155. So the palmate leaf, if again compounded in the same way, becomes twice palmate, or, as we say when the divisions are in threes, twice ternate (in Latin form biternate); if a third time compounded, thrice ternate or triternate. But if the division goes still further, or if the degree is variable, we simply say that the leaf is decompound; either palmately or pinnately decompound, as the case may be. Thus, Fig. 161 represents a four times ternately compound (in other words a ternately decompound) leaf of a common Meadow Rue.

158. Perfoliate Leaves. In these the stem that bears them seems to run through the blade of the leaf, more or less above its base. A common Bellwort (Uvularia perfoliata, Fig. 162) is a familiar illustration. The lower and earlier leaves show it distinctly. Later, the plant is apt to produce some leaves merely clasping the stem by the sessile and heart-shaped base, and the latest may be merely sessile. So the series explains the peculiarity: in the formation of the leaf the bases, meeting around the stem, grow together there.

159. Connate-perfoliate. Such are the upper leaves of true Honeysuckles. Here (Fig. 163) of the opposite and sessile leaves, some pairs, especially the uppermost, in the course of their formation unite around the stem, which thus seems to run through the disk formed by their union.

160. Equitant Leaves. While ordinary leaves spread horizontally, and present one face to the sky and the other to the earth, there are some that present their tip to the sky, and their faces right and left to the horizon. Among these are the equitant leaves of the Iris or Flower-de-Luce. Inspection shows that each leaf was formed as if folded together lengthwise, so that what would be the upper surface is within, and all grown together, except next the bottom, where each leaf covers the next younger one. It was from their straddling over each other, like a man on horseback (as is seen in the cross-section, Fig. 165), that Linnæus, with his lively fancy, called these Equitant leaves.

161. Leaves with no distinction of Petiole and Blade. The leaves of Iris just mentioned show one form of this. The flat but narrow leaves of Jonquils, Daffodils, and the cylindrical leaf of Onions are other instances. Needle-shaped leaves, like those of the Pine, Larch, and Spruce, and the awl-shaped as well as the scale-shaped leaves of Junipers, Red Cedar, and Arbor-Vitæ (Fig. 166), are examples.

162. Phyllodia. Sometimes an expanded petiole takes the place of the blade; as in numerous New Holland Acacias, some of which are now common in greenhouses. Such counterfeit blades are called phyllodia,—meaning leaf-like bodies. They may be known from true blades by their standing edgewise, their margins being directed upwards and downwards; while in true blades the faces look upwards and downwards; excepting in equitant leaves, as already explained.

163. Falsely Vertical Leaves. These are apparent exceptions to the rule, the blade standing edgewise instead of flatwise to the stem; but this position comes by a twist of the stalk or the base of the blade. Such leaves present the two faces about equally to the light. The Compass-plant (Silphium laciniatum) is an example. So also the leaves of Boltonia, of Wild Lettuce, and of a vast number of Australian Myrtaceous shrubs and trees, which much resemble the phyllodia of the Acacias of the same country. They are familiar in Callistemon, the Bottle-brush Flower, and in Eucalyptus. But in the latter the leaves of the young tree have the normal structure and position.

164. Cladophylla, meaning branch-leaves. The foliage of Ruscus (the Butcher's Broom of Europe) and of Myrsiphyllum of South Africa (cultivated for decoration under the false name of Smilax) is peculiar and puzzling. If these blades (Fig. 167, 168) are really leaves, they are most anomalous in occupying the axil of another leaf, reduced to a little scale. Yet they have an upper and lower face, as leaves should, although they soon twist, so as to stand more or less edgewise. If they are branches which have assumed exactly the form and office of leaves, they are equally extraordinary in not making any further development. But in Ruscus, flowers are borne on one face, in the axil of a little scale: and this would seem to settle that they are branches. In Asparagus just the same things as to position are thread-shaped and branch-like.

§ 2. LEAVES OF SPECIAL CONFORMATION AND USE.

165. Leaves for Storage. A leaf may at the same time serve both ordinary and special uses. Thus in those leaves of Lilies, such as the common White Lily, which spring from the bulb, the upper and green part serves for foliage and elaborates nourishment, while the thickened portion or bud-scale beneath serves for the storage of this nourishment. The thread-shaped leaf of the Onion fulfils the same office, and the nourishing matter it prepares is deposited in its sheathing base, forming one of the concentric layers of the onion. When these layers, so thick and succulent, have given up their store to the growing parts within, they are left as thin and dry husks. In a Houseleek, an Aloe or an Agave, the green color of the surface of the fleshy leaf indicates that it is doing the work of foliage; the deeper-seated white portion within is the storehouse of the nourishment which the green surface has elaborated. So, also, the seed-leaves or cotyledons are commonly used for storage. Some, as in one of the Maples, the Pea, Horse-chestnut, Oak, etc., are for nothing else. Others, as in Beech and in our common Beans, give faint indications of service as foliage also, chiefly in vain. Still others, as in the Pumpkin and Flax, having served for storage, develop into the first efficient foliage. Compare 11, 22-30, and the accompanying figures.

166. Leaves as Bud-Scales serve to protect the forming parts within. Having fulfilled this purpose they commonly fall off when the shoot develops and foliage-leaves appear. Occasionally, as in Fig. 170, there is a transition of bud-scales to leaves, which reveals the nature of the former. The Lilac also shows a gradation from bud-scale to simple leaf. In Cornus florida (the Flowering Dogwood), the four bud-scales which through the winter protect the head of forming flowers remain until blossoming, and then the base of each grows out into a large and very showy petal-like leaf; the original dry scale is apparent in the notch at the apex.

167. Leaves as Spines occur in several plants. A familiar instance is that of the common Barberry (Fig. 171). In almost any summer shoot, most of the gradations may be seen between the ordinary leaves, with sharp bristly teeth, and leaves which are reduced to a branching spine or thorn. The fact that the spines of the Barberry produce a leaf-bud in their axil also proves them to be leaves.

168. Leaves for Climbing are various in adaptation. True foliage-leaves serve this purpose; as in Gloriosa, where the attenuated tip of a simple leaf (otherwise like that of a Lily) hooks around a supporting object; or in Solanum jasminoides of the gardens (Fig. 172), and in Maurandia, etc., where the leaf-stalk coils round and clings to a support; or in the compound leaves of Clematis and of Adlumia, in which both the leaflets and their stalks hook or coil around the support.

169. Or in a compound leaf, as in the Pea and most Vetches, and in Cobæa, while the lower leaflets serve for foliage, some of the uppermost are developed as tendrils for climbing (Fig. 167). In the common Pea this is so with all but one or two pairs of leaflets.

170. In one European Vetch, the leaflets are wanting and the whole petiole is a tendril, while the stipules become the only foliage (Fig. 173).

171. Leaves as Pitchers, or hollow tubes, are familiar in the common Pitcher-plant or Side-saddle Flower (Sarracenia, Fig. 174) of our bogs. These pitchers are generally half full of water, in which flies and other insects are drowned, often in such numbers as to make a rich manure for the plant. More curious are some of the southern species of Sarracenia, which seem to be specially adapted to the capture and destruction of flies and other insects.

172. The leaf of Nepenthes (Fig. 175) combines three structures and uses. The expanded part below is foliage: this tapers into a tendril for climbing; and this bears a pitcher with a lid. Insects are caught, and perhaps digested, in the pitcher.

173. Leaves as Fly-traps. Insects are caught in another way, and more expertly, by the most extraordinary of all the plants of this country, the Dionæa or Venus's Fly-trap, which grows in the sandy bogs around Wilmington, North Carolina. Here (Fig. 176) each leaf bears at its summit an appendage which opens and shuts, in shape something like a steel-trap, and operating much like one. For when open, no sooner does a fly alight on its surface, and brush against any one of the two or three bristles that grow there, than the trap suddenly closes, capturing the intruder. If the fly escapes, the trap soon slowly opens, and is ready for another capture. When retained, the insect is after a time moistened by a secretion from minute glands of the inner surface, and is digested. In the various species of Drosera or Sundew, insects are caught by sticking fast to very viscid glands at the tip of strong bristles, aided by adjacent gland-tipped bristles which bend slowly toward the captive. The use of such adaptations and operations may be explained in another place.

§ 3. STIPULES.

174. A leaf complete in its parts consists of blade, leaf-stalk or petiole, and a pair of stipules. But most leaves have either fugacious or minute stipules or none at all; many have no petiole (the blade being sessile or stalkless); some have no clear distinction of blade and petiole; and many of these, such as those of the Onion and all phyllodia (166), consist of petiole only.

175. The base of the petiole is apt to be broadened and flattened, sometimes into thin margins, sometimes into a sheath which embraces the stem at the point of attachment.

176. Stipules are such appendages, either wholly or partly separated from the petiole. When quite separate they are said to be free, as in Fig. 112. When attached to the base of the petiole, as in the Rose and in Clover (Fig. 177), they are adnate. When the two stipules unite and sheathe the stem above the insertion, as in Polygonum (Fig. 178), this sheath is called an Ocrea from its likeness to a greave or leggin.

177. In Grasses, when the sheathing base of the leaf may answer to petiole, the summit of the sheath commonly projects as a thin and short membrane, like an ocrea: this is called a Ligula or Ligule.

178. When stipules are green and leaf-like they act as so much foliage. In the Pea they make up no small part of the actual foliage. In a related plant (Lathyrus Aphaca, Fig. 173), they make the whole of it, the remainder of the leaf being tendril.

179. In many trees the stipules are the bud-scales, as in the Beech, and very conspicuously in the Fig-tree, Tulip-tree, and Magnolia (Fig. 179). These fall off as the leaves unfold.

180. The stipules are spines or prickles in Locust and several other Leguminous trees and shrubs; they are tendrils in Smilax or Greenbrier.

§ 4. THE ARRANGEMENT OF LEAVES.

181. Phyllotaxy, meaning leaf-arrangement, is the study of the position of leaves, or parts answering to leaves, upon the stem.

182. The technical name for the attachment of leaves to the stem is the insertion. Leaves (as already noticed, 54) are inserted in three modes. They are

Alternate (Fig. 181), that is, one after another, or in other words, with only a single leaf to each node;

Opposite (Fig. 182), when there is a pair to each node, the two leaves in this case being always on opposite sides of the stem;

Whorled or Verticillate (Fig. 183) when there are more than two leaves on a node, in which case they divide the circle equally between them, forming a Verticel or whorl. When there are three leaves in the whorl, the leaves are one third of the circumference apart; when four, one quarter, and so on. So the plan of opposite leaves, which is very common, is merely that of whorled leaves, with the fewest leaves to the whorl, namely, two.

184. No two or more leaves ever grow from the same point. The so-called Fascicled or Clustered leaves are the leaves of a branch the nodes of which are very close, just as they are in the bud, so keeping the leaves in a cluster. This is evident in the Larch (Fig. 184), in which examination shows each cluster to be made up of numerous leaves crowded on a spur or short axis. In spring there are only such clusters; but in summer some of them lengthen into ordinary shoots with scattered alternate leaves. So, likewise, each cluster of two or three needle-shaped leaves in Pitch Pines (as in Fig. 185), or of five leaves in White Pine, answers to a similar extremely short branch, springing from the axil of a thin and slender scale, which represents a leaf of the main shoot. For Pines produce two kinds of leaves,—1. primary, the proper leaves of the shoots, not as foliage, but in the shape of delicate scales in spring, which soon fall away; and 2. secondary, the fascicled leaves, from buds in the axils of the former, and these form the actual foliage.

185. Phyllotaxy of Alternate Leaves. Alternate leaves are distributed along the stem in an order which is uniform for each species. The arrangement in all its modifications is said to be spiral, because, if we draw a line from the insertion (i. e. the point of attachment) of one leaf to that of the next, and so on, this line will wind spirally around the stem as it rises, and in the same species will always bear the same number of leaves for each turn round the stem. That is, any two successive leaves will always be separated from each other by an equal portion of the circumference of the stem. The distance in height between any two leaves may vary greatly, even on the same shoot, for that depends upon the length of the internodes, or spaces between the leaves; but the distance as measured around the circumference (in other words, the Angular Divergence, or angle formed by any two successive leaves) is uniformly the same.

186. Two-ranked. The greatest possible divergence is, of course, where the second leaf stands on exactly the opposite side of the stem from the first, the third on the side opposite the second, and therefore over the first, and the fourth over the second. This brings all the leaves into two ranks, one on one side of the stem and one on the other, and is therefore called the Two-ranked arrangement. It occurs in all Grasses,—in Indian Corn, for instance; also, in the Basswood (Fig. 181). This is the simplest of all arrangements, and the one which most widely distributes successive leaves, but which therefore gives the fewest vertical ranks. Next is the

187. Three-ranked arrangement,—that of all Sedges, and of White Hellebore. Here the second leaf is placed one third of the way round the stem, the third leaf two thirds of the way round, the fourth leaf accordingly directly over the first, the fifth over the second, and so on. That is, three leaves occur in each turn round the stem, and they are separated from each other by one third of the circumference. (Fig. 186, 187.)

188. Five-ranked is the next in the series, and the most common. It is seen in the Apple (Fig. 188), Cherry, Poplar, and the greater number of trees and shrubs. In this case the line traced from leaf to leaf will pass twice round the stem before it reaches a leaf situated directly over any below (Fig. 189). Here the sixth leaf is over the first; the leaves stand in five perpendicular ranks, with equal angular distance from each other; and this distance between any two successive leaves is just two fifths of the circumference of the stem.

190. Eight-ranked arrangement. In this the ninth leaf stands over the first, and three turns are made around the stem to reach it; so it is expressed by the fraction 3/8. This is seen in the Holly, and in the common Plantain. Then comes the

191. Thirteen-ranked arrangement, in which the fourteenth leaf is over the first, after five turns around the stem. The common Houseleek (Fig. 191) is a good example.

192. The series so far, then, is 1/2, 1/3, 2/5, 3/8, 5/13; the numerator and the denominator of each fraction being those of the two next preceding ones added together. At this rate the next higher should be 8/21, then 13/34, and so on; and in fact just such cases are met with, and (commonly) no others. These higher sorts are found in the Pine Family, both in the leaves and the cones and in many other plants with small and crowded leaves. But in those the number of the ranks, or of leaves in each cycle, can only rarely be made out by direct inspection. They may be indirectly ascertained, however, by studying the secondary spirals, as they are called, which usually become conspicuous, at least two series of them, one turning to the right and one to the left, as shown in Fig. 191. For an account of the way in which the character of the phyllotaxy may be deduced from the secondary spirals, see Structural Botany, Chapter IV.

193. Phyllotaxy of Opposite and whorled Leaves. This is simple and comparatively uniform. The leaves of each pair or whorl are placed over the intervals between those of the preceding, and therefore under the intervals of the pair or whorl next above. The whorls or pairs alternate or cross each other, usually at right angles, that is, they decussate. Opposite leaves, that is, whorls of two leaves only, are far commoner than whorls of three or four or more members. This arrangement in successive decussating pairs gives an advantageous distribution on the stem in four vertical ranks. Whorls of three give six vertical ranks, and so on. Note that in descriptive botany leaves in whorls of two are simply called opposite leaves; and that the term verticillate or whorled, is employed only for cases of more than two, unless the latter number is specified.

194. Vernation or Præfoliation, the disposition of the leaf-blades in the bud, comprises two things; 1st, the way in which each separate leaf is folded, coiled, or packed up in the bud; and 2d, the arrangement of the leaves in the bud with respect to one another. The latter of course depends very much upon the phyllotaxy, i. e. the position and order of the leaves upon the stem. The same terms are used for it as for the arrangement of the leaves of the flower in the flower-bud. See, therefore, "Æstivation, or Præfloration."

195. As to each leaf separately, it is sometimes straight and open in vernation, but more commonly it is either bent, folded, or rolled up. When the upper part is bent down upon the lower, as the young blade in the Tulip-tree is bent upon the leaf-stalk, it is said to be Inflexed or Reclined in vernation. When folded by the midrib so that the two halves are placed face to face, it is Conduplicate (Fig. 193), as in the Magnolia, the Cherry, and the Oak. When folded back and forth like the plaits of a fan, it is Plicate or Plaited (Fig. 194), as in the Maple and Currant. If rolled, it may be so either from the tip downwards, as in Ferns and the Sundew (Fig. 197), when in unrolling it resembles the head of a crosier, and is said to be Circinate; or it may be rolled up parallel with the axis, either from one edge into a coil, when it is Convolute (Fig. 195), as in the Apricot and Plum; or rolled from both edges towards the midrib,—sometimes inwards, when it is Involute (Fig. 198), as in the Violet and Water-Lily; sometimes outwards, when it is Revolute (Fig. 196), in the Rosemary and Azalea. The figures are diagrams, representing sections through the leaf, in the way they were represented by Linnæus.

Section VIII. FLOWERS.

196. Flowers are for the production of seed (16). Stems and branches, which for a time put forth leaves for vegetation, may at length put forth flowers for reproduction.

§ 1. POSITION AND ARRANGEMENT OF FLOWERS, OR INFLORESCENCE.

197. Flower-buds appear just where leaf-buds appear; that is, they are either terminal or axillary (47-49). Morphologically, flowers answer to shoots or branches, and their parts to leaves.

198. In the same species the flowers are usually from axillary buds only, or from terminal buds only; but in some they are both axillary and terminal.

199. Inflorescence, which is the name used by Linnæus to signify mode of flower-arrangement, is accordingly of three classes: namely, Indeterminate, when the flowers are in the axils of leaves, that is, are from axillary buds; Determinate, when they are from terminal buds, and so terminate a stem or branch; and Mixed, when these two are combined.

200. Indeterminate Inflorescence (likewise, and for the same reason, called indefinite inflorescence) is so named because, as the flowers all come from axillary buds, the terminal bud may keep on growing and prolong the stem indefinitely. This is so in Moneywort (Fig. 199).

201. When flowers thus arise singly from the axils of ordinary leaves, they are axillary and solitary, not collected into flower-clusters.

202. But when several or many flowers are produced near each other, the accompanying leaves are apt to be of smaller size, or of different shape or character: then they are called Bracts, and the flowers thus brought together form a cluster. The kinds of flower-clusters of the indeterminate class have received distinct names, according to their form and disposition. They are principally Raceme, Corymb, Umbel, Spike, Head, Spadix, Catkin, and Panicle.

203. In defining these it will be necessary to use some of the following terms of descriptive botany which relate to inflorescence. If a flower is stalkless, i. e. sits directly in the axil or other support, it is said to be sessile. If raised on a naked stalk of its own (as in Fig. 199) it is pedunculate, and the stalk is a Peduncle.

204. A peduncle on which a flower-cluster is raised is a Common peduncle. That which supports each separate flower of the cluster is a Partial peduncle, and is generally called a Pedicel. The portion of the general stalk along which flowers are disposed is called the Axis of inflorescence, or, when covered with sessile flowers, the Rhachis (back-bone), and sometimes the Receptacle. The leaves of a flower-cluster generally are termed Bracts. But when bracts of different orders are to be distinguished, those on the common peduncle or axis, and which have a flower in their axil, keep the name of bracts; and those on the pedicels or partial flower-stalks, if any, that of Bractlets or Bracteoles. The former is the preferable English name.

205. A Raceme (Fig. 200) is that form of flower-cluster in which the flowers, each on their own foot-stalk or pedicel, are arranged along the sides of a common stalk or axis of inflorescence; as in the Lily of the Valley, Currant, Barberry, one section of Cherry, etc. Each flower comes from the axil of a small leaf, or bract, which, however, is often so small that it might escape notice, and even sometimes (as in the Mustard Family) disappears altogether. The lowest blossoms of a raceme are of course the oldest, and therefore open first, and the order of blossoming is ascending from the bottom to the top. The summit, never being stopped by a terminal flower, may go on to grow, and often does so (as in the common Shepherd's Purse), producing lateral flowers one after another for many weeks.

206. A Corymb (Fig. 202) is the same as a raceme, except that it is flat and broad, either convex, or level-topped. That is, a raceme becomes a corymb by lengthening the lower pedicels while the uppermost remain shorter. The axis of a corymb is short in proportion to the lower pedicels. By extreme shortening of the axis the corymb may be converted into

207. An Umbel (Fig. 203) as in the Milkweed, a sort of flower-cluster where the pedicels all spring apparently from the same point, from the top of the peduncle, so as to resemble, when spreading, the rays of an umbrella; whence the name. Here the pedicels are sometimes called the Rays of the umbel. And the bracts, when brought in this way into a cluster or circle, form what is called an Involucre.

208. The corymb and the umbel being more or less level-topped, bringing the flowers into a horizontal plane or a convex form, the ascending order of development appears as Centripetal. That is, the flowering proceeds from the margin or circumference regularly towards the centre; the lower flowers of the former answering to the outer ones of the latter.

209. In these three kinds of flower-clusters, the flowers are raised on conspicuous pedicels (204) or stalks of their own. The shortening of these pedicels, so as to render the flowers sessile or nearly so, converts a raceme into a Spike, and a corymb or an umbel into a Head.

210. A Spike is a flower cluster with a more or less lengthened axis, along which the flowers are sessile or nearly so; as in the Plantain (Fig. 204).

211. A Head (Capitulum) is a round or roundish cluster of flowers, which are sessile on a very short axis or receptacle, as in the Button-ball, Button-bush (Fig. 205), and Red Clover. It is just what a spike would become if its axis were shortened; or an umbel, if its pedicels were all shortened until the flowers became sessile. The head of the Button-bush is naked; but that of the Thistle, of the Dandelion, and the like, is surrounded by empty bracts, which form an Involucre. Two particular forms of the spike and the head have received particular names, namely, the Spadix and the Catkin.

212. A Spadix is a fleshy spike or head, with small and often imperfect flowers, as in the Calla, Indian Turnip, (Fig. 206), Sweet Flag, etc. It is commonly surrounded or embraced by a peculiar enveloping leaf, called a Spathe.

213. A Catkin, or Ament, is the name given to the scaly sort of spike of the Birch (Fig. 207) and Alder, the Willow and Poplar, and one sort of flower-clusters of the Oak, Hickory, and the like,—the so-called Amentaceous trees.

214. Compound flower-clusters of these kinds are not uncommon. When the stalks which in the simple umbel are the pedicels of single flowers themselves branch into an umbel, a Compound Umbel is formed. This is the inflorescence of Caraway (Fig. 208), Parsnip, and almost all of the great family of Umbelliferous (umbel-bearing) plants.