him. He'll never do such a thing again, I fancy. Where is Merton?"

They drew off and went to help their comrade, who lay senseless.

and clang of hoofs. "Good God!" he exclaimed, and forgetting everything else, sprung from the bed and ran to the window outside his door. The light of their lanterns dimly showed a confused crowd in the yard of the mews, and amid the hellish uproar of their coarse voices he could hear Kelpie plunging and kicking. Again she uttered the same ringing scream. He threw the window open and cried to her that he was coming, but the noise was far too great for his enfeebled voice. Hurriedly he added a garment or two to his half-dress, rushed to the stair, passing his new friend, who watched anx-self could have been his witness how she iously at the head of it, without seeing her, and shot from the house.

CHAPTER L.

THE DEMONESS AT BAY.

When Malcolm would have led Kelpie in, she stopped suddenly at the stabledoor, and started back shuddering as if the memory of what she had endured there overcame her. Every fibre of her trembled. He saw that she must have been pitifully used before she broke loose and got out. But she yielded to his coaxing, and he led her to her stall without difficulty. He wished Lady Clementina her

knew her friend and trusted him. Had she seen how the poor bleeding thing rejoiced over him, she could not have doubted that his treatment had been in part at least a success.

Kelpie had many enemies amongst the men of the mews. Merton had gone out for the evening, and they had taken the opportunity of getting into her stable and tormenting her. At length she broke her fastenings: they fled, and she rushed out after them.

They carried the maimed man to the hospital, where his leg was immediately amputated.

WHEN he reached the yard of the mews the uproar had nothing abated. But when he cried out to Kelpie, through it all came a whinny of appeal, instantly followed by a scream. When he got up to the lanterns he found a group of wrathful men with stable-forks surrounding the poor animal, from whom the blood was streaming before and behind. Fierce as she was, she dared not move, but stood Malcolm washed and dried his poor antrembling, with the sweat of terror pour-imal, handling her as gently as possible, ing from her. Yet her eye showed that for she was in a sad plight. It was plain not even terror had cowed her. She was he must not have her here any longer: but biding her time. Her master's first worse to her at least was sure to follow. impulse was to scatter the men right and He went up, trembling himself now, to left, but on second thoughts, of which he Mrs. Merton. She told him she was just was even then capable, he saw that they running to fetch him when he arrived: might have been driven to apparent bru- she had no idea how ill he was. But he tality in defence of their lives, and be- felt all the better for the excitement, and sides, he could not tell what Kelpie might after he had taken a cup of strong tea do if suddenly released. So he caught wrote to Mr. Soutar to provide men on her by the broken halter and told them to whom he could depend if possible the fall back. They did so, carefully it same who had taken her there before — seemed unwillingly. But the mare had to await Kelpie's arrival at Aberdeen. eyes and ears only for her master. What There he must also find suitable housing she had never done before, she nosed and attention for her at any expense until him over face and shoulders, trembling all further directions, or until, more probably, the time. Suddenly one of her torment- he should claim her himself. He added ors darted forward and gave her a terrible many instructions to be given as to her prod in the off hind-quarter. But he treatment. paid dearly for it. Ere he could draw back she lashed out and shot him half across the yard with his knee-joint broken. The whole set of them rushed at her. "Leave her alone," shouted Malcolm, "or I will take her part. Between us we'll do for a dozen of you."

"The devil's in her," said one of them. "You'll find more of him in that rascal groaning yonder. You had better see to

Until Merton returned he kept watch, then went back to the chamber of his torture, which, like Kelpie, he shuddered to enter. The cook let him in and gave him his candle, but hardly had he closed his door when a tap came to it, and there stood Rose, his preserver. He could not help feeling embarrassed when he saw her. I see you don't trust me," she said. "I do trust you," he answered. "Will

you bring me some water? I dare not drink anything that has been standing."

She looked at him with inquiring eyes, nodded her head and went. When she returned he drank the water.

"There! you see I trust you," he said with a laugh. "But there are people about who for certain reasons want to get rid of me: will you be on my side?"

"That I will," she answered eagerly. "I have not got my plans laid yet; but will you meet me somewhere near this tomorrow night? I shall not be at home, perhaps, all day.”

whistle had ceased ringing in Malcolm's own ears. He left him with his horse, went on board and gave various directions to Travers; then took Davy with him, and bought many things at different shops, which he ordered to be delivered to Davy when he should call for them. Having next instructed him to get everything on board as soon as possible, and appointed to meet him at the same place and hour he had arranged with Rose, he went home.

A little anxious lest Florimel might have wanted him, for it was now past the hour She stared at him with great eyes, but at which he usually waited her orders, agreed at once, and they appointed time he learned to his relief that she was gone and place. He then bade her good-night, shopping with Lady Bellair, upon which and the moment she left him lay down on he set out for the hospital whither they the bed to think. But he did not trouble had carried the man Kelpie had so terribly himself yet to unravel the plot against mauled. He went, not merely led by him, or determine whether the violence sympathy, but urged by a suspicion also he had suffered had the same origin with which he desired to verify or remove. On the poisoning. Nor was the question the plea of identification he was permitted merely how to continue to serve his sister to look at him for a moment, but not to without danger to his life; for he had just speak to him. It was enough: he recoglearned what rendered it absolutely im-nized him at once as the same whose secperative that she should be removed from her present position. Mrs. Merton had told him that Lady Lossie was about to accompany Lady Bellair and Lord Liftore to the Continent. That must not be, whatever means might be necessary to prevent it. Before he went to sleep things had cleared themselves up considerably.

He woke much better, and rose at his usual hour. Kelpie rejoiced him by affording little other sign of the cruelty she had suffered than the angry twitching of her skin when hand or brush approached a wound. The worst fear was that some few white hairs might by-and-by in consequence fleck her spotless black. Having urgently committed her to Merton's care, he mounted Honor and rode to the Aberdeen wharf. There, to his relief, time growing precious, he learned that the same smack in which Kelpie had come was to sail the next morning for Aberdeen. He arranged at once for her passage, and saw, before he left, to every contrivance he could think of for her safety and comfort. He warned the crew concerning her temper, but at the same time prejudiced them in her favor by the argument of a few sovereigns. He then rode to the Chelsea Reach, where the Psyche had now grown to be a feature of the river in the eyes of the dwellers upon its banks.

At his whistle Davy tumbled into the dinghy like a round ball over the gunwale, and was rowing for the shore ere his

ond attack he had foiled in the Regent's Park. He remembered having seen him about the stable, but had never spoken to him. Giving the nurse a sovereign and Mr. Soutar's address, he requested her to let that gentleman know as soon as it was possible to conjecture the time of his leaving. Returning, he gave Merton a hint to keep his eye on the man, and some money to spend for him as he judged best. He then took Kelpie for an airing. To his surprise, she fatigued him so much that when he had put her up again he was glad to go and lie down.

When it came near the time for meeting Rose and Davy he got his things together in the old carpet-bag, which held all he cared for, and carried it with him. As he drew near the spot, he saw Davy already there, keeping a sharp lookout on all sides. Presently Rose appeared, but drew back when she saw Davy. Malcolm went to her. "Rose," he said, "I am going to ask you to do me a great favor. But you cannot except you are able to trust me."

"I do trust you," she answered.

"All I can tell you now is that you must go with that boy to-morrow. Before night you shall know more. Will you do it?"

"I will," answered Rose. "I dearly love a secret."

"I promise to let you understand it if you do just as I tell you." "I will."

"Be at this very spot, then, to-morrow

visited by insects. Thus, in some of our colonies, the common red clover sets no seeds on account of the absence of humblebees; for the proboscis of the hive bee is not long enough to effect the object. According to Mr. Belt, the same is the case, for the same reason, in Nicaragua, with "the scarlet-runner. But even in those instances in which it is not absolutely necessary, it is a great advantage that the flowers should be fertilized by pollen brought from a different plant, and, with this object in view, insects are tempted to visit flowers for the sake of the honey and pollen; while the colors and scents are useful in making the flowers more easy to find.

of insects are of importance to the plant. | do our ants perform for them a very similar Sprengel was the first to point out the real function, by keeping down the number of office of honey, but his views were far small insects, which would otherwise rob from meeting with general consent, and them of their sap and strip them of their even so lately as 1833 were altogether leaves. rejected by Kurr, who came to the conclu- Forel watched, from this point of view, sion that the secretion of honey is the a nest of Formica pratensis. He found result of developmental energy, which that the ants brought in dead insects, afterwards concentrates itself on the ovary. small caterpillars, grasshoppers, cercopis, No doubt, however, seems any longer etc., at the rate of about twenty-eight a to exist that Sprengel's view is right; and minute, or more than one thousand six that the true function of honey is to attract hundred in an hour. When it is coninsects and thus to secure cross-fertiliza- sidered that the ants work not only all tion. Thus most of the Rosacea are fer- day, but in warm weather often all night tilized by insects and possess nectaries; too, it is easy to see how important a funcbut, as Delpino has pointed out, the genus tion they fulfil in keeping down the numPoterium is anemophilous, or wind-fertil-bers of small insects. ized, and possesses no honey. So also Some of the most mischievous insects, the maples are almost all fertilized by indeed certain species, for instance, of insects and produce honey; but Acer aphis and coccus - have turned the tables

the little brown garden ant, for instance, assiduously running up the stems of plants, to milk their curious little cattle. By this ingenious idea, not only do the aphides and cocci secure immunity from the attacks of the ants, but even turn them from foes into friends. They are subject to the attacks of a species of ichneumon, which lays its eggs in them, and Delpino has seen the ants watching over them with truly maternal vigilance, and driving off the ichneumons whenever they attempted to approach.

negundo is anemophilous and honeyless. on the plants, and converted the ants Again among the Polygonacea some from enemies into friends, by themselves species are insect-fertilized and mellifer- developing nectaries, and secreting honey, ous, while on the other hand certain which the ants love. We have all seen genera, Rumex and Oxyria, have no honey, and are fertilized by the wind. At first sight it might appear an objection to this view, and one reason perhaps why the earlier botanists missed the true explanation may have been the fact, that some plants secrete honey on other parts besides the flowers. Belt and Delpino have, I think, suggested the true function of these extra-floral nectaries. The former of these excellent observers describes a South American species of acacia, which, if unprotected, is apt to be stripped of the leaves by a leaf-cutting ant, which uses the leaves, not directly for food, but, according to Mr. Belt, to grow mushrooms on. The acacia, however, bears hollow thorns, and each leaflet produces honey in a crater-formed gland at the base, and a small, sweet, pear-shaped body at the tip. In consequence, it is inhabited by myriads of a small ant, Pseudomyrma bicolor, which nests in the hollow thorns, and thus finds meat, drink, and lodging all provided for it. These ants are continually roaming over the plant, and constitute a most efficient body-guard, not only driving off the leaf-cutting ants, but, in Belt's opinion, rendering the leaves less liable to be eaten by herbivorous mammalia. Delpino mentions that on one occasion he was gathering a flower of Clerodendron fragrans, when he was suddenly attacked by a whole army of small

ants.

I am not aware that any of our English plants are protected in this manner from the browsing quadrupeds, but not the less

But though ants are in some respects very useful to plants, they are not wanted in the flowers. The great object is to secure cross-fertilization; but for this purpose winged insects are almost necessary, because they fly readily from one plant to another, and generally confine themselves for a certain time to the same species. Creeping insects, on the other hand, naturally would pass from each floret to the next; and, as Mr. Darwin has shown in his last work, it is of little use to bring pollen from a different flower of the same plant; it must be from a different plant altogether. Moreover, creeping insects when they quitted a plant would naturally creep up another close by, without any regard to species. Hence, even to small flowers (such as many cruciferæ, compositæ, saxifrages, etc.), which, as far as size is concerned, might well be fertilized by ants, the visits of flying insects are much more advantageous. Moreover, if larger flowers were visited by ants, not only would they deprive the flowers of their

honey without fulfilling any useful function | corolla. The nectar is not protected by in return, but they would probably prevent any special arrangement of the flower the really useful visits of bees. If you itself, and is accessible even to very small touch an ant with a needle or a bristle, insects. The stamens ripen before the she is almost sure to seize it in her jaws; pistil, and any flying insect, however and if bees when visiting any particular small, coming from above, would assist in plant were liable to have the delicate tip cross-fertilization. Creeping insects, on of their proboscis seized on by the horny the contrary, which in most cases would jaws of an ant, we may be sure that such enter from below, would rob the honey a plant would soon cease to be visited. without benefiting the plant. P. amphi bium, as its name denotes, grows sometimes in water, sometimes on land. So long, of course, as it grows in water, it is thoroughly protected, and the stem is smooth; while, on the other hand, those specimens which live on land throw out certain hairs which terminate in sticky glands, and thus prevent small insects from creeping up to the flowers. In this case, therefore, the plant is not sticky, except just when this condition is useful.

On the other hand, we know how fond ants are of honey, and how zealously and unremittingly they search for food. How is it, then, that they do not anticipate the bees and secure the honey for themselves? Kerner has recently published a most interesting memoir on this subject, and pointed out a number of ingenious contrivances by which flowers protect themselves from the unwelcome visits of such intruders. The most frequent are the interposition of chevaux de frise, which ants cannot penetrate, glutinous parts which they cannot traverse, slippery slopes which they cannot climb, or barriers which close the way.

All these viscous plants, as far as I know, have upright or horizontal flowers. On the other hand, where the same object is effected by slippery surfaces, the flowers are often pendulous; creeping creatures being thus kept out of them, just as the pendulous nests of the weaver bird are a protection from snakes and other enemies. As instances of this kind, I may mention the common snowdrop, or the cyclamen.

Firstly, then, as regards chevaux de frise. In some respects these are the most effectual protection, since they exclude not only creeping insects, but also other creatures, such as slugs. With this object, it will be observed that the hairs I have elsewhere suggested that the sowhich cover the stalks of so many herbs called "sleep" of flowers had reference to usually point downwards. A good exam- the habits of insects, on the ground that ple of this is afforded, for instance, by a flowers which are fertilized by night-flying plant allied to our common blue scabious, insects would derive no advantage by beKnautia dipsacifolia. The heads of the ing open in the day; while, on the other common carline (Carlina vulgaris), again, hand, those which are fertilized by bees present a sort of thicket, which must offer would gain nothing by being open at night. an impenetrable barrier to ants. Some I confess that I suggested this with much species of plants are quite smooth, ex- diffidence, but it may now, I think, be recepting just below the flowers. The com-garded as well established. mon but beautiful cornflower (Centaurea Silene nutans, the Nottingham catchfly, cyanus) is quite smooth, but the involu- is a very instructive species from this cres forming the flower-head are bordered with recurved teeth. In this case neither the stem nor the leaves show a trace of such prickles. In this species the stigma projects about one-fifth of an inch above the flower, so that if ants could obtain access they would steal the honey without fertilizing the flower; a flying insect, on the contrary, alighting on the flower, could scarcely fail to touch the stigma.

Another common mode of protection is by glutinous surfaces.

Kerner has called attention to a very interesting illustration afforded by Polygonum amphibium. The beautiful rosy flowers of this species are rich in nectar; the stamens are short; the pistil, on the contrary, projects considerably above the

point of view, and indeed illustrates a number of interesting points in the relations between plants and insects. Its life history has recently been well described by Kerner. The upper part of the flowering stem is viscid, from which it has derived its local name, the Nottingham catchfly. This prevents the access of ants and other small creeping insects. Each flower lasts three days, or rather three nights. The stamens are ten in number, arranged in two sets, the one set standing in trout of the sepals, the other in front of the petals. Like other night flowers, it is white, and opens towards evening, when it also becomes extremely fragrant. The first evening, towards dusk, the stamens in front of the sepals