, consist of several little green florets, arranged in small clusters, known as spikelets, along a single common axis. Of this arrangement, the head of wheat itself offers a familiar and excellent example. If we pull to pieces one of the spikelets composing such a head, we find it to consist of four or five distinct florets. Omitting special features and unnecessary details, we may say that each floret is made up of two chaffy scales, known as pales, and representing the calyx, together with a pair of small white petals known as lodicules, three stamens, and an ovary with two feathery styles. Moreover, the two pales or calyx-pieces are not similar and symmet-
- The sedges are not, in all probability, a real natural family, but are a group of heterogeneous, degraded lilies, containing almost all those kinds in which the reduced florets are covered by a single conspicuous glume-like bract. It will be seen from the sequel that these bracts are not truly analogous to the glumes or outer palese of grasses.