Lewis
Acknowledgement
Some years ago, after a seminar in Glasgow, A. MacCoinnich was kind enough to share his research findings with me. They appear as Appendix D8 in A. MacCoinnich, Plantation and Civility in the North Atlantic World – The case of the Northern Hebrides, 1570-1639, (Brill, Leiden, 2015) pp 455-460. Since he and I both used one of the principal primary sources our conclusions are similar. However I am grateful to him for generously providing his data and note that he too arrives at a total of about 300 pennylands for Lewis.
Principal Sources
OSA Vol 19, No 6, Stornoway, published 1797, pp 241-262
OSA Vol 19, No 7, Barvas, published 1797, pp 263-273
OSA Vol 19, No 8, Lochs, published 1797, pp 274-279
OSA Vol 19, No 9, Uig, published 1797, pp 280-288
NSA Vol XIV, Stornoway, published 1845, pp 115-140
NSA Vol XIV, Barvas, written 1836, published 1845, pp 141-150
NSA Vol XIV, Uig, written 1833, published 1845, pp 151-156
NSA Vol XIV, Lochs, written 1833, published 1845, pp 157-169
RMS I App 2 No 859, 1343
RMS I (742) 1382
RMS II (3578) 1511
RMS IV (2019) 1571-2
RMS VI (465) 1596, (1879, 1981-2) 1607
RMS VII (149, 167) 1609, (341, 346) 1610
RMS IX (672) 1637, (2150) 1649
RMS X (646) 1658
RSS I (1332) 1506, (1662, 1690) 1508
RSS II (4371) 1541
Acts of Parliament of Scotland Vol IV pp 160-4, 248-251
Register of the Privy Seal Vol LXI f 39
E655/1/2 Rental 1718
E655/2/2 Rental 1718
RH2/8/48 Rental 1718 – which is a photocopy of GB 248 GUA 00300 in Glasgow University Archives
Report to the Secretary for Scotland by the Crofters Commission on the social condition of the people of Lewis in 1901 as compared with twenty years ago. Appendix O (i) – Judicial Rental of the Island of Lewis, taken for the Forfeited Estates Commissioners in 1718 – published 1902.
GD1/400/2/3 Section B p 29 1573 (From Protocol Book of William Cuming)
GD 46/17/46 (C14538) – Seaforth Papers 1817
GD 305/1/1/5 1573-1590
GD 305/1/7/1 1596
GD 427/1/1 Rental Lewis, 1740
GD 427/11/1-7 Rentals etc 1769-1776
GD 427/2 Judicial Rental of The Lewis, 1754
GD 427/5/2 Pennylands in Lewis, 1765
Retours (Ross & Cromarty) (37) 1615, (79) 1633
NLS MS 19308 1739
RHP 43267 The Island of Lewis reduced from Mr Chapman’s Survey – William Johnson, 1821. National Library of Scotland also have a copy of this (EMS.s.543) available via their online Digital Map Library.
JB Caird, Early 19th century Estate Plans, in Togail Tir, ed. F. Macleod, Stornoway, 1989.
- Gibb’s copy of James Chapman’s MS map of Lewis – now in Stornoway Public Library – see Togail Tir p 51 ff.
Valuation Roll of the County of Ross for the Year Ending Whitsunday 1880, Inverness, 1879
WF Skene, Celtic Scotland Vol III pp 429-31
Highland Papers Vol II, Edinburgh, 1916, pp 284-8, 312-341
E Cregeen (ed.), Argyll Estate Instructions, SHS, 1964
Lawson, Bill, Lewis in History and Legend – The West Coast, Edinburgh, 2008
Lawson, Bill, Lewis in History and Legend – The East Coast, Edinburgh, 2011
A MacCoinnich, Plantation and Civility in the North Atlantic World – The case of the Northern Hebrides, 1570-1639, (Brill, Leiden, 2015) pp 455-460
J Macdonald, General View of the Agiculture of the Hebrides, Edinburgh, 1811
D Maciver, Place Names of Lewis and Harris, Stornoway, 1934
F MacLeod, Togail Tir, Stornoway, 1989
F MacLeod, The Chapels in the Western Isles, Stornoway, Lewis, 1997
F MacLeod, The Healing Wells of the Western Isles, Stornoway, Lewis, 2000
Martin Martin, A Description of the Western Islands of Scotland c. 1695, Birlinn, Edinburgh, 1994.
M Oftedal, The Village Names of Lewis in the Outer Hebrides, Norsk Tidsskrift for Sprogvidenskap, Bind XVII, Oslo, 1954, pp 363-409.
G Ponting, A Mini-Guide to Eoropie Teampull, Stornoway, 1989
M Robson, A Desert Place in the Sea, Ness, Lewis, 1997
I Stewart-Hargreaves & R Barrowman, Early Christianity in Lewis – The Ness Connection, Lewis, 2008
R Cochrane, Report of the Excursion of the Cambrian Archaeological Association, Dublin, 1900. (= Cambrian Scotch Excursion).
Early evidence of the pennyland system in Lewis is very scarce. At the end of the sixteenth century the Macleods of Lewis, the traditional owners, became involved in a long struggle against the Fife Adventurers who were supported by the Scottish realm. In the early seventeenth century the Macleods eventually succumbed to the Mackenzies. This change in ownership, after many years of conflict, probably explains the lack of surviving mediaeval documentation.
However we do have several documents from around 1600 which establish that Lewis was worth £40 or 60 merks. (APS IV p 163 (1598) & p 250 (1600); Skene III p 429). If the rate of exchange between ouncelands and merklands was the same in Lewis as in Harris and Skye then we should expect Lewis to be 15 davachs or ouncelands. (£40 = 60 merks; exchange rate of 1 ounceland to 4 merks; 60 divided by 4 = 15). Can we find any supporting evidence for this?
We have three documents from the end of the sixteenth century which demonstrate that Lewis was composed of 20d ouncelands. The first of these can be found in the research notes of David Murray Rose found under GD1/400/2/3 in the National Records of Scotland. Section B, p 29 gives a sasine from the Protocol Book of William Cuming dated 12/10/1573 at the Chanonry of Ross (Fortrose). It is from Torquil McLeod of Lewis in favour of his wife Margaret Nyn Angus McAlexander in a liferent of 4 davachs in Coigach and 6 davachs in Lewis. The latter comprise 20d Uige, 20d Valtos, 30d Loch, 20d Brogair, 20d Fallis, 10d Cullestaf on the north side of Lewis. We may conclude that 6 davachs amounted to 120d and that therefore a davach was the same as an ounceland of 20d in this part of the Hebrides. As we shall see below, Fallis is probably a misreading or mistranscription of Gallis, and Cullestaf of Tullestaf (t’s and c’s are often confused). The text makes clear that the Tullestaf mentioned is in North Lewis. This identifies it as the Tolsta in Stornoway parish rather than the one in Carloway parish.
Highland Papers II pp 284-8 prints a contract of 1576 between Rorie M’Cloid (MacLeod) of Lewis and Torquill M’Cloid, his son. This refers to the 20d of Croager and the 20d of the Lochis (along with other lands).
Origines Parochiales Scotiae, II, I pp 385-9 refer to a source in the Register of the Privy Seal Vol LXI f 39 (=P51/61 f 39 r) for 1590. This is a precept for a charter of confirmation of Torquil Macleod’s liferent grant to his wife Margarete nyne Angus MakAlexander. As in 1573 (above) it specifies 6 davachs in Lewis viz.:
20d Uge
20d Valtis (OPS reads Vallis. Having studied a copy of the original I read Valtis).
30d Lochis
20d Broagir
20d Gallis
10d Tolestaff
See also GD 305/1/1/5 1573-1590 which is the charter of confirmation and where some of the place-names appear to have corrupt spellings.
If we assume that Croager and Broagir are the same then we can modernise the spellings as follows:
20d Valtos (probably in Uig rather than Lochs)
30d (1573), 20d (1576) or 30d (1590) Lochs
20d Uig
20d Bragar
20d Galson
10d (North) Tolsta
Since the stated total is 120d, Lochs has to be 30d. That doesn’t necessarily mean the 1576 figure is a mistake, possibly not all of Lochs is being referred to. Most of these units are ouncelands or davachs. North Tolsta is half an ounceland whilst Lochs is one-and-a-half ouncelands.
Further indications of ounceland units are found in RMS VI (465) 1596, (1879, 1982) 1607 which all refer to the Castle of Stornoway, with 20 merklands adjacent to the castle, (which are reserved to the Crown). (See also GD 305/1/7/1 1596, Retours (Ross & Cromarty) (37) 1615, (79) 1633). This 20 merkland sounds like a castle ‘bordland’ – that is land allocated to provide resources specifically for the maintenance of the castle. We know of it in other Highland contexts where it has a Gaelic cloak in the form of the place-name ‘Borlum’. 20 merklands would be equivalent to 5 davachs or ouncelands on the basis that each davach or ounceland was worth 4m (which rate prevailed in the Northern Hebrides of Lewis, Harris and Skye).
Secondly we have some eighteenth-century evidence which helps us pin down the pennyland assessment more precisely. Unfortunately the rentals from the first three decades of the century give us a great deal of information about rents and very little about land-assessment. (It is tempting to try and deduce the latter from the former but there are too many variables for it to be more than educated guesswork). From 1754 we have a judicial rental (GD 427/2) which specifies at least 181⅛d. From 1765 we have a summary listing (GD 427/5/2) of the pennylands of Lewis which also gives the following totals by parish:
Lochs (excluding the Park) 18¼d
Uig 48¼d
Carloway 10½d
Cladach & Ness 137d
Eye (including Stornoway) 86½d
Total 300½d (½d over 15 ouncelands of 20d each).
The table shows 302⅝d which must allow for some possible duplication in Carloway parish. From this data it seems pretty clear that Lewis was made up of 15 x 20d ouncelands or 15 davachs.
There is one additional piece of evidence which reinforces the above conclusion. In an article in ‘The Scotsman’ in December 1887, Donald MacKinnon writes that “the people of Harris still speak of the five davachs of Harris and the fifteen of Lewis”.
In Place-names of Lewis & Harris p 86, D. Maciver mentions Geodha na Dabhach but it must be admitted that this may represent a tub-like topographical feature on the coast rather than a land-assessment unit. Ungeshader (in Uig parish at NB 1229) has a first element which may refer to an ‘Ung’ (ie tirung) or ounceland. (However, Oftedal – No 50 – thought a man’s name was the first element).
NLS MS19308 f4 1739 refers to ‘The 16 Davochs of the Lews’ (see also f11). This is late evidence and I am inclined to discount it since I doubt both this and some of the other figures given therein.
Unfortunately we have something of a disconnect between the late sixteenth-century evidence which talks in terms of ouncelands and the eighteenth-century rental evidence which is down at the level of pennylands and fractions thereof. We may not be seeing the intermediate holdings, the quarterlands or 5 pennyland units, which probably once also obtained.
J Macdonald, General View of the Agiculture of the Hebrides, writes (p 314):
Lands are subdivided not only as in Uist into pennies, half-pennies, and farthings, (the last of which ought not on any account to be allowed, being too small a farm for the maintenance of a family) but also further, into what are called clitigs, cianags, [see Glossary] etc. or half farthings, and the half of half farthings. No smaller subdivision of lands than half-penny lands ought to be permitted;
Macdonald’s concern about the minimum size of farms was shared elsewhere. Among the Instructions issued to the Chamberlain of Tiree in 1801, was one laying down a minimum extent of 4 mail lands. (Cregeen p 55).
This issue goes to the heart of the question of ‘extent’. Originally, davachs, pennylands and hides probably all embodied a vision of the amount of land required to sustain a family, or a group of families. (This, of course, begs the question of how big these families, or extended families, were). But whilst the conceptual framework might remain constant the stresses of economic progress and population growth would continuously put pressure on it. In the Highlands and Islands there is plenty of evidence of the tensions caused by the process of subdivision. Rising population levels in the eighteenth and nineteenth centuries caused many to emigrate, either voluntarily or, in the case of Clearance, by compulsion. For those who remained there was a natural tendency to subdivide the family holding amongst the children. These subdivisions were formalised with the names of fractions. It may be that this process operated more strongly in the later centuries of ‘extent’. We don’t just see it in the Western Isles. Islay had its own unique set of subdivisions. Perhaps the tiniest subdivisions are features of more recent times.
The Parishes of Lewis
It appears that the structures of land-assessment (davachs and ouncelands) were closely linked to parish organisation along the west coast and through the islands. What can we say about this in the context of Lewis? One source is the set of notes made by Timothy Pont and printed in Macfarlane’s Geographical Collections Vol II. Pont lists 8 parishes and makes it clear that the island of Pabbay was part of Uig, not Bernera. The Lewis Contract of 1598 printed in Volume IV of the Acts of Parliament of Scotland (p 160 ff) speaks of the need to build four parish churches in Lewis which was perhaps based on Dean Monro’s assertion that there were four parish kirks there. By the time of the ‘Ratification of the Infeftment of Lewis’ in 1600 (printed on pp 248-251 of the same volume) there had obviously been some feedback from the men on the ground. Firstly they had come to realise that Lewis was much less fertile than had been previously claimed and so a reduction in rent was needed! Secondly they must have learned that Monro’s information was misleading with regard to local religious requirements. Now they were given power to build ten parish churches on Lewis.
Two other important sources are the rentals of 1718 and 1726. The first is available under NRS E655/1/2 and E655/2/2 and there is a printed version with sterling values as Appendix O to the 1901 Report by the Crofters Commission on the social condition of the people of Lewis. The 1726 rental is printed in SHS Highland Papers Vol II pp 312-341. These sources complement and reinforce each other and the 1726 rental is very specific about which farm belonged to which parish – allowing us to map the constituent farms for six parishes. It seems that the long slow process of amalgamation meant that the original 15 ouncelands had become 8 parishes by the 1590s and 6 by 1726. I have summarised this data in Table A.
Johnson’s map of 1821 lends these parishes some geographical boundaries although Barvas seems to have replaced Cladach and Bragar from the seventeenth century. OPS II, I, p 388 suggests that Stornoway was composed of two parts – Stornoway proper and Gress – so that we can surmise the existence of about 10 early parishes. If we look at the pennyland data in the table we can also draw some interesting conclusions as to how parishes were composed. My parish totals are:
Lochs (excluding the Park) 18¼d
Uig (including the two Bernerays) 42⅞d
Carloway 13½d
Barvas 76½d
Ness 60d
Stornoway 60d
Eye 31½d
Total 302⅝d
It is striking that Ness and Stornoway are each exactly 3 ouncelands or 3 davachs. Lochs was possibly an ounceland and Carloway (where there may be some duplication) was substantially the smallest parish at probably only half an ounceland. The evidence does suggest that parishes were collections of ouncelands or davachs.
It is also clear that the concept of ounceland chapels (as suggested for the Northern Isles) has no meaning in Lewis. There are at least 37 early religious sites in Lewis which would work out at over 2 per ounceland or davach. It is also striking that, in Lewis, sites with Early Christian associations are not built into the land-assessment system. Virtually everywhere else in the area under study I can point to farms bearing ecclesiastical names and fiscal valuations that suggest they were religious endowments. (In other words, a Kil- site with a valuation of ½d, 1d, 5d, 10d and even 20d). The only island for which I struggle is Lewis. There are a few such farms. We have Baile na Cille (Uig) at 2¼d, Kirkibost (Bernera) at 2½d, Cross (Ness) at 4d and Pabail (Eye) at 7 or 7½d. Moreover most of the church sites begin with the element Teampull (from Latin templum), rather than Cille (from Latin cella), which might reflect a different process of conversion. Lewis is not short of church sites, it is just that they do not seem to have been built into the fiscal system as elsewhere in the Hebrides or facing mainland.
However we should acknowledge that early religious establishments may have survived into and through the Norse period even though we know nothing about the circumstances. So, in addition to the above list we have St Columba’s Isle and the Shiants in Lochs parish and Pabbay by Uig. If these bore assessments I do not know them.
In his study of the village-names of Lewis, M Oftedal concluded that the overwhelming majority were of Norse origin. If we look at the names of those farms which have fiscal values we would draw exactly the same conclusion. It would appear that when pennylands were first established, c. 1000 AD, almost every farm in Lewis had a Norse name and that there was not yet the fiscal support for the Christian church that obtained elsewhere amongst the Norse colonies of the Hebrides and west coast. This does suggest a ruling stratum that was Norse and probably still pagan.
The agricultural division of Lewis
The agricultural division of Lewis as represented by the rentals of 1718 and 1726 is not identical with the mediaeval pattern. We can glimpse the earlier framework through these and other rentals, the Blaeu map of Lewis and Harris which was based on Pont’s survey of the 1580s-1590s, and Johnson’s map of 1821 which was based on Chapman’s Book of Plans dated 1807-9. The eighteenth century rentals often show farms which were already composites. We can tell the older units in Johnson’s map which shows the boundaries of many of the Lewis farms. Although these may not have been specified in 1718 and 1726 we can assume that they were still known and regarded as components of the larger farms there named. The memory of an older framework survived and was recorded by successive cartographers and surveyors. In Togail Tir (pp 54-5), J. B. Caird has elaborated this earlier pattern for part of Lochs and the same process can be applied to the other parishes in the island.
It is striking how Pont’s survey (as it survives in Blaeu’s map of Lewis and Harris), is reinforced by Johnson’s map. In fact we can even tell something of the routes Pont is likely to have travelled. In the parish of Lochs for instance he only names four farms south of Loch Erisort. Pont picked up the names of many farms and gives us an invaluable insight into the mediaeval situation in Lewis but he is least well-informed about Uig and South Lochs.
Carloway offers another example of the slow process of amalgamation. In 1718 and 1726 it appears that this half-ounceland parish only consisted of Tolsta Chaolais, Kerevig, Upper Carloway and Dell (more & beg). Yet if we study Johnson’s map it apparently consisted of these same four units plus distinct farms called Sandwick, Limshade(r), a blank unit that was probably Borghastan, Garinnin and Knock. (I have disregarded S. Shawbost since this looks like an anomaly). We find confirmation of two of these latter units in Pont(Blaeu) where 5 separate farms are named (Tollosta=Tolsta, Kadlowa oc=Carloway Upper, Knockille=Lower Knock, Knock, Ghearen Vyg (Garinnan Uig to distinguish it from Ghearen in Ness), Dallenille=Lower Dell and Dall=Dell). So the four farm units of 1718-26 were, in mediaeval times, at least eight, possibly nine.
However, just as the process of amalgamation went on in the seventeenth and eighteenth centuries, so, earlier, did the process of decomposition. If we look at the list of farms in 1718 and 1726, or in the maps of Johnson or Blaeu, it is immediately apparent than many include the word shader. This comes from the Norse setr or sheiling which means summer pasture ground. So any place-name including the element shader was probably once just the shieling ground for another more important unit elsewhere, probably a primary farm on the coast. Over time these sheilings became permanently settled and independent of their former parent. In Norse or Dark Age times there may well have been fewer but larger farms whose sheilings subsequently became separate.
The parish of Eye seems to offer a good example of this process. Garrabost and Paible may both include the element bolstadr or farm whilst Suardail includes the element dalr or dale which is, by definition, an important topographical feature. Sulaisiadar, Siadar and Seisiadar all include the element shader or shieling. Knock and Aignish may well be later sub-divisions in which case it seems likely that the Eye peninsula in the Norse period included three primary farms towards its western end, (Garrabost, Suardail and Paible), with three complementary sheilings towards the eastern side (Sulaisiadar, Siadar and Seisiadar).
Leave a Reply