Donald’s sister, Elyne (the spelling of the name used for Eleyne’s grand-daughter), married Sir John Menteith. Their mother Christian Bruce survived her imprisonment and married for the third time, Andrew Murray of Bothwell. She died about 1357 and was buried in her chapel of the Blessed Virgin of the Garioch.

Although King Robert and Queen Elizabeth had a son, who inherited his father’s kingdom in 1329 as King David II, he died without issue, and it was Eleyne’s great-grandson, the son of Marjorie Bruce and Walter the Steward of Scotland, who next inherited the throne as King Robert II, the first of the Stewart line, in fulfilment of Einion’s prophecy.

AUTHOR’S NOTE

The story of Eleyne of Mar is the result of a pilgrimage deep into the remote archives of my family, and is a part of a legend with which I grew up. This legend told of romance and excitement in Scotland in the time of Robert the Bruce and Isobel of Buchan, which I first wrote about in Kingdom of Shadows. It was while researching the historical sections of that novel that I realised how closely Eleyne (or Helen or Ellen), the great- grandmother of Isobel and the mother of Isabella, Robert’s first wife, was bound up with their story and I began to wonder what kind of a woman she was. It was the start of a quest which turned out to be convoluted and full of irreconcilable puzzles and which led in the end to the writing of this novel.

Like any good detective I began my research into Helen’s life with the part I already knew, or thought I knew, when she was the Countess of Mar. Thirteenth-century Scotland is fairly well documented. There are chronicles, there are records, there is the wonderful narrative poem ‘The Bruce’, written by John Barbour, the Archdeacon of Aberdeen, only seventy or so years after the siege of Kildrummy. I felt it would be easy to find out about ‘Helen’ and her world.

There she was, in the Peerage: ‘Donald, Earl of Mar m. Helen, widow of Malcolm, Earl of Fife (who d. 1266), and da. of Llywelyn, Prince of North Wales. She was living in Feb. 1294/5.’ The entry under Fife confirms her name. It was those words ‘daughter of Llywelyn’ which intrigued me. How had this daughter of a Welsh prince ended up the great-grandmother of a Scots king?

I was to discover, however, that there are very few mentions of her extant in original records. One of the few is to be found in the Pipe Roll of King Edward I where we have, in an account by Walter de Cambo of the Issues of Lands and Tenements belonging to Duncan, Earl of Fife: ‘Et Elenae, comitissae de Mar, pro parte dotis suae xl s. per idem tempus’, an entry which is repeated a few months later, and lists the ‘pension’ which followed Ellen/Helen into her last marriage. This entry was brief, but appeared to establish her existence beyond any doubt. It was a start. But how had this daughter of Wales reached Scotland in the first place?

I turned to Welsh history. There were two Llywelyns who might be called ‘Prince of North Wales’, although the term was not strictly used for either of them. Llywelyn ap Iorwerth, or Llywelyn the Great, was Prince of Aberffraw and Lord of Snowdon. He was the ruler of Gwynedd and so could be said to have been Prince of North Wales. Then there was his grandson, Llywelyn ap Gruffydd, who in the latter part of his rule was to call himself Prince of Wales. The former appeared to be the most likely candidate for Helen’s father and there indeed in all the textbooks and family trees was the information that Llywelyn ap Iorwerth and his wife, Joan, had a daughter called Ellen or Helen. Her brother and two of her sisters were married to de Braoses, a fact which I registered with a shock of recognition as I had become so familiar with that family while researching Lady of Hay. Ellen/Helen herself was married to John the Scot, Earl of Huntingdon and Chester. This seemed at once to establish a Scots connection, albeit a somewhat tenuous one. I read on.

Ellen/Helen married John in 1222. But my Scots Ellen/Helen married the Earl of Mar sometime after 1266 and went on to bear him five children. If she was the same person the dates did not fit unless she married John of Huntingdon as a baby. There is a great deal of information available about this wedding between the daughter of Prince Llywelyn and the heir to the great and powerful earldom of Chester. We know where and when, we know some of the gifts, we know who the witnesses were. But nowhere does it say the bride was a baby or even a child or that the wedding was by proxy. By now doubting that this could after all be my Ellen/ Helen I read on about the Countess of Huntingdon, looking for clues. There were for instance no children by this marriage. That would fit if most of it was spent growing up. If she was a small child in 1222, at her husband’s death in 1237 she would have been in her teens. Of course there could be many reasons for her childlessness, such as his ill health – he was a comparatively young man when he died. (It was later that I was to find the intriguing information that at the time she was suspected of having procured his death by poison.)

The records chart Ellen/Helen’s removal to Chester Castle, where she was to be kept in honourable and fitting state until Henry decided what to do with her, and her swift remarriage to Robert de Quincy. We read from the Dunstable annalist the remark about Llywelyn’s indignation at the haste of the wedding and at his new son-in-law’s low rank. We learn of her two daughters and can determine their ages. We know about Joanna’s marriage, we lose track of Hawisa while still a minor shortly after her father’s death. We know that Robert de Quincy took the Cross (but not if he actually went to the Holy Land) and we know when he died.

I assumed that at this point, if Ellen/Helen was my Ellen/Helen, she would now have remarried for the third time, on this occasion to Malcolm, Earl of Fife. There would have been plenty of time then for her to have had two sons by him before he died in 1266.

But no.

Following the records further, I found that in 1253 the story of the Countess of Huntingdon and Chester abruptly ended when her dower lands were redistributed amongst her heirs. The obvious inference was that she was dead and I must admit I was very disappointed. I had become fascinated with my Welsh Ellen/Helen and now I was left, so it seemed, with two Ellen/Helens and an unbridgeable gap between them.

Where to go from there?

Families in medieval times commonly and confusingly had siblings with the same Christian name. Did Llywelyn ap Iorwerth have two daughters called Ellen? Back came an unequivocal answer from North Wales, no, although at this point some information emerged about an interesting variation on Welsh Ellen’s birth. Two sources, one collected in the sixteenth or seventeenth century, state that she was not Joan’s daughter, as all the modern history books said, but the eldest daughter of Tangwystl, Llywelyn’s mistress or first wife. (Margaret and Gwladus were also, according to this source, daughters of Tangwystl.) This record does agree however that she married John the Scot. Intriguing, but the authorities who claim the three girls as Joan’s daughters seem to outnumber those who uphold that they were Tangwystl’s.

At this point I took a step back to take a more oblique look at Welsh history. There were at least a dozen Llywelyns, lesser princelings and lords, extant at the relevant period. Could my Scots Ellen/Helen be the daughter of one of them, and had he become transmogrified over the centuries into ‘the’ Llywelyn by her descendants?

I put this idea on hold whilst considering the other major contender for the title of Llywelyn, Prince of North Wales, Llywelyn ap Gruffydd. I had at first discounted him because, as far as the majority of history books and records are concerned, he had only one child, the unfortunate Gwenllian who died without issue. But…

In a collection of traditional Welsh pedigrees in the College of Arms in London, consulted for me by Peter Gwynn-Jones, the Lancaster Herald, there are at least two which contain a record of Llywelyn’s marriage to Eleanor, daughter of Simon de Montfort, and go on to speak of their only child, Catherine Lackland, who married first Philip ap Ifor and second MALCOLM, EARL OF FIFE!

Catherine?

This was a shock. And it couldn’t be right! Llywelyn married Eleanor in 1278, twelve years after Malcolm died! Perhaps Catherine was a bastard daughter of Llywelyn ap Gruffydd, conceived in his youth? (That might explain her name of Lackland.) Works based on pedigrees in the National Library of Wales do not mention her at all, speaking only of Gwenllian.

I was by now feeling very confused. Obviously there had been a marriage between some member of the house of Gwynedd and Malcolm, Earl of Fife. At the Scots end, the lady’s name was believed to have been Ellen/Helen; at the Welsh end, the only actual mention of Malcolm of Fife is linked with a putative daughter of Llywelyn ap Gruffydd, called Catherine.

It seemed to be a good time to return to the Scots records and once again to consult the experts, confronting them with Catherine. There is quite a lot of information in existence about the Mar family at that date. We know Donald was knighted rather late in life (but not why). We know who his children were and whom they married. Obviously their close relationship with the Bruces is well documented. We know Alexander was kept in the Tower (the cost of the Scots nobles and their retinues ‘staying’ in the Tower after the Battle of Dunbar was noted in Edward I’s account book as being £407 6s. ½d) and we know he is not heard of again. But sadly it proved impossible to learn any more than I already knew about Donald’s wife, namely that she was Malcolm’s widow, that her name was Ellen or Helen and that she was a daughter of a Llywelyn.