Who killed the Red Fox? What was the secret that the Celts would not communicate to Mr. R.L. Stevenson, when he was writing Kidnapped? Like William of Deloraine, 'I know but may not tell'; at least, I know all that the Celt knows. The great-grandfather and grandfather of a friend of mine were with James Stewart of the Glens, the victim of Hanoverian injustice, in a potato field, near the road from Ballachulish Ferry to Appin, when they heard a horse galloping at a break-neck pace. 'Whoever the rider is,' said poor James, 'he is not riding his own horse.' The galloper shouted, 'Glenure has been shot!'

'Well,' said James to his companion, 'whoever did it, I am the man that will hang for it.'

Hanged he was. The pit in which his gibbet stood is on the crest of a circular 'knowe,' or hummock, on the east side of the Ballachulish Hotel, overlooking the ferry across the narrows, where the tide runs like a great swift river.

I have had the secret from two sources; the secret which I may not tell. One informant received it from his brother, who, when he came to man's estate, was taken apart by his uncle. 'You are old enough to know now,' said that kinsman, 'and I tell you that it may not be forgotten.' The gist of the secret is merely what one might gather from the report of the trial, that though Allan Breck was concerned in the murder of Campbell of Glenure, he was not alone in it.

The truth is, according to tradition, that as Glenure rode on the fatal day from Fort William to his home in Appin, the way was lined with marksmen of the Camerons of Lochaber, lurking with their guns among the brushwood and behind the rocks. But their hearts failed them, no trigger was drawn, and when Glenure landed on the Appin side of the Ballachulish Ferry, he said, 'I am safe now that I am out of my mother's country,' his mother having been of clan Cameron. But he had to reckon with the man with the gun, who was lurking in the wood of Letter More ('the great hanging coppice'), about three-quarters of a mile on the Appin side of Ballachulish Ferry. The gun was not one of the two dilapidated pieces shown at the trial of James of the Glens, nor, I am told, was it the Fasnacloich gun. The real homicidal gun was found some years ago in a hollow tree. People remember these things well in Appin and Glencoe, though the affair is a hundred and fifty years old, and though there are daily steamers bringing the newspapers. There is even a railway, not remarkable for speed, while tourists, English, French, and American, are for ever passing to view Glencoe, and to write their names in the hotel book after luncheon, then flying to other scenes. There has even been a strike of long duration at the Ballachulish Quarries, and Labour leaders have perorated to the Celts; but Gaelic is still spoken, second sight is nearly as common as short sight, you may really hear the fairy music if you bend your ear, on a still day, to the grass of the fairy knowe. Only two generations back a fairy boy lived in a now ruinous house, noted in the story of the Massacre of Glencoe, beside the brawling river: and a woman, stolen by the fairies, returned for an hour to her husband, who became very unpopular, as he neglected the means for her rescue; I think he failed to throw a dirk over her shoulder. Every now and then mysterious lights may be seen, even by the Sassenach, speeding down the road to Callart on the opposite side of the narrow sea-loch, ascending the hill, and running down into the salt water. The causes of these lights, and of the lights on the burial isle of St. Mun, in the middle of the sea strait, remain a mystery. Thus the country is still a country of prehistoric beliefs and of fairly accurate traditions. For example, at the trial of James Stewart for the murder of Glenure, one MacColl gave damaging evidence, the MacColls being a sept subordinate to the MacIans or Macdonalds of Glencoe, who, by the way, had no hand in the murder. Till recently these MacColls were still disliked for the part played by the witness, and were named 'King George's MacColls.'

map

[Enlarge]

But we must come to the case of Allan Breck. To understand it, some knowledge of topography is necessary. Leaving Oban by steamer, you keep on the inside of the long narrow island of Lismore, and reach the narrow sea inlet of Loch Creran on your right. The steamer does not enter it, but, taking a launch or a boat, you go down Loch Creran. On your left is the peninsula of Appin; its famous green hills occupy the space bounded by Loch Creran on the south and Glencoe on the north. Landing near the head of Loch Creran, a walk of two miles takes you to the old house of Fasnacloich, where Allan Breck was wont to stay. Till two or three years ago it belonged to the Stewarts of Fasnacloich, cadets of the chief, the Laird of Appin; all Appin was a Stewart country and loyal to the King over the Water, their kinsman. About a mile from Fasnacloich, further inland, is the rather gloomy house of Glenure, the property of Campbell of Glenure, the Red Fox who was shot on the road under Letter More. Walking across the peninsula to Appin House, you pass Acharn in Duror, the farm of James Stewart of the Glens, himself an illegitimate kinsman of the Laird of Appin. To the best of my memory the cottage is still standing, and has a new roof of corrugated iron. It is an ordinary Highland cottage, and Allan, when he stayed with James, his kinsman and guardian, slept in the barn. Appin House is a large plain country house, close to the sea. Further north-east, the house of Ardshiel, standing high above the sea, is visible from the steamer going to Fort William. At Ardshiel, Rob Roy fought a sword and target duel with the laird, and Ardshiel led the Stewarts in the rising of 1745; Appin, the chief, held aloof. The next place of importance is Ballachulish House, also an old house of Stewart of Ballachulish. It is on the right hand of the road from Ballachulish Pier to Glencoe, beneath a steep wooded hill, down which runs the burn where Allan Breck was fishing on the morning of the day of Glenure's murder, done at a point on the road three-quarters of a mile to the south-west of Ballachulish House, where Allan had slept on the previous night. From the house the road passes on the south side of the salt Loch Leven (not Queen Mary's Loch Leven). Here is Ballachulish Ferry, crossing to Lochaber. Following the road you come opposite the House of Carnoch, then possessed by Macdonalds (the house has been pulled down; there is a good recent ghost story about that business), and the road now enters Glencoe. On high hills, well to the left of the road and above Loch Leven, are Corrynakeigh and Coalisnacoan (the Ferry of the Dogs), overtopping the narrows of Loch Leven. Just opposite the House of Carnoch, on the Cameron side of Loch Leven, is the House of Callart (Mrs. Cameron Lucy's). Here and at Carnoch, as at Fasnacloich, Acharn, and Ballachulish, Allan Breck was much at home among his cousins.

From Loch Leven north to Fort William, with its English garrison, all is a Cameron country. Campbell of Glenure was an outpost of Whiggery and Campbells, in a land of loyal Stewarts, Camerons, and Macdonalds or MacIans of Glencoe. Of the Camerons, the gentle Lochiel had died in France; his son, a boy, was abroad; the interests of the clan were represented by Cameron of Fassifern, Lochiel's uncle, living a few miles west by north of Fort William. Fassifern, a well-educated man and a burgess of Glasgow, had not been out with Prince Charles, but (for reasons into which I would rather not enter) was not well trusted by Government. Ardshiel, also, was in exile, and his tenants, under James Stewart of the Glens, loyally paid rent to him, as well as to the commissioners of his forfeited estates. The country was seething with feuds among the Camerons themselves, due to the plundering by ——, of ——, of the treasure left by Prince Charles in the hands of Cluny. The state of affairs was such that the English commander in Fort William declared that, if known, it 'would shock even Lochaber consciences.' 'A great ox hath trodden on my tongue' as to this business. Despite the robbery of Prince Charles's gold, deep poverty prevailed.

In February, 1749, Campbell of Glenure had been appointed Factor for Government over the forfeited estates of Ardshiel (previously managed by James Stewart of the Glens), of Lochiel, and of Callart. In the summer of 1751, Glenure evicted James from a farm, and in April, 1752, took measures for evicting other farmers on Ardshiel estates. Such measures were almost unheard of in the country, and had, years before, caused some agrarian outrages among Gordons and Camerons; these were appeased by the King over the Water, James VIII. and III. James Stewart, in April, 1752, went to Edinburgh, and obtained a legal sist, or suspension of the evictions, against Glenure, which was withdrawn on Glenure's application, who came home from Edinburgh, and intended to turn the tenants out on May 15, 1752. They were assailed merely as of Jacobite name and tendencies. Meanwhile Allan Breck—who had deserted the Hanoverian army after Prestonpans, had joined Prince Charles, fought at Culloden, escaped to France, and entered the French army—was lodging about Appin among his cousins, perhaps doing a little recruiting for King Louis. He was a tall thin man, marked with smallpox.

Cruising about the country also was another Jacobite soldier, 'the Sergent More,' a Cameron, later betrayed by ——, of ——, who robbed the Prince's hoard of gold. But the Sergeant More had nothing to do, as has been fancied, with the murder of Glenure. The state of the country was ticklish; Prince Charles expected to invade with Swedish forces, under the famous Marshal Keith, by the connivance of Frederick the Great, and he had sent Lochgarry, with Dr. Archibald Cameron and others, to feel the pulse of the western clans. As Government knew all about these intrigues from Pickle the Spy, they were evicting Jacobite tenants from Ardshiel's lands, and meant to do the same, by agency of Campbell of Glenure, in Lochaber, Lochiel's country.

On Monday, May 11, Campbell, who intended to do the evictions on May 15, left Glenure for Fort William, on business; the distance is computed at sixteen miles, by the old hill road. Allan Breck, on the 11th, was staying at Fasnacloich, near Glenure, where the fishing is very good. When Glenure moved north to Fort William, Allan went to James Stewart's cottage of Acharn. Glenure's move was talked of, and that evening Allan changed his own blue coat, scarlet vest, and black velvet breeches for a dark short coat with silver buttons, a blue bonnet, and trousers (the Highlanders had been diskilted), all belonging to James Stewart. He usually did make these changes when residing with friends. In these clothes next day (Tuesday, May 12) Allan, with young Fasnacloich, walked to Carnoch, the house of Macdonald of Glencoe, situated just where the Water of Coe or Cona enters Loch Leven. The dowager of the house was natural sister of James of the Glens, and full sister of the exiled Stewart of Ardshiel. From Carnoch, Allan, on the same day, crossed the sea-strait to Callart opposite, where Mrs. Cameron was another half-sister to James of the Glens. On Wednesday Allan recrossed, called at Carnoch, and went to stay at Ballachulish House. On Thursday, when Glenure would certainly return home by Ballachulish Ferry, Allan, about mid-day, was seen to go fishing up Ballachulish burn, where he caught no trout, and I do not wonder at it.

The theory of the prosecution was that, from the high ground to the left of the burn he watched the ferry, having one or two guns, though how he got them unobserved to the place is the difficulty; he could not have walked the roads from Acharn unobserved with a gun, for the Highlanders had been disarmed. At this point he must have had the assistance and the gun of the other man. Allan came down from the hill, asked the ferryman if Glenure had crossed, and returned to his point of observation. About five o'clock in the afternoon, Glenure, with a nephew of his, Mungo Campbell, a 'writer' or solicitor, crossed the ferry, and was greeted and accompanied for three-quarters of a mile on his homeward way by old Stewart of Ballachulish, who turned back and went to his house. A sheriff's officer walked ahead of Glenure, who, like Mungo, was mounted. Behind both, mounted, was Campbell's servant, John Mackenzie. The old road was (and is) a rough track, through thick coppice. There came a shot, and Glenure, pierced by two balls, fell and died.

John Mackenzie, Glenure's servant, now rode onwards at a great gallop to find Campbell of Ballieveolan, and on his way came to Acharn and met James Stewart, with the two ancestors of my friend, as already described. He gave the news to James, who 'wrung his hands and expressed great concern at what had happened, as what might bring innocent people to trouble.' In fact, he had once, or oftener, when drinking, expressed a desire to have a shot at Glenure, and so had Allan. But James was a worthy, sensible man when sober, and must have known that, while he could not frighten the commissioners of forfeited estates by shooting their agent, he was certain to be suspected if their agent was shot. As a matter of fact, as we shall see, he had taken active steps to secure the presence of a Fort William solicitor at the evictions on Friday, May 15, to put in a legal protest. But he thought it unadvisable to walk three or four miles and look after Glenure's corpse; the Highlanders, to this day, have a strong dread or dislike of corpses. That night James bade his people hide his arms, four swords, a long Spanish gun, and a shorter gun, neither of which weapons, in fact, did the trick, nor could be depended on not to miss fire.

Where, meanwhile, was Allan? In the dusk, above Ballachulish House, he was seen by Kate MacInnes, a maid of the house; they talked of the murder, and she told Donald Stewart, a very young man, son-in-law of Ballachulish, where Allan was out on the hillside. Donald Stewart averred that, on hearing from Kate that Allan wanted to see him (Kate denied that she said this), he went to the hill, accused Allan of the crime, and was told, in reply, that Allan was innocent, though, as a deserter from the Hanoverian army, and likely to be suspected, he must flee the country. Other talk passed, to which we shall return. At three in the morning of Friday, May 15, Allan knocked at the window of Carnoch House (Glencoe's), passed the news, was asked no questions, refused a drink and made for the sheiling, or summer hut, high on the hill side of Coalisnacoan, whence you look down on the narrows of Loch Leven.

There we leave Allan for the moment, merely remarking that he had no money, no means of making his escape. As he is supposed by the prosecution to have planned the slaying of Glenure with James Stewart on May 11, it seems plain that James would then have given him money to use in his escape, or, if he had no money by him, would have sent at once to Fort William or elsewhere to raise it. He did not do this, and neither at Carnoch, Callart, nor Ballachulish House did Allan receive any money.

But, on May 12, when Allan went to Carnoch and Callart, James sent a servant to a very old Mr. Stewart, father of Charles Stewart, notary public. The father was a notary also, and James, who wanted a man of law to be at the evictions on May 15, and thought that Charles Stewart was absent in Moidart, conceived that the old gentleman would serve the turn. But his messenger missed the venerable sportsman, who had gone a-fishing. Learning later that Charles had returned from Moidart, James, at 8 a.m. on May 14 (the day of the murder), sent a servant to Charles at Fort William, bidding him come to the evictions on May 15, 'as everything must go wrong without a person that can act, and that I can trust.' In a postscript he added, 'As I have no time to write to William (Stewart), let him send down immediately 8l. to pay for four milk cows I bought for his wife at Ardshiel.' His messenger had also orders to ask William Stewart for the money.

Nothing could seem more harmless, but the prosecution might have argued that this letter was, as to the coming of the notary, a 'blind,' and that the real object was, under the plea of sending for the notary, to send the messenger for William Stewart's 8l., destined to aid Allan in his escape.[8] There was no proof or even suggestion that, on May 12, James had asked old Mr. Stewart to send money for Allan's use, or had asked William Stewart, as having none by him he would have done—that is, if James had concerted the murder with Allan. If, on May 14, James was trying to raise money to help a man who, as he knew, would need it after committing a murder on that day, he showed strange want of foresight. He might not get the money, or might not be able to send it to Allan. In fact, that day James did not get the money. The prosecution argued that the money was sent for on May 14, to help Allan Breck, and did not even try to show that James had sent for money on May 12; when it would have arrived in good time. Indeed James did not, on May 12, send any message to William Stewart at Fort William, from whom, not from Charles or the old gentleman, he tried to raise the cash on May 14. A friendly or a just jury would have noted that if James planned a murder on the night of May 11, and had no money, his very first move, on May 12, would be to try to raise money for the assassin's escape. No mortal would put off that step till the morning of the crime; indeed, it is amazing that Allan, if he meant to do the deed, did not first try to obtain cash for his escape. The relations of Glenure suspected, at the time, that Allan was not the assassin, that he fled merely to draw suspicion away from the real criminal (as he does in Kidnapped), and they even wished to advertise a pardon for him, if he would come in and give evidence. These facts occur in a copious unpublished correspondence of the day between Glenure's brothers and kinsmen; Mr. Stevenson had never heard of these letters.[9] Thus, up to the day of the murder, Allan may not have contemplated it; he may have been induced, unprepared, to act as accessory to the other man.

The point where, according to the prosecution, the evidence 'pinched' James of the Glens was his attempt to raise money on May 14. What could he want with so large a sum as 8l., so suddenly, as he had no bill to meet? Well, as a number of his friends were to be thrown out of their farms, with their cattle, next day, James might need money for their relief, and it seems certain that he had made no effort to raise money at the moment when he inevitably must have done so, if guilty, that is, on May 12, immediately after concerting, as was alleged, the plot with Allan Breck. Failing to get money from William Stewart at Fort William on May 14, James did on May 15 procure a small sum from him or his wife, and did send what he could scrape together to Allan Breck at Coalisnacoan. This did not necessarily imply guilt on James's part. Allan, whether guilty or not, was in danger as a suspected man and a deserter; James was his father's friend, had been his guardian, and so, in honour, was bound to help him.

But how did he know where Allan was to be found? If both were guilty they would have arranged, on May 11, a place where Allan might lurk. If they did arrange that, both were guilty. But Donald Stewart, who went, as we have said, and saw Allan on the hillside on the night of the murder, added to his evidence that Allan had then told him to tell James of the Glens where he might be found, that is, at Coalisnacoan. These tidings Donald gave to James on the morning of May 15. James then sent a pedlar, Allan's cousin, back to William Stewart, got 3l., added, in the evening of the 16th, more money of his own, and sent it to Allan. There was a slight discrepancy between the story of the maid, Kate MacInnes, and that of Donald Stewart, as to what exactly passed between them, concerning Allan, on the night of the murder, and whether Allan did or did not give her a definite message to Donald. The prosecution insisted on this discrepancy, which really, as James's advocate told the jury, rather went to prove their want of collusion in the manufacture of testimony. Had their memories been absolutely coincident, we might suspect collusion—that they had been 'coached' in their parts. But a discrepancy of absolutely no importance rather suggests independent and honest testimony. If this be so, Allan and James had arranged no trysting-place on May 11, as they must have done if Allan was to murder Glenure, and James was to send him money for his escape.

But there was a discrepancy of evidence as to the hour when the pedlar sent by James to Fort William on May 15 arrived there. Was he despatched after the hour when Donald Stewart swore that he gave Allan's message to James of the Glens, or earlier, with no knowledge on James's part of the message carried by Donald? We really cannot expect certainty of memory, after five months, as to hours of the clock. Also James did not prove that he sent a message to Allan at Coalisnacoan, bidding him draw on William Stewart for money; yet on Friday, May 15, James did, by the pedlar, bid William Stewart give Allan credit, and on Saturday, May 16, Allan did make a pen from a bird's feather, and ink with powder and water, and write a letter for money, on the strength of James's credit, to William Stewart. This is certainly a difficulty for James, since he suggested John Breck MacColl, a tenant of Appin's at Coalisnacoan, for the intermediary between Allan and William Stewart, and Allan actually did employ this man to carry his letter. But Allan knew this tenant well, as did James, and there was nobody else at that desolate spot, Coalisnacoan, whom Allan could employ. So lonely is the country that a few years ago a gentleman of my acquaintance, climbing a rocky cliff, found the bones of a man gnawed by foxes and eagles; a man who never had been missed or inquired after. Remains of pencils and leather shoe strings among the bones proved that the man had been a pedlar, like James Stewart's messenger, who had fallen over the precipice in trying to cross from Coalisnacoan to the road through Glencoe. But he never was missed, nor is the date of his death known to this day.

The evidence of the lonely tenant at Coalisnacoan, as to his interviews with Allan, is familiar to readers of Kidnapped. The tenant had heard of the murder before he saw Allan. Two poor women, who came up from Glencoe, told the story, saying that 'two men were seen going from the spot where Glenure was killed, and that Allan Breck was one of them.' Thus early does the mysterious figure of the other man haunt the evidence. The tenant's testimony was not regarded as trustworthy by the Stewart party; it tended to prove that Allan expected a change of clothes and money to be sent to him, and he also wrote the letter (with a wood-pigeon's quill, and powder and water) to William Stewart, asking for money. But Allan might do all this relying on his own message sent by Donald Stewart, on the night of the murder, to James of the Glens, and knowing, as he must have done, that William Stewart was James's agent in his large financial operations.

On the whole, then, the evidence, even where it 'pinches' James most, is by no means conclusive proof that on May 11 he had planned the murder with Allan. If so, he must have begun to try to raise money before the very day of the murder. James and his son were arrested on May 16, and taken to Fort William; scores of other persons were arrested, and the Campbells, to avenge Glenure, made the most minute examinations of hundreds of people. Meanwhile Allan, having got 5l. and his French clothes by the agency of his cousin the pedlar, decamped from Coalisnacoan in the night, and marched across country to the house of an uncle in Rannoch. Thence he escaped to France, where he was seen in Paris by an informant of Sir Walter Scott's in the dawn of the French Revolution; a tall, thin, quiet old man, wearing the cross of St. Louis, and looking on at a revolutionary procession.

The activities of the Campbells are narrated in their numerous unpublished letters. We learn from a nephew of Glenure's that he had been 'several days ago forewarned,' by whom we cannot guess; tradition tells, as I have said, that he feared danger only in Lochiel's country, Lochaber, and thought himself safe in Appin. The warning, then, probably came from a Cameron in Lochaber, not from a Stewart in Appin. In coincidence with this is a dark anonymous blackmailing letter to Fassifern, as if he had urged the writer to do the deed:

'You will remember what you proposed on the night that Culchena was buried, betwixt the hill and Culchena. I cannot deny but that I had breathing' (a whisper), 'and not only that, but proposal of the same to myself to do. Therefore you must excuse me, when it comes to the push, for telling the thing that happened betwixt you and me that night.... If you do not take this to heart, you may let it go as you will.' (June 6, 1752.)

Fassifern, who had no hand in the murder, 'let it go,' and probably handed the blackmailer's letter over to the Campbells. Later, ——, —— of ——, the blackest villain in the country, offered to the Government to accuse Fassifern of the murder. The writer of the anonymous letter to Fassifern is styled 'Blarmachfildich,' or 'Blarmackfildoch,' in the correspondence. I think he was a Mr. Millar, employed by Fassifern to agitate against Glenure.

In the beginning of July a man, suspected of being Allan, was arrested at Annan on the Border, by a sergeant of the Royal Welsh Fusiliers. He really seems to have changed clothes with Allan; at least he wore gay French clothes like Allan's, but he was not that hero. Young Ballachulish, at this time, knew that Allan was already across the sea. Various guesses occur as to who the other man was; for example, a son of James of the Glens was suspected, so there was another man.

The 'precognitions,' or private examinations of witnesses before the trial, extended to more than seven hundred persons. It was matter of complaint by the Stewart party that 'James Drummond's name appeared in the list of witnesses;' this is Mr. Stevenson's James More, really MacGregor, the son of Rob Roy, and father of Catriona, later Mrs. David Balfour of Shaws, in Kidnapped and Catriona. 'James More's character is reflected upon, and I believe he cannot be called worse than he deserves,' says one of the Campbells. He alleges, however, that in April, before the murder, James of the Glens visited James More, then a prisoner in Edinburgh Castle, 'caressed him,' and had a private conversation with him. The abject James More averred that, in this conversation, James of the Glens proposed that James More's brother, Robin Oig, should kill Glenure for money. James More was not examined at the trial of James of the Glens, perhaps because he had already escaped, thanks to Catriona and collusion; but his evidence appears to have reached the jury, almost all of them Campbells, who sat at Inveraray, the Duke of Argyll on the bench, and made no difficulty about finding James of the Glens 'Guilty.' To be sure, James, if guilty, was guilty as an accessory to Allan, and that Allan was guilty was not proved; he was not even before the court. It was not proved that the bullets which slew Glenure fitted the bore of James's small gun with which Allan was alleged to have perpetrated the murder, but it was proved that the lock of that gun had only one fault—it missed fire four times out of five, and, when the gun did not miss fire, it did not carry straight—missed a blackcock, sitting! That gun was not the gun used in the murder.

The jury had the case for James of the Glens most clearly and convincingly placed before them, in the speech of Mr. Brown for the accused. He made, indeed, the very points on which I have insisted; for example, that if James concerted a murder with Allan on May 11, he would not begin to hunt for money for Allan's escape so late as May 14, the day of the murder. Again, he proved that, without any information from James, Allan would naturally send for money to William Stewart, James's usual source of supply; while at Coalisnacoan there was no man to go as messenger except the tenant, John Breck MacColl. A few women composed his family, and, as John MacColl had been the servant of James of the Glens, he was well known already to Allan. In brief, there was literally no proof of concert, and had the case been heard in Edinburgh, not in the heart of the Campbell country, by a jury of Campbells, a verdict of 'Not Guilty' would have been given: probably the jury would not even have fallen back upon 'Not Proven.' But, moved by clan hatred and political hatred, the jury, on September 24, found a verdict against James of the Glens, who, in a touching brief speech, solemnly asserted his innocence before God, and chiefly regretted 'that after ages should think me guilty of such a horrid and barbarous murder.'

He was duly hanged, and left hanging, on the little knoll above the sea ferry, close to the Ballachulish Hotel.

And the other man?

Tradition avers that, on the day of the execution, he wished to give himself up to justice, though his kinsmen told him that he could not save James, and would merely share his fate; but, nevertheless, he struggled so violently that his people mastered and bound him with ropes, and laid him in a room still existing. Finally, it is said that strange noises and knockings are still heard in that place, a mysterious survival of strong human passions attested in other cases, as on the supposed site of the murder of James I. of Scotland in Perth.

Do I believe in this identification of the other man? I have marked every trace of him in the documents, published or unpublished, and I remain in doubt. But if Allan had an accessory in the crime, who was seen at the place, an accomplice who, for example, supplied the gun, perhaps fired the shot, while Allan fled to distract suspicion, that accessory was probably the person named by legend. Though he was certainly under suspicion, so were scores of other people. The crime does not seem to me to have been the result of a conspiracy in Appin, but the act of one hot-headed man or of two hot-headed men. I hope I have kept the Celtic secret, and I defy anyone to discover the other man by aid of this narrative.

That James would have been quite safe with an Edinburgh jury was proved by the almost contemporary case of the murder of the English sergeant Davies. He was shot on the hillside, and the evidence against the assassins was quite strong enough to convict them. But some of the Highland witnesses averred that the phantasm of the sergeant had appeared to them, and given information against the criminals, and though there was testimony independent of the ghost's, his interference threw ridicule over the affair. Moreover the Edinburgh jury was in sympathy with Mr. Lockhart, the Jacobite advocate who defended the accused. Though undeniably guilty, they were acquitted: much more would James of the Glens have obtained a favourable verdict. He was practically murdered under forms of law, and what was thought of the Duke of Argyll's conduct on the bench is familiar to readers of Kidnapped. I have never seen a copy of the pamphlet put forth after the hanging by the Stewart party, and only know it through a reply in the Campbell MSS.

The tragedy remains as fresh in the memories of the people of Appin and Lochaber as if it were an affair of yesterday. The reason is that the crime of cowardly assassination was very rare indeed among the Highlanders. Their traditions were favourable to driving 'creaghs' of cattle, and to clan raids and onfalls, but in the wildest regions the traveller was far more safe than on Hounslow or Bagshot Heaths, and shooting from behind a wall was regarded as dastardly.