be to regard the leaf as consisting of two parts(1) the base, with or without foliar expansions, the stipules; and (2) the upper part or petiole, with or without a foliar expansion, the leaf-blade.

Sometimes all four divisions are present, as in the Rose; sometimes the stipules are absent, as in Maples; sometimes the leaf is sessile, the stalk, stipules and base being undeveloped, as in Gentians; sometimes the blade is absent, and its functions are performed by the flattened petiole, as in most of the Australian Acacias; sometimes the stipules perform the function of the leaf-blade, as in Lathyrus Aphaca (fig. 294).

That Helianthemum vulgare (fig. 22, p. 20) has stipules and H. celandicum (fig. 23, p. 21) has none is clear and simple enough, but there are many cases in which the presence or absence of stipules is far from being so easy to determine, and which have been the subject of much difference of opinion amongst botanists.

For instance, the Composites generally (Daisies, Chrysanthemums, Dandelions, Lettuces, &c.) are said to be exstipulate, but in some the petioles are dilated at the base into appendages which can scarcely be distinguished from stipules.

In some Crucifers, as, for instance, in Cardamine impatiens, some of the lower leaves have rounded and thickened auricles, which in the upper leaves become prolonged into subulate, obtuse, falcate processes clasping the stem. These are often called stipules. They

correspond to the auricles of other Crucifers which are continuous with the margin of the leaf.

The Umbellifers (Carrot, Parsnip, Parsley, &c.), again, have the petioles flattened and broad, often In some the upper or smaller part of the sheath becomes detached. When this has

forming a sheath.

FIG. 314.-LEAF OF PARSLEY.

FIG. 315.-LEAF OF OSTRUTHIUM.

proceeded sufficiently far the free lobes may fairly be called stipules. Indeed, it is obvious that there may be every intermediate gradation between a species with a merely flattened petiole and one with true stipules. Nay, such transitions may and do occur in the same plant, and even in the same bud.

I have figured leaves of three members of this family to illustrate this relation between leaf-sheath and stipule. In the Parsley (fig. 314) the sheath is quite entire; in Ostruthium (fig. 315) there is a distinct shoulder on each side; while in Fool's Parsley (Ethusa Cynapium) (fig. 316), a common British species, the shoulder has grown out into a small but distinct stipule.

FIG. 316.-LEAF OF ETHUSA.

Speaking of Canarium, Bentham and Hooker say ('Genera Plant.,' i. 324): 'Folia exstipulata, vel pinnulis inferioribus sessilibus stipulæformibus, imparipinnata,' and again, Foliola infima Canarii interdum stipulas simulant' (p. 321).

The small primary leaves of the Mistletoe have been mistaken by some authors for stipules.

Aristolochia elegans, while really exstipulate, has a small cordate, membranous, subsessile leaf in the axil, which resembles a single axillary stipule; sometimes a pair are present. Close examination shows that this stipule-like process is really the first leaf of an axillary axis or bud. It clasps the main axis with its auricles, and has two buds in different stages of advancement lying between it and the petiole of the leaf in whose axil it occurs. The small leaf belongs to the larger and more advanced bud.

The early unfolding of the first pair of leaves of an axillary bud may often give the appearance of stipules. For instance, in Tutsan (Hypericum Androsæmum) the leaves are opposite and exstipulate, but the first two leaves of the axillary branch, before any further growth of the bud takes place, stand right and left above the base of each leaf, and by a careless observer may be taken for stipules.

The genus Lotus (Bird's Foot Trefoil) is quinquefoliate. The lower folioles have been sometimes regarded as leaflets, sometimes as stipules. Bentham and Hooker, in the 'Genera Plantarum,' say, 'Folia 4-5foliolata, foliolis integerrimis 3 ad apicem petioli confertis, 1-2 juxta caulem stipulas simulantibus.'

Speaking of the corresponding organs in Tetragonolobus, Norman says that they' par leur structure, leur

coloration, leur forme et leur position, sont entièrement différentes des folioles, et s'harmonisent tout à fait avec les stipules que l'on trouve dans la plupart des Légumineuses' (28).

The same may also be said of the allied genera Dorycnium, Bonjeania, &c.

Cambessèdes, however, maintains that the true stipules of Lotus and the other genera mentioned are two minute glandular elevations at the base of the leaf.

The Leguminosae, to which these genera belong, are as a rule stipulate; so that if this view is correct we have an interesting case in which the stipules have dwindled away, and their function is performed by the lower leaflets, which have to some extent wandered away from their usual position in order effectively to take on this new duty.

Again, the tendrils of Cucumbers (Cucurbitaceae) and Smilax have been regarded by some botanists as stipules, while this has been denied by others; and the same difference of opinion has existed with reference to the ligule of grasses.

I shall presently return to some of these doubtful

cases.

Helianthemum guttatum is a very interesting species. As Dr. Stenzel (29) points out, the lower part of the shoots bears lancet-shaped leaves without stipules. Nearer the flower-buds the leaves tend to become smaller, and have stipules. These are linear, and attached quite at the