This great group contains no less than forty British genera, and a very large number of species. It comprises the Daisy (Bellis), Dandelion (Taraxacum), Groundsel (Senecio), Chrysanthemum, Thistle (Carduus), Lettuce (Lactuca), Hawkweed (Hieracium), etc. Though there are many differences in the structure of the flowers, as might naturally be expected in so large a group, still in many respects, they are very uniform. The florets are so closely united on a common head, that each group is commonly, though of course incorrectly, spoken of as a single flower. The so-called flower of the Daisy, for instance, is in reality a group of flowers; the outer row of which, or ray florets as they are termed, are unlike the rest and terminate on the outer side in a white leaf or "ray."

The advantages of this arrangement are: 1. That the flowers become much more conspicuous than would be the case if they were arranged singly,

2. That the facility with which the honey is obtained renders them more attractive to insects.

3. That the visits of the insects are more likely to be effectual, since the chances are that an insect which once alights, touches several, if not many, florets.

No wonder, therefore, that the Compositse are the most extensive family among flowering plants, are represented in every quarter of the globe and in every description of station (Bentham, "Handbook of the British Flora," vol. i. p. 408; Jour. Linn. Soc. 1873, p. 335,) and contain nearly ten thousand species.

The principal differences among the Compositae, regarded from the point of view of the present work, consist in the different length of the florets, rendering the honey more or less accessible to insects; in the arrangement of the stamens and pistil; and in the character of the outer, or "ray" florets.

In some of the Compositae the florets all contain both anthers and stigma. Generally, however, the ray florets develope no anthers, but a pistil only; while in some species of Centaurea they are barren, and merely serve as flags. It is remarkable that in C. nigra, while the outer row of florets generally resemble the rest, they are sometimes enlarged and neuter, as in C. cyanus etc. As regards the relation to insects, we find every gradation, between Taraxacum, Cirsium arvense, and Achillea, on the one hand which are conspicuous, rich in honey and much visited by insects; and on the other, Senecio vulgaris, which is rarely visited by insects, and the species of Artemisia, which are said to be wind fertilised.

In Tussilago farfara the disk florets give honey and pollen; the ray florets contain neither, but render the flower-head conspicuous, and produce seed.

In the common Feverfew, or large white Daisy (Fig. 86), (Chrysantliemum parthenium), which has been well described by Dr. Ogle, "Popular Science Review," April 1870, the flower-heads consist of an outer row of female florets, in which the tubular corolla terminates on the outer side in a white leaf or ray, which doubtless is useful in making the flower conspicuous. The inner florets are also tubular, but are small, yellow, and without a ray. Each of these florets is furnished with stamens as well as a pistil. The anthers are united at their sides so as to form a closed tube, within which the pistil lies. They ripen before the pistil, and open on their inner sides, so that the pollen is discharged into the upper end of the tube above the head of the pistil. When the flower opens, the pollen is already ripe, and fills the upper part of the stamen tube. A floret in this condition is represented in (Fig. 87). The pistil, however, continues to elongate, and at length pushes the pollen against the upper end of the tube, which gives way, and thus the pollen is forced out of the tube, as shown in (Fig 88). The pistil itself terminates in two branches, which at first are pressed closely to one another, and each of which terminates in a brush of hairs (Fig. 89). As the style elongates this brush of hairs sweeps the pollen cleanly out of the tube, and it is then removed by insects. When the pistil has attained its full length, the two branches open and curve downwards, so as to expose the stigmatic surfaces (Fig. 89 st) which had previously been pressed closely to one another, and thus protected from the action of the pollen. From this arrangement it is obvious that any insect alighting on the flower-head of the Chrysanthemum would dust its under-side with the pollen of the younger flowers, which then could not fail to be brought into contact with the stigmatic surfaces of the older ones. As the expansion of the flowers begins at the outside and thence extends to the centre, it is plain that the pollen of any given floret cannot be used to fertilise one situated on its inner side. Consequently, if the outer row of florets produced pollen, it would, in the great majority of cases, be wasted. I have, however, already mentioned that these florets do not produce pollen, while the saving thus effected enables them to produce a larger corolla. It is also interesting to observe that in these outer flowers the branches of the pistil do not possess the terminal brush of hairs which, in the absence of pollen, would be useless.

Chrysanthemum parihenium.

Fig. 86. - Chrysanthemum parihenium.

Floret of Chrysanthemum parthenium, ]ust opened.

Fig. 87.

Ditto, somewhat more advanced

Fig. 88.

Ditto, with the stigmas expanded

Fig. 89.

Fig. 87. - Floret of Chrysanthemum parthenium, ]ust opened. Fig. 88. - Ditto, somewhat more advanced. Fig. 89. - Ditto, with the stigmas expanded.

In other Composite, as in the Marigold, while the ray flowers produce no pollen, the disk flowers develop no stigmas. In this case, as in the Feverfew, the pistil of the ray flowers does not require or possess the terminal brushes of hairs, there being no pollen to be swept out. The central flowers, on the other hand, though they develop no stigmas, require a pistil in order to force the pollen out of the anther tube. Hence the pistil is present as usual, but the head is simple and not bifid. This complete alteration of the function of the pistil is extremely curious.