him, convicted him, not only of garbling and altering the text, but of deliberately appending spurious notes to many of the songs; he therefore stands condemned for all time by the text of the long-lost Glenriddel volume, recovered by Mr Dick, and sold by auction at Sotheby's on October 30th, 1903, for £610. Whatever may have been the grounds of Scott Douglas's judgment, the note which appears in the Reliques cannot be received as a fact to found upon till the missing leaf is forthcoming or its contents otherwise corroborated. We know for certain that the ending of the "Episode" must have been subsequent to 1782, for the evidence of the Bibles in the Alloway Monument is incontrovertible, and that is the date upon the title-page. It is equally certain that Highland Mary, whose surname was Campbell, died at Greenock, in the autumn of 1786, with the Bibles in her possession; hence the parting must have taken place, between 1782 and 1786, somewhere in the neighbourhood of Mauchline, with or without the negligible adjuncts of place and circumstance with which Cromek has invested it. The evidence at command is too meagre to justify any dogmatising on the precise year in which the incident occurred; and this applies with greater force to the beginning of the "Episode," regarding which there is no information on the record, save what may be ascertained from relative contemporary facts outside the narrative proper. Burns persistently refers the "Episode" to "his early years,” and neither editor nor writer on the subject has ever made any serious attempt at investigating what this vague expression really means. After a century and a quarter of intervening years, the beginning of the "Episode" is as much in the dark as ever. In all attempts to dispel it, we are confronted with two outstanding facts, vouched for by a weight and unanimity of evidence which put them beyond doubt, but whose chronological sequence and exact dates puzzle us at the very outset. Highland Mary, unless there has been a lying conspiracy from the beginning, undoubtedly acted in the capacity of a servant

at Coilsfield, and also in Gavin Hamilton's house in Mauchline, at different periods of her sojourn in Ayrshire -whether in the capacity of "dairymaid" in the one and "nursemaid" in the other is immaterial. The points to be determined are, which of them was the prior engagement, and when were they entered upon? Writer after writer places her in Gavin Hamilton's house in the summer of 1785, for no other apparent reason than the birth of Alexander, his second son, which took place on July 13th of that year; a fact which very likely originated the "nursemaid" tradition in order to square matters and give a seeming of reality to a mere guess. The written record. neither corroborates nor contradicts the surmise. A minute of the Kirk-Session of Mauchline, dated January 20th, 1785, informs us that James Wylie, Sorn; Agnes Cameron, New Street; Mary Vallance, Cumnock; Flora Weir, Woodend ; and Janet Caldwell, Maybole, "late servants to Mr Hamilton," are to be called as witnesses in the case then pending between the Session and Mr Hamilton; and Hugh Sterling, Helen Herris, and Jean Rennie are named as his “present servants"; from which it would appear that Mr Hamilton employed a groom, or orra man," and two servant maids, on a half-yearly engagement, as was then, and still is, the custom in Ayrshire. Sterling must have been succeeded, in May, 1786, by the lad from Mossgiel, whom Maister Tootie" had endeavoured to wheedle into his service. The girls, Herris and Rennie, may have left Mr Hamilton's service at Whitsunday, 1785; while the periods covered by the service of the other four may reasonably be supposed to take us back to Martinmas, 1783. Between that date, therefore, and Whitsunday, 1785, Mary Campbell could not possibly have been in the service of Gavin Hamilton. If she entered his service at Whitsunday, 1785, the "pretty long tract of reciprocal attachment " narrows itself down to a twelvemonth; if in July of that year, it dwindles to ten monthsproceeding on the implied assumption of so many writers on the subject that she never served at Coilsfield at all.

66

Is there any justification of this assumption and its further demand on our credulity that Mary Campbell was in Gavin Hamilton's service in the autumn of 1785 and spring of 1786? Almost every incident in the "Episode" is so befogged with rumour, gossip, and vain imaginings that the explorer very soon finds himself lost in an almost impenetrable jungle of conjectures, contradictions, and bald assertions, for never was the “bletherin' b--h" of all sorts and conditions so characteristically and constantly in evidence as in the terra incognita of Highland Mary. Thrusting the undergrowth aside, it is vain to deny the tradition of "Montgomerie Castle" and its connection with the story of Highland Mary. Mrs Begg assured Robert Chambers, in 1850, on the authority of her mother, that about the time of his desertion by Jean Armour, Burns "became acquainted with Mary Campbell, who was acting as nursemaid in the family of Gavin Hamilton, which situation she left to become dairymaid at Coilsfield. I said he just then became acquainted with her, but he must have known her previous to that, though his love-fit had only begun then." We have already endeavoured to show that Mary's term of service with Mr Hamilton was either previous to Martinmas, 1783, or subsequent to Whitsunday, 1785. If Mrs Begg is correct in saying that she left Mr Hamilton's service to become dairymaid at Coilsfield, she must have been there, making fullest allowance, during the half-year ending Whitsunday, 1786, the alleged date of the farewell interview, thus reducing the period of her residence in Mauchline to six, or, alternatively, to four months..

*

The following testimony by Andrew Smith, the founder of the Boxmaking industry in Mauchline, who was born. in 1797, and died in 1869, was communicated to me by the late James Wilson, Bank Agent, Sanquhar, some years

*The surmise has been ventilated that Burns might have become acquainted with her in Irvine, a part of Dundonald Parish being within the Burgh Boundaries. Burns left Irvine at the end of 1781; the Dundonald Session case was begun in April, 1784,

by a pursuer resident in Mauchline.

before his death, who must have procured it from Mr Smith at a date subsequent to 1858, the year in which the Mrs Todd referred to, who was Gavin Hamilton's daughter, Wilhelmina, departed this life. "Next to

Jean Armour," he says, Burns has bestowed the finest effusions of his Muse upon a Highland girl of the name of Mary Campbell, and we think we are in possession of some information regarding this simple maiden not generally known to the world. The Poet first became acquainted with her when she was a domestic servant in the family of his first friend and patron, Gavin Hamilton; she left this family and entered the service of General Montgomerie, then living at Coilsfield, on which Burns has bestowed the more euphonious name of the Castle of Montgomerie.' About this time, Burns's intercourse with Jean Armour had become of such a binding nature as to lead to a discontinuance of his love affair with Mary Campbell. When, however, James Armour, the father of his betrothed, compelled his daughter to destroy the written evidence of her marriage with the Poet, the courtship with Highland Mary appears for a time to have been renewed, and it was during the period of this second courtship that these interviews took place which furnished him with the themes for the very finest of his love songs. Mary Campbell was not what could be called a beauty, but the late Mrs Todd, who remembered her when she was a servant in her father's family, said she was a very pleasing and winning girl.' A thorn bush near the front of Coilsfield House has been pointed out as the spot where the parting interview between the two lovers took place; this is sheer nonsense. The Poet describes the meeting as having taken place in the day-time, and therefore it was not likely to have happened on the lawn in front of a gentleman's house. Burns says it was on the banks of the Ayr. The scene near to the confluence of the Fail with the Ayr.... is, without a doubt, the place where this memorable interview took place. At the time when this parting did take place, Mary Campbell was going to see

her friends in the Highlands; she soon afterwards returned to Greenock, took trouble, and died there; and now over her revered remains a fine tombstone has been erected

by public subscription." In an accompanying note Mr Wilson states that Andrew Smith was, during the whole of his life, on intimate terms with the Armours, if not, indeed, a blood relation of that family. Mr Wilson further states that he was a man of position and exceptional ability, whose word was his bond in all his business and social relations.

Robert Chambers says that there is much obscurity about the situations and movements of Mary in Ayrshire; he also considers it probable that she acted as nurse to Gavin Hamilton's son in July, 1785; and he makes no reference whatever to her sojourn in Coilsfield.

John Blane, who was "gaudsman " to Burns at Mossgiel in 1785, at which date he must have been a very young man,* always asserted that he was in the service of Gavin Hamilton along with Highland Mary, which statement, if reliable, fixes the date of her engagement there possibly at Whitsunday, 1785, but certainly before Martinmas of that year.

The preposterous and discredited Grierson-Train Histoires Scandaleuses need only be referred to here as corroborative of the tradition of Highland Mary's association with Coilsfield.† John Richmond left Mauchline for Edinburgh towards the end of 1785; all that he personally knew of Mary Campbell, therefore, is necessarily of a date anterior to Martinmas of that year.

66

Let us now take a glance at the chronology of the course of events in which the "Episode" is said to have occurred. Burns took up residence in Mossgiel in March, 1784; and he wrote a letter to Thomas Orr from Mossgavil" the farm-name which appears upon the Bibleson November 11th of that year. In the Poet's letters of 1786, the name is always spelt "Mossgiel," from which

*He died sixty years afterwards.

† See Wallace's Chambers Edition, Vol. I., Appendix viii.