THE MURRAY OF TULLIBARDINE is one of a relatively few clan tartans with a design that can be dated with certainty to the period of the 1745 Rising. Many of the early references are connected with the Murrays which demonstrates a long association of the tartan with the family and/or Perthshire.

This tartan is unique in being shown in not just one, or even two, but five 18th-century portraits. Only one is of these is definitely of a Murray and it is slightly later than the others, all of which date to within approximately five years of the '45.

The claim by James Grant ‘That tartan called Tullibardine .. was adopted and worn by Charles, first Earl of Dunmore, second son of the first Marquis of Tullibardine..’ and that he ‘..in 1679 was Lieutenant-Colonel of the Royal Grey Dragoons..’ has been taken by later writers as evidence of the antiquity of the Tullibardine tartan. However, he appears to have misunderstood the information given by the Smiths (known to be the source for much of his work) who wrote "We found this very pretty pattern of Tartan in the market, but we can say nothing more anent it, than that the proprietors of the respectable Tartan Warehouse from which we obtained it assured us it is the Tartan used by the Earl of Dunmore". The respectable Tartan Warehouse mentioned by the Smiths was probably Wilsons of Bannockburn’s. A recently discovered sample book of their tartans circa 1830-40 includes a piece named simply Tullibardine meaning that we know that the leading manufacturer of the day from whom the Smiths obtained their specimens was selling this tartan.

Probably the first evidence of the pattern is to be found in Cosmo Alexander’s unidentified Portrait of a Jacobite Lady circa 1740-46 which is one of a small number of pre-Proscription pictures to show tartan wore by women. For a long time it was assumed that the sitter was wearing an unidentified red tartan but recent research by the author proved that she is in fact wearing the Tullibardine tartan. 

Roughly contemporary with the Jacobite Lady are three portraits by the famous Scottish artist Allan Ramsay: John Campbell, 4th Earl of Loudoun; Norman MacLeod of MacLeod, 22nd Chief; and Flora MacDonald. This is where matters become confused with various claims concerning the origins and ownership of the plaid worn by the sitters. It is known that Ramsay painted the composition, face, and hands and that he sub-contracted the costume painting to another artist Joseph Van Aken. He died in July 1749 so the costume in all three portraits must have been completed by then if the costume detail was by Van Aken. 

Lord Loudoun’s portrait was probably painted in 1747 when he was in Edinburgh before sailing to Flanders in late June. He is referred to as having been painted ‘in his Regimentals’ which must have referred to his scarlet jacket and insignia of rank, but not the tartan. Although the regiment wore plaids we know from orders and samples of the time that his regimental plaids were a blue, green, and black type tartan very different from the one in his portrait. In his fulsome paper on the portrait Ruairidh MacLeod identified the tartan as Tullibardine and postulated two possible sources: Loudon’s mistress, Anne Farquarson, wife of MacKintosh of MacKintosh, is said to have woven Loudoun a plaid ‘which became his “favourite” garment. Alternatively, his cousin, Lady Margaret MacDonald, wife of Alexander MacDonald of Sleat, wove him a plaid in 1746. In neither case do we know what the tartan was and, although it is tempting to conclude that he must have been wearing his ‘favourite’ plaid, we just do not know.

There is also the very real possibility that this is neither of the plaids but the ‘brighter’ sett that Loudoun supposedly sought, possibly from the Crieff weavers, in 1747. There is a poor quality portrait of one of his officers, Lt. Reid, wearing a red tartan waistcoat and it is possible that Loudoun had a plaid in the same pattern.

The following year Ramsay painted Norman MacLeod of MacLeod, 22nd Chief. Again Van Aken painted the clothing; MacLeod’s coat and trews are of a red and black check now commonly called Rob Roy but once again the plaid is of Tullibardine tartan. Coincidence? Stewart iv suggested that MacLeod wore the tartan to denote the connection with his grandmother, a Murray. Elsewhere it is claimedv that the plaid was from an 11 yard bolt of fabric ordered by MacLeod from Skye in 1747 and that the same fabric was also used by Ramsay/Van Aken for the portraits the Earl of Loudoun and of Sir Francis Charteris with his sister. No evidence is offered to support this claim. We know that Loudoun was painted a year earlier and that Charteris, 7th Earl of Wemyss, wears a similar style suit of Rob Roy with a matching plaid not one of Tullibardine.

Finally, the third of Ramsay’s portraits, that of Flora MacDonald, was painted in London and signed by him Ramsay pinxit anno 1749. The first owner, and presumably commissioner, of the portrait was Dr. Richard Mead, physician to King George II. Why Mead would want a portrait of Flora MacDonald is unclear but in 1749 she was something of a popular figure whose actions in assisting Charles Edward Stuart were already beginning to be romanticised. She had no close connections with the Murrays yet the detail of the Tullibardine tartan is very clear in the portrait.

Some may argue that someone such as Loudoun or MacLeod would not have wanted to be painted in someone else’s tartan but that presupposes its use as a form of clan symbol, something which we know did not develop until the early 19th-century. Very little is known about the use of tartans in portraits beyond the fact that in combination with the clothes, it denoted political (Highland or Scottish), and social status, particularly the use of large amounts of red. The likelihood that all three sitters owned a length of the same tartan is not credible and given Ramsay/Van Aken’s use of the same tartan in three portraits it seems more likely that they were working from a single source. Possibly Loudoun, as the first of the three sitters, owned a plaid whose design was used a default tartan or perhaps a length had been acquired by Ramsay or Van Aken to be used as a studio prop. More important in terms of this paper is the fact that all three confirm the existence of the Tullibardine tartan in the mid-18th-century.

The last of the portraits in which the sitter wears the Tullibardine is that of John Murray 4th Earl of Dunmore by Sir Joshua Reynolds in which Murray wears a coat and waistcoat of Tullibardine tartan and belted plaid of 42nd (Black Watch) tartan. This is the first positive evidence of the use of the tartan by a member of the clan with which it is generally associated. Although painted in 1770 the style of the jacket is earlier, circa 1750-60, so itis possible that Murray wore a jacket that he had had for some time.

In every case examined the tartan has been painted with such clarity as to allow the details to be extracted and confirmed as Tullibardine. If one accepts the hypothesis that the plaids in the Ramsay portraits were one and the same then there is evidence of at least three pieces of the tartan in use during the 18th-century and that this use was not exclusively by Murrays with whom it is now commonly associated.

Having demonstrated the early use of the tartan in portraiture what about actual specimens of the period? The oldest known named reference to the tartan is a piece in a sample book of Wilsons of Bannockburn’s patterns circa 1830-40 although older unnamed specimens survive. Blair Castle, clan seat of the Murrays, has a set of old bed hangings made from a length of Tullibardine tartan circa 1800. The amount of material and, the offset nature of the pattern and use of a herringbone selvedge are indicative of the cloth having been a copy of an older, probably early-mid 18th-century, plaid rather than being the original. Whilst examining a number of old pieces at Blair the author discovered a fragment of tartan, a portion of a plaid that dates to the first half of the 18th-century, which was probably the original piece from which the bed hangings were copied. The structure of the Tullibardine tartan is considered in a companion paper.

In conclusion, the Tullibardine tartan can be dated with certainty to the mid-18th century when it appeared in a number of portraits. Based on the available evidence I believe that the Ramsay portraits used the same plaid as the source for all three portraits in which it appears. Whether Loudoun owned such a plaid or whether Ramsay/Van Aken obtained a length to use as a studio prop is unclear. The use of the Tullibardine tartan in the Reynolds’ portrait of John Murray, together with the early samples at Blair, make a strong case for its traditional use by the family and/or being a pattern associated with that part of Perthshire. The sett and colours in the coat and the old sample are very different suggesting that the family had at least two lengths of the material during the 18th century which furthers support to this proposition. Irrespective of its origins, this is one of the very few tartans worn today that can be shown to date to the era of the clan system and to have been used continuously to the present. © Peter Eslea MacDonald Nov 2010.

THE AIM of this paper is to examine some old specimens of the tartan now known as Murray of Tullibardine, why it’s associated with the Murrays and to consider what may be the original setting and why it differs from the version seen today. 

In another paper, I noted that this is one of a relatively few clan tartans whose design can be dated with certainty to the period of the ’45 with at least four portraits of the time and one slightly later all showing the subject wearing this tartan. The painted representations are detailed and the sett quite clear although the colour of some of the finer lines is open to interpretation. The significance of this is discussed in the later. There is no surviving 18th-century reference to what the tartan was called and it was not until the mid-19th-century that written references list this amongst other clan tartans.

The design was shown by James Grant where his statement ‘That tartan called Tullibardine is a red tartan, and was adopted and worn by Charles, first Earl of Dunmore, second son of the first Marquis of Tullibardine..’ was taken by later writers as evidence that the Tullibardine tartan was in existence by the late 1600s. Grant took his information from the Smithsii Early Specimens who wrote ‘We found this very pretty pattern of Tartan in the market, but we can say nothing more anent it, than that the proprietors of the respectable Tartan Warehouse from which we obtained it assured us it is the Tartan used by the Earl of Dunmore. The Smiths called their pattern Murray of Tullibardine and went on to mention ‘...the present Earl.....’ making it clear that they were not referring to Charles, the 1st Earl.

The tartan is not included in any of the early 19th-century collections of tartan specimens but three examples at Blair Castle and a later one from Wilsons of Bannockburn show the development of the design and how it became the Murray of Tullibardine.

For a long time the oldest surviving piece was what was thought to a set of tartan hangings on what is known as the Tullibardine Bed at Blair Castle. The hangings are made from a single run of hard tartan approximately 28 yards long by 27” wide. The bed is not the original, which was a huge round one that was moved to Blair circa 1829 and the hangings are said to be a copy of the older ones on the original bed. The replacement tartan dates to circa 1800 but what is unknown is whether it was on the bed when it was moved to Blair circa 1829 or was already owned by the Atholls. The tartan was woven off-set with a broad black herringbone selvedge and an unusual colouration giving an overall red, blue and black appearance that lead to some researchers mistakenly classifying it as a different tartan altogether. However, recent research by the author confirmed that the broad dark stripes together with their flanking dark stripes are in fact dark green and thus the sett conforms to the standard setting. This is confirmed by a later length of material at Blair in which the setting, including the herringbone selvage, are matched but in much brighter shades. This circa 1820-40 hard tartan length ) is wholly consistent with the quality and colours of the major tartan weavers of the time, Wilsons of Bannockburn.

Having shown that the bed hangings and the later copy follow the standard colouration and setting what can be determined about the way in which the cloth was set to the loom? As the 1820-40 copy matches the hangings I shall deal simply with them as the older. The material is 27” wide and off-set (meaning the pattern is not balanced i.e. repeating evenly from the centre outwards) with a black selvedge mark. It will be recalled that the hangings are a copy of an earlier set and the above details indicate that the original was intended to be joined to make double width plaid. Such off-set designs with selvedge patterns were common in the early 18th-century and it seems likely that an original plaid was later used as hangings or a bed throw, perhaps during the Proscription era when plaids could not openly worn.

There is however a problem with the circa 1800 replacement material used for the hangings. It doesn’t follow the standard practice of traditional off-set plaids where the pattern finishes in the middle of a pivot on the side to be joined, in this case it should be the narrow black stripe, meaning that the sett will repeat normally until the selvedge mark at the top and bottom of the cloth. Here the design continues through the green pivot, the next red and finally into the blue stripe with the result that when the cloth is joined there is a triple blue stripe that throws out the normal repeat and creates a false pivot . There are a number of potential reasons why the cloth was woven like this:

• It is possible that the original plaid was a different width and the circa 1800 weavers simply fitted the design to the width that they had available.

• The original plaid/bed hangings were incomplete and so the reconstruction was a best guess.

• The original plaid was also set like this.

Given the selvedge pattern the first possibility seems unlikely. If the original plaid survived it would be a simple matter to amend the count to ensure that the design fitted the width available.

Perhaps then the original plaid was incomplete with the joining selvedge missing and so the weaver had to fit the design as best as they could. Although this is a fairly busy design there is more than a full repeat of the sett which makes the pivot points very obvious and it’s inconceivable that a weaver would make such an error unless he was only working from a scrap and had to guess the second pivot. This is a possibility but the inclusion of the selvedge mark means that there must have been a reasonable amount of the original to work with.

Finally, there is the possibility that the original plaid was woven this way too. That would be very unusual but not unique. The original MacDonald of Glenaladale plaid is similarly off-set but for some inexplicable reason not finished on a pivot at the joining selvedge.

But for a chance discovery we might be left to ponder the conundrum. A few years ago I examined a number of tartan fragments in the Blair Castle attic amongst which was a 18th-century fragment 26.5 x 1.75 inches from selvage to torn edge. The pattern is clearly the Tullibardine with the exception that the usual fine black stripes are green although their narrowness makes them appear black. Interestingly, this piece is also off-set with the joining edge not at the pivot and the frayed edge stopping short of what should be the selvedge mark. Comparison of the full width of the fragment and the bed hangings shows the striking similarity between the two with the pivots marked at X and the joining edge on the right.

It is unfortunate that the fragment stops short of the selvedge leaving us to speculate as to the original width and selvedge arrangement. There is no evidence of herringboning at the ragged edge so we have no way of knowing how it was finished or what width the original was. If it was the same width as the hangings then there would have been one additional red bar and that that would not have been herringboned due to its narrow width. Alternatively, the older warp may have had the same selvedge arrangement as the hangings in which case the cloth would have been about 33” wide, untypical but unknown for early 18th-century tartan material.

So far all the pieces examined and examples in early portraits have been unnamed. It has already been mentioned that the tartan was not included in early collections nor was it listed in Wilsons’ 1819 Key Pattern Book, however, a recently discovered circa 1830-40 sample book of Wilsons’ tartans. Conclusions includes a piece named simply Tullibardine. Although speculation, it is reasonable to conclude that Wilsons were asked to copy the tartan from the Tullibardine Bed at some point in the early 1800s and, always looking for genuinely old Highland patterns, they later sold it under that name. The respectable Tartan Warehouse mentioned by the Smiths was probably Wilsons’. Presumably they made the connection between the Wilsons’ name and the family and added the family name Murray for their 1850 publication.

This paper and the related one on the Early Use of the Tullibardine tartan have shown beyond doubt that the pattern can be dated to circa1746 with certainty and that the oldest surviving examples of cloth, and one of the portraits, are connected with the Murrays. A logical explanation has been offered for the naming of the pattern as Tullibardine and later Murray of Tullibardine although the evidence for its historic use by that branch is circumstantial at best whereas the Blair Castle and Dunmore associations indicate a pattern more likely to be connected with Atholl and/or Highland Perthshire in some way.

The fragment and bed hangings at Blair appear structurally related. In particular, the way in which they are off-set and the use of the unusual blue shade seem beyond coincidence. Possibly the hangings were an incorrect copy, either by design or omission, or there was more than one original piece with different settings. The 4th Earl of Dunmore’s jacket in the circa 1770 portrait is unlikely to have been made from an off-set length so there were presumably other weavings of the material available in the late 18th-century. With the colouring of the fine stripes in the early portraits by no means clear it is possible that the assumed black was in fact blue and that Blair fragment represents the original setting where the simplified red, blue and green setting is visually more pleasing.

We will probably never know the origins of the Tullibardine tartan before circa 1746. It is logical to assume that if it existed then, then it existed before but for how long is unclear. What is certain is that with so many portraits and old specimens depicting the design it has, with the exception of the simpler red and black Rob Roy type check, the unique position of being the most widely recorded and accurately documented of our original clan era tartans. © Peter Eslea MacDonald Nov 2010.

THIS FINE unidentified Portrait of a Jacobite Lady by Cosmo Alexander circa 1740-46 is one of a small number of pre-Proscription pictures to show tartan wore by women. Images of this one are generally of poor quality and for a long time researchers assumed that the sitter was wearing an unidentified red tartan. The original portrait, part of the Drambuie Collection, is on display at the NTS Culloden exhibition where I had an opportunity to examine it and working from a good quality photograph was able to analyse the sett and identify it as Murray of Tullibardine. The extract below shows a portion of the picture with a colour stripe superimposed and the sett matched to the tartan of the jacket. As can be seen it is possible to make out every line of the pattern which is truly remarkable for such a complex design. Having identified the tartan are we able to hazard a guess at the identification of the sitter? She appears to be aged roughly 18-35 and the style and fineness of the jacket indicates a lady/family of wealth. The white rose represents a Jacobite supporter and was painted circa 1740-50 with the period 1745-46 being most likely. She has been identified as Jenny Cameron, daughter of Cameron of Locheil and alleged mistress of Bonnie Prince Charlie but the evidence is by no means definitive. Whilst we cannot be certain who the sitter was, nor whether she was a Murray, there are several possible contenders amongst this strongly Jacobite clan: Lady Mary Murray (1720-95) daughter of the 1st Duke of Atholl; Lady Jane Murray (1730-47) daughter of the 2nd Duke of Atholl; Lady Charlotte Murray (1731-1805) daughter of the 2nd Duke of Atholl; Amelia Murray (1710-66), Lord George Murray’s wife. Amelia Murray (1732-1777) daughter of Lord George Murray. Irrespective of the sitter’s identity it is rewarding to be able to identify a tartan with such an historic pedigree. © Peter Eslea MacDonald Nov 2010.