Gathering Mosses

In his blog post Gatherings about Moss, Thomas Clancy follows up John MacQueen’s (1956, p. 141) suggestion that Old English mos or Old Norse mosi – cognate with Scots MOSS ‘(peat) bog’ – were borrowed into Gaelic. There are names in the Place-Names of the Galloway Glens database which appear to use this borrowed element *mos: Moss Roddock Loch & Moss Raploch. Names of the type elsewhere in Galloway include Moss Nae, Mossmaul and Mossfeather.

I’ve found three Moss- names which weren’t recorded on the 1st edition of the Ordnance Survey six-inch to the mile maps and which have therefore flown under the radar: Mossbowie, Kirkmaiden WIG; Moss Glairn, Lochrutton, KCB; and Mossband, Kirkgunzeon KCB.

Mossbowie, Kirkmaiden WIG

Mossbowie is recorded on John Ainslie’s (1782) A Map of the County of Wigton. This is the most securely Gaelic of the three Moss-names gathered here. It is hard to analyse it as anything other than Gaelic mos + buidhe ‘yellow’.

John Ainslie (1782) ‘A Map of the County of Wigton’
Reproduced with the permission of the National Library of Scotland

Moss Glairn, Lochrutton KCB

I think Moss Glairn, only recorded on William Mounsey’s (1815) A Plan of the Lochrutton Estate, is fairly likely to be another Gaelic Moss- name too. Here is what I wrote about it in my Estate Map Gleanings 3 post:

“The second element of Moss Glairn looks like it could be Gaelic gleadhran ‘wood sorrel; yellow rattle; cockscomb’. (Alan James has suggested that cockscomb here could be Pedicularis palustris.) If not named from a flower, Thomas Clancy suggests a possible derivation from gleadhar ‘noise, rattle’.

Another possibility is that Moss Glairn is a Scots name in inverted element order (see Drumcushat Park below). Moss is fine, but what would Glairn be? Bobby Clark suggests a derivation from GLAUR ‘soft, sticky mud; ooze, slime’. Alan James suggests glairn could be ‘an analogical adjectival form’ of GLAUR. ‘Slimy bog’ would be something of a tautology, but tautologous names are not uncommon. However, I think the balance is probably in favour of this being a Gaelic name.”

William Mounsey (1815) ‘A Plan of the Loch Rutton Estate comprending the property in Loch Rutton and Urr parishes in the Stewartry of Kirkcudbright belonging to Marmaduke Constable Maxwell’
Reproduced with the permission of the National Library of Scotland

Mossband, Kirkgunzeon KCB

Mossband appears on the map as Mossband Cottages in 1957 on OS sheet NX86NE. This site subsequently became Mossband Caravan Park and is now Mossband Residential Park. I imagine Mossband Cottages took their name from the field on which they were built, which sits beside a moss. Mossband could represent Gaelic mos + bàn ‘white’. Band would be an unremarkable reanalysis of bàn. A parallel can be found at Castle Ban (Gaelic caiseal bàn) in Kirkcolm, WIG which was recorded in the OS Name Books as Castle Band OS1/35/15/8.

Another option is that Mossband is Scots moss + band. There are hundreds of Scots/Scottish Standard English names where moss appears in first position, followed (in order of frequency) by elements such as side, end, head, plantation, wood, house, bank etc. Mossband feels like a natural Scots/Scottish Standard English name. However, it’s not clear what the element band would be doing here. In northern England, band is used in relatively young names in the sense ‘the ridge of a small hill’. [See this tweet for the reference; thanks to Paul Carbuncle for providing it.] This seems to be the meaning in The Band, Closeburn: “A name given to a rocky Brow [on] the farm of Townhead & South slope of Auchenleck Hill” OS1/10/6/50 (also described at OS1/10/6/53).

However, beyond The Band in Closeburn, band only appears as the final element in Mossband/Moss Band. In addition to the Kirkgunzeon Mossband, there are a handful of other names of this type across Scotland and northern England:

  • Mossband, Dumfries, DMF NX 996 752
  • Mossband, Kirkandrews Nether CMB NY 350 654
    • The Survey of English Place-Names records the following forms: Mosband, 1590; the Mosbande, 1602; Mosse Band near Carlisle 1677; two names (Mossband and Mosband) are marked on the south-west and south sides of Solway Flow, now Rosetrees Moss, on Roy’s Military Map (1752-55).
  • Mossband, Bothwell LAN NS 795 606.
    • Moss Band on Roy’s Military Map (1752-55).
  • Moss Bands, Mossband, Mid Moss Band, Airth STL are marked on Roy’s Military Map (1752-55) on the east side of what is now Leatham Moss. They aren’t recorded on the OS.

Paul Carbuncle notes that the 1602 form of the Mosbande (Kirkandrews Nether), which includes the definite article might pose a hurdle to a Gaelic etymology here. If *mosbàn had been reanalysed as Mossband by this time, I don’t think this is to much of an issue. It may be that a name interpreted as English would ‘invite’ the definite article (though I think this probably depends on the type of name as well as the grammatical context of the record.) However, it’s worth bearing in mind that not all these names need be Gaelic; they might represent a mix of Gaelic, Scots and (Scottish Standard) English names.

If (some of) these names are Scots/(Scottish Standard) English, the fact that band appears only to collocate with Moss- implies that they weren’t being coined as moss + band but that *mossband was a ready-made compound. What would that be? Band ‘the ridge of a small hill’ is unlikely to be relevant, as Mossbands appear to be located at low-lying, moss-side sites. Perhaps band here referred to a ‘strip of land’ or ‘land bounding/bordering something’.

It’s not easy to settle the matter, but I think the weight of probability lies with a Gaelic etymology: ‘white moss’ would be a perfectly ordinary name, and one with a parallel in Mossbowie ‘yellow moss’. A Scots/(Scottish Standard) English etymology would require either a formation with band in an otherwise unrecorded toponymic sense or the use of an equally unrecorded pre-existing compound *mossband. Not impossible, but the path of least resistance appears to be Gaelic.

NX86NE – A Surveyed / Revised: Pre-1930 to 1956, Published: 1957
Reproduced with the permission of the National Library of Scotland


  • MacQueen, John, 1956, ‘Kirk- and Kil- in Galloway Place-Names’, Archivum Linguisticum 8, pp. 135-149

Two Angelic Place-Names in Kirkpatrick Irongray

I recently got hold of a copy of EMH M’Kerlie’s (1916) Pilgrim Spots in Galloway. In the chapter on ‘Kirkpatrick Durham and Kirkpatrick-Irongray’ it mentions Angel Well (NX 8733 7770) and Angel Chapel (NX 8739 7764). Neither of these names are recorded on the 1st edition or subsequent Ordnance Survey maps (though they do get a mention in the Name Books). Both the well and chapel have CANMORE entries but as the the original sources aren’t that easy to access, I’ve copied them below.

‘Angel’ place-names are, unsurprisingly, not particularly common. The Saints in Scottish Place-Names database records eight: Cnoc nan Aingeal (North Uist); Angels’ Hill (Lochalsh); Tom Aingil (Kilmonivaig); Cnoc Aingil (Lismore); ?Tom nan Ainil (Balquhidder); Cnoc an Aingil (Glassary); Cnoc Aingil (Kildalton & Oa); Angel Hill (Kirkcudbright). In addition to these, the Ordnance Survey Name Books include Angels Burn (Aboyne & Glentanar: OS1/1/4/62, OS1/1/9/89).

All but two of these names are Gaelic and only one (Angels Burn) isn’t a hill name. In this context, Angel Well and Angel Chapel are notable not just for adding to the corpus of Scottish angel place-names but for being Scots/English names and for referring to features that aren’t hills.


E. Marianne H. M’Kerlie (1916) Pilgrim Spots in Galloway, London: Sands & Co., pp. 211-213

In the absence of all other tradition, who so [212] fitly as St Patrick may be connected with what was once the Angel Chapel and the Angel Well at Barnsoul? — he who lived in such familiar converse with his angel guardian.
Now hardly known, unmarked on any map, the spot is chiefly regarded as an ancient fort; an, were it not for the sweet-sounding title of the Angle Chapel, the interest would be chiefly for the archæologist or the lover of beautiful scenery. Bu the title is a magnet; and the road, at least from Dumfries, most beautiful all the way to the romantic “Routin Brig,” under which the old Water of Cluden tumbles over the rocks, in miniature cascades, to join the Cluden water; then the wild hills, of which one is Skeoch, as seen from Barnsoul, lying a little to the south.
On arriving here, the usual inquiries were necessary. The farmer’s wife knew nothing of the Angel Chapel and Well, but a bright little girl did, and her father knew more. He was the fetched, and conducted me tot he site. This is in a big, rich, sloping field, and on the June day of my visit it was a veritable carpet of buttercups and daisies. Six or eight Ayrshire calved tried hard to follow us into it; and the man, remarking on its richness, said that he had heard that it was blessed by the monks long ago. Then he pointed to a mound, with some old thorns on either side, and on going up into the grassy enclosure, it presented the form [213] of a horse-shoe, and this is the site of the Angel Chapel. A little below is a circle of rounded stones, — the basin of a spring. This was the Angel Well, whose waters have been diverted for the purpose of draining the land, and which now find their outlet in an adjoining burn.

F. Coles (1893) ‘The Motes, Forts, and Doons in the East and West Divisions of the Stewartry of Kirkcudbright’, Proceedings of the Society of Antiquaries of Scotland, 27, pp. 92-182 [at p.112, n.1]<>

Half a mile N.E., and much lower down, on Barnsoul, there is marked on the O.M. the site of a chapel. From Mr Welsh, proprietor of Macnaughton, I learned that there were records extant in his family bearing on this point. So far as may be judged by actual survey of the remains as they now are, the notion of an ecclesiastical or of any other rectangular walled building, indeed, would be the very last to be suggested. The site is a horseshoe-shaped flattish space, within what certainly seems to be nothing more or less than a rather unusually broad rampart of earth and stone — in parts quite 20 feet wide — and having interior diameters of 75 x 75 feet. wing, however, to ravages made by ploughing and sundry unequal parts which incline to the angular, and help to render this curious site incompletely curvilinear, I do not feel justified in assigning it a place in my survey. Mr Welsh avers that it was known as The Angel Chapel, and a spring of water in the hollow to the N. goes by the name of The Angel Well to this day.

Ordnance Survey Name Book: Chapel (site of) OS1/20/64/12 (1848)

The site of a chapel on the f[arm] of Drumcloyor house of Barnsoul, no remains of the […] can now be traced out, althoug[h] the spot is surrounded by an [em]bankment of earth and stones […] covered with grass and thorn but W. Alex. Welsh & his brother Jam[es] Weslh of Macnaughton whose ancestors have been residents in the locality for more than a c[en]tuary past, say that they have heard it as a tradition from their father & from others; that a chapel existed here at some remote peri[od.] The name Chapel Rig which applies to the eminence would in a deg[ree] confirm this tradition also sometime ago, a well near this place which [is] now closed up was called Angel Well. & which is evident had some connection with the chapel.

Scots Place-Names on the Map

The trend for Scots place-names on maps is towards anglicisation. This process accelerated and was largely fixed by the Ordnance Survey’s first edition six-inch to the mile maps. Comparing earlier and contemporary maps – and the Various Modes of Spelling entries in OS Name Books – to the names on the 1st edition shows a concerted shift from Scots to English forms: stone > stane; laigh > low; brig > bridge; co > cave; lang > long; water > river, etc.

The standardisation of Scots names in English spelling on maps and road signs hasn’t just visually altered the linguistic landscape, it’s changing the way Scots names are spoken. Unless you’re used to hearing Auld Brig o Urr, you’re going to follow Google, the OS and the road signs and call the place Old Bridge of Urr. There’s no way to read Colvend as Co’en unless you seen or heard the latter. And being local only gets you so far. I know the Scots pronunciations of the Kirkcudbrightshire names I grew up around but I gang agley with Wigtownshire names – mostly (I assume) without even knowing it – as I have to follow the spelling on the map or road signs. The process is self-sustaining, the more a name gets ‘airtime’ in English form, the less familiar the Scots form becomes. The path from unofficial, to unused, to lost is short.

If you think that Scots matters, it’s hard not to despair at this. You can’t change the OS and while digitally editing road signs is a diverting hobby, it’s not likely to effect any real change. Fortunately, OpenStreetMap offers a way to get Scots place-names back on the map.

River Bladnoch road sign, edited to show Water o Blaidnoch, using GIMP and and Transport Heavy font which “Contains public sector information licensed under the Open Government Licence v3.0”.

OpenStreetMap, a free, open source, editable map of the world – think Wikipedia for maps – has the facility for storing multilingual names for every place in its database. Scots names can be added to the map using the tag name:sco, just as you can add names in any other language. What this means in practice is that if you search for Auld Brig o Urr in OpenStreetMap it takes you to Old Bridge of Urr, unlike Google Maps which will take you to a Wetherspoons in Irvine. (You can also change your language preferences to display the user interface in Scots giving you Sairch instead of Search, Eedit instead of Edit, etc.)

OpenStreetMap showing Old Bridge of Urr when Auld Brig o Urr is searched for. Google Maps doesn’t do this.

This is a fantastic start, but although Scots names can be stored in OpenStreetMap so that they can be searched for, the names on the map appear in English by default [1]. What we need is a way of displaying the Scots names first, with the map falling back on English names where no Scots form is recorded. The good news is that this is completely achievable.

All that is needed is someone to host a tile-server which will display name:sco names first in the hierarchy.[2] Anyone with the hardware and know-how could do this (I don’t but Chris Fleming has kindly offered to help). In terms of hardware, the Scots Language Centre seems like the perfect place to host a Scots OpenStreetMap. They’ve already produced a map and a gazetteer of Scots place-names. A Scots OpenStreetMap would take this excellent work up several notches, with thousands of names – from major settlements to minor names – displayed in Scots, with the added bonus of all the interactive functionality OpenStreetMap has to offer.

I’m hopeful that this will get off the ground soon. In the meantime, there’s plenty of work to do adding Scots names to OpenStreetMap. The first thing to do if you’re interested is sign up for an account, after which it’s just a case of searching for a place > clicking ‘Edit’ > clicking the ‘+’ under ‘Name’ to enter a multilingual name, and you’re away. The Scots Language Centre’s gazetteer is a great first point of call for finding Scots names. For Galloway, Trotter’s Galloway Gossip and S.R. Crockett’s books (available as free pdfs at the S.R. Crockett Archive) are great sources. However, local pronunciations are probably going to get you the most coverage.[3]

I’ve put links to some names I’ve recently updated below as examples:


[1] A way round this is to change the name tag to the Scots form. This is probably the best solution with minor names: there is no reason why Sliddery Stane should be entered as it is on the OS as Sliddery Stone, when the second word is pronounced locally as stane. However, settlements should probably match their official form on the OS and road signs. Even if you were to enter the Scots form as the ‘name’ of a place with the English form listed under name:en, OpenStreetMap will default to name:en over name, as is the case with Sliddery Stane (which gets displayed as Sliddery Stone).

[2] A workaround is to use, which will display the Scots names recorded on OpenStreetMap. While this sounds like what we’d want, the map is let down by the limited features it displays (the names of burns, lochs and various other major and minor features aren’t shown) and the length of time it takes to update (often weeks if not months).

[3] Local pronunciations will vary and as Scots doesn’t have a fully standardised spelling system, choosing which form and spelling of a name could cause some systems an issue. As it happens, OpenStreetMap allows you to store alternative names, using the alt_name:sco tag. Cackerbush or Calkerbush? Enter them both. Doing this will improve the search functionality. Nonetheless, one form will have to be selected for the name:sco tag, which is the form that will be displayed on the map.

Eel Place-Names

Eels were an important part of the medieval and early modern economy, so it’s surprising that we don’t see more of them in place-names. There are only two Dumfries and Galloway eel place-names recorded on the first edition Ordnance Survey six-inch to the mile maps: Blue Eel Pool, Dryfesdale/Tundergarth DMF and Eel Spring Strand, Balmaclellan KCB.

The Ordnance Survey is a fantastic record of place-names, but it isn’t the only one. There are plenty of place-names which didn’t make onto this map, including those that mention eels. So far I’ve found three in D&G. There are probably more out there to be found.

Eelburn, Troqueer KCB NX 963 694

Eelburn is marked as two buildings – no doubt named from the burn – on Ainslie’s 1797 map of the Stewartry of Kirkcudbright. Neither the buildings nor the name make it to the OS. The burn itself isn’t given a name either.

Stewartry of Kirkcudbright [south east section], John Ainslie 1797
Reproduced with the permission of the National Library of Scotland

Eeel Hole, Kirkmahoe DMF NX 957 897

Eeel Hole is now buried under forestry. It’s not on a watercourse but the 1779 estate map – the only record of the name – shows that it was located at the south west end of a moss [a (peat) bog]. It holds the record for most consecutive <e>s in a Scottish place-name. (If we’re ignoring spaces, Little Eela Water, Wee Eldrick, Blue Eel Pool, and Bee Edge would tie it.)

The Farm of Auchengieth [Kirkmahoe], the Property of Robt Brown Esq, 1779
Reproduced with the permission of the National Library of Scotland

Eelholes, Tongland KCB NX 714 589

This name is recorded in the OS Name Books as Eelhole, but for whatever reason doesn’t appear on the map. The Name Book description is as follows: “A small cottage in indiffer[ent] repair on the farm of H[igh] Barncrosh & the prope[rty] of Jas. Carrickmoore of Corswall Wigtown Shir[e]” OS1/20/130/28.

Stewartry of Kirkcudbright [South east section], John Ainslie 1797
Reproduced with the permission of the National Library of Scotland