In other words, each of the eight free maenols contained sixteen homesteads, which sixteen homesteads were first classified in groups of four called trevs. Or, to put the case the other way, the eight free maenols, were divided into quarters or trevs, and these trevs again each contained four homesteads.
It is evidently a tribal arrangement, clustering the homesteads numerically for purposes of the payment of gwestva, and probably the discharge of other [p203] public duties, and not a natural territorial arrangement on the basis of the village or township.
Turning now to the Dimetian and Gwentian Codes, according to which the free trev instead of the maenol is the gwestva-paying unit:255 there is first the group of twelve trevs (instead of twelve maenols) under a single maer, and under the name of maenol instead of cymwd; but apparently all the trevs in the group of twelve256 are free trevs. There are other groups of seven taeog-trevs making a taeog-maenol, and the maenol (instead of the cymwd) has its court, and becomes the unit of legal jurisdiction.257
Confining attention to the free maenol, the first thing to notice is that each of the twelve free trevs of which it was composed paid its gwestva, or tunc pound in lieu of it. The trev, therefore, was the gwestva-paying unit.
And as to the interior of the trev we read,—
'There are to be four randirs in the trev, from which the king's gwestva shall be paid.'
'312 erws are to be in the randir between clear and brake, wood and field, and wet and dry, except a supernumerary trev [the upland has in addition].' 258
In this case the 'tunc pound' of 240d. was paid by each trev of 4 randirs, each randir containing 312 erws, and the trev 1,248 erws in all. The trev in South Wales is, therefore, slightly larger than the [p204] Venedotian maenol. Here we are bound by no law that the pence in the gwestva should exactly correspond with the number of erws. But in the other versions the 12 odd erws in the randir are stated to be for 'domicilia,' 259 or buildings, and 12 erws would allow of 3 tyddyns of the requisite 4 erws each.
This fixes for us the number of homesteads or tyddyns in the trev. There were 3 tyddyns to each randir, and 4 randirs to the trev, and so there were 12 tyddyns in each trev, and to each tyddyn there were appendant 100 erws in the arable, pasture, and waste.
The trev which paid its tunc pound of 240d. was thus made up of 12 holdings, each paying a score pence. And as in the Latin version of the Dimetian Laws (p. 825) a score pence is translated uncia argenti, the connexion is at once made clear between the system of grouping the holdings so as to pay the tunc pound, and the monetary system which prevailed in Wales, viz., that according to which 20d. made an ounce, and 12 ounces one pound. The 12 holdings each paying a score of pence, or ounce of silver, made up between them the tunc pound of the trev.
This curious geometrical arrangement or classification of tyddyns and trevs, with an equal area of land to each, is at first sight entirely inconsistent with the division of the family land among the heirs of the holder, inasmuch as the great grandchildren when they divided the original family holding must, one would suppose, have held smaller shares than their great [p205] grandfather. And there is only one answer to this. It would have been so if the tribe, and the families composing it, were permanently fixed and settled on the same land, and pursuing a regular agriculture, with an increasing population within certain boundaries. But the Welsh were still a pastoral people, and, as we shall see when we come to examine the Irish tribal system, while the homesteads and land divisions were fixed, the occupants were shifted about by the chiefs from time to time, each sept, or clan, or family receiving at each rearrangement a certain number of tyddyns or homesteads, according to certain tribal rules of blood relationship of a very intricate character.
This permanence of the geographical divisions and homesteads, and shifting of the tribal households whenever occasion required it, was only possible with a pastoral and scanty population. Long before the fourteenth century the households were settled in their homesteads, geometrical regularity had ceased, and the land was divided and subdivided into irregular fractions. This is the state of things disclosed in the Record of Carnarvon. But in the tenth century, according to the Welsh laws, the old tribal rules were apparently still in force.
Without pretending to have mastered all the details of these obscure tribal arrangements, the point to be noted is that the scattering of the tyddyns all over the country side, and the clustering of them by fours and sixteens, or twelves, into the group which was the unit paying the gwestva or tunc pound, and again into clusters of twelve or thirteen260 under a [p206] maer, as the unit of civil jurisdiction, were obviously distinctive features arising from the tribal holding of land, and that the system was adopted apparently to facilitate the division of the land among the families in the tribe somewhat in the same way as in the open field system the division of the arable land by turf balks into actual erws facilitated the division of the ploughed land among the contributors to the plough team.
Bearing this in mind we may now turn back to the Domesday Survey, and compare its description of the land system of Gwent and Archenfield with the results obtained from the Welsh laws.
In order, however, to make this comparison the Welsh terms must be translated into Latin, otherwise it will be difficult to recognise the trev, and maer, and maenol, and gwestva in the Domesday description.
The before-mentioned Latin version of the Dimetian Code, the MS. of which dates from the early thirteenth century, will do this for us.261
It translates trev, the unit of the tunc pound, by villa. It takes the Welsh word 'maenol' as equivalent to manor, and indeed it did resemble the Saxon and Norman manor in this, that it was the unit of the jurisdiction of each single steward or villicus of the chief. This officer was called in Welsh the maer, which was translated into the Latin præpositus. He did to some extent resemble the English præpositus, but he differed in this—that instead of being set over the 'trev' or 'villata' of a single manor, [p207] the Welsh maer was, as we have seen, set over a number of 'villas' or trevs—thirteen free trevs or seven taeog-trevs, in Gwent—each free trev of which rendered its 'tunc pound' or 'gwestva,' and each taeog or villein-trev its 'dawn-bwyd' of food.
Now, this is precisely what is described in the Domesday Survey of Gwent.
There are four groups of thirteen or fourteen 'villas' or trevs, each group under a 'præpositus' or maer; and these four groups, which were in fact Gwentian 'maenols,' rendered as gwestva a food-rent amounting to 47 sextars of honey, 40 pigs, 41 cows, and 28 shillings for hawks.
In the district of Archenfield the clusters of trevs do not appear, but the food-rents were similar—honey being a marked item throughout.
In the Welsh gwestva, also, honey was an important element. It is mentioned as such in the Welsh codes, and it is conspicuous also in the Domesday Survey both of Gwent and Archenfield.
Its importance is shown by the fact that in the Gwentian Code a separate section was devoted to 'The Law of Bees.' It begins as follows:—'The origin of bees is from Paradise, and on account of the sin of man they came from thence, and they were blessed by God, and, therefore, the mass cannot be without the wax.' 262
The price of a swarm of bees in August was equal to the price of an ox ready for the yoke, i.e. ten or fifteen times its present value, in proportion to the ox.
Honey had, in fact, two uses, besides its being the [p208] substitute for the modern sugar—one for the making of mead, which was three times the price of beer; the other for the wax for candles used in the chief's household, and on the altar of the mass.263 The lord of a taeog had the right of buying up all his honey;264 and in North Wales, according to the Venedotian Code, all the honey of the king's aillts or taeogs was reserved for the court.265 The mead brewer was also an important royal officer in all the three divisions of Wales.
It is not surprising, then, that the tribute of honey, which formed so important a part of the Welsh gwestva, should be retained as an item in the tribute of the trevs of Gwent after their conquest by Harold.
From the combined evidence of the Domesday Survey and the 'Ancient Laws of Wales,' the fact has now been learned that in the eleventh century, as it had done previously probably for 400 years, the river Wye separated by a sharp line the Saxon land, on which the manorial land system prevailed, from the Welsh land, on which the Welsh tribal land system prevailed. On the one side of the river, at the date of the Survey, clusters of scattered homesteads of free Welshmen contributed food-rents in the form of gwestva to the conqueror of their chief, and taeogs their dawn-bwyds. On the other side the villata of geneats and geburs, besides paying gafol, performed servile week-work upon the demesne lands of the lord of the [p209] village or manor. It may be well, however, to seek for some earlier evidence of the payment of gwestva on the Welsh side of the river.
Documentary evidence of the manorial system on the Saxon side was forthcoming as early as the seventh century, in the laws of King Ine. How far back can documentary evidence be traced of the Welsh system?
In the possession of the church of Llandaff there was long preserved an ancient MS. of the Gospels in Latin, called the Book of St. Chad.266 This MS. appears to date back to the eighth century. And it was for long the custom to enter on its margin a record of solemn compacts sworn upon it, as in the similar case of the Book of Deer. It thus happens to contain (inter alia) two short records of grants to the church of St. Teilo (or Llandaff). One of these gifts is as follows:267—
'This writing showeth that Ris and the family of Grethi gave to God and St. Teilo, Treb guidauc. . . and this is its census: 40 loaves and a wether sheep in summer; and in winter, 40 loaves, a hog, and 40 dishes of butter. . . .'
Another is in these words:—
'This writing showeth that Ris and Hirv . . . . gave Bracma as far as Hirmain Guidauc, from the desert of Gelli Irlath as far as Camdubr, its "hichet" [food-rent?], 3 score loaves and a wether sheep, [p210] and a vessel of butter. And then follow the witnesses.' 268
Rhys ap Ithael, the donor in these two cases, was king of the district of Glewyssig in the middle of the ninth century, about the time of Alfred the Great. Now, a king or chief would hardly be likely to transfer to the church of Llandaff a free trev and the gwestva paid therefrom. This would have involved the severance of free members of the tribe from the tribe, to put them under an ecclesiastical lordship. We should expect then to find that the Trev 'Guidauc' was a taeog-trev on the chief's own land, and according to the description given in the grants, the census corresponds not with the gwestva of a free trev under the Welsh laws, but with the 'dawn-bwyd' of the taeog-trev.
The food tribute in these grants was divided into summer and winter payments, and so, as we have seen, were the dawn-bwyds of the taeogs in the Welsh laws; the scores of loaves, the sow, the wether sheep, and the tubs of butter, correspond also with the food-gifts from the taeog-trevs, as described in the laws, though with varying quantities.269
These grants in the margin of the Book of St. Chad may, therefore be taken as evidence that the system of food-rents was prevalent in Wales in the middle of the ninth century.
There is still earlier evidence of the prevalence of the system of food-rents where we should little expect to find it, viz., in the laws of King Ine. Ine being King of Wessex, and Wessex shading off as it [p211] were into the old British districts both south and east of the Severn, it was but natural that some old Welsh or British customs should have survived in certain places; as Walisc men here and there survived amongst the conquering English. These Welshmen were allowed under Ine's laws to hold half-hides and hides of land. We have only to examine the Domesday Survey for Gloucestershire and Herefordshire to find traces even at that date of survivals of Welsh and Saxon customs in exceptional cases, even outside those districts which had only just been conquered.
In some places where Saxon customs had long prevailed a little community of Welshmen remained under Welsh customs. In other places the customs were partly Welsh and partly English.270 [p212]
In precisely the same way survivals such as these must have existed in King Ine's time. There must have been then, as 400 years afterwards, at the date of the Survey, places in Wessex where Welshmen predominated and Welsh customs survived. There must have been, in other words, manors which paid Welsh gwestva instead of Saxon services. There is a remarkable passage in King Ine's laws which can only be thus explained. On the same page, and in the next paragraph but two to the law about the yard-land set to 'gafol' and to 'weorc,'271 there is a clause apparently out of place, which begins abruptly with this heading: 'Æt x. hidum to fostre.' 272 In the Latin version this is rendered 'De x. hides ad corredium.' 273 Now, there is a passage in a charter of Louis VII. of France, anno 1157, given by Du Cange under the word 'Corredium,' in which certain 'villas' are freed from the exaction of 'quædam convivia, quæ vulgo Coreede vel Giste vocantur.' This definition of corredium and of 'giste,' as a contribution of food exacted from tenants, corresponds exactly to the Welsh 'gwestva.' And the Saxon word fostre also means food. So that this heading to the passage in question may be translated—'from x. hides paying gwestva.' And so interpreted the following list [p213] becomes perfectly intelligible, for it describes what the gwestva consisted of.
Now, if the system of gwestva payment or food-rent described in this passage of the laws of King Ine be evidence of the survival of the Welsh custom after the Saxon conquest, it is at the same time equally clear documentary evidence of the seventh century that the system of gwestva or food-rents was prevalent outside Wales in the west of Britain before the Saxon conquest.274
203. Liber Landavensis, p. 545. Ordericus Vitalis, ii. 190. It may have been conquered in 1049, after Gruffydd and Irish pirates had, according to Florence, crossed the Wye and burned 'Dymedham' (see Freeman's Norman Conquest, ii. App. P); but most likely shortly before A.D. 1065, under which date is the following entry in the Saxon Chronicle:—
'A. 1065. In this year before Lammas, Harold the Eorl ordered a building to be erected in Wales at Portskewith after he had subdued it, and there he gathered much goods and thought to have King Edward there for the purpose of hunting; but when it was all ready, then went Cradock, Griffin's son, with the whole force which he could procure, and slew almost all the people who there had been building.'
204. Domesday, i. 162 a.
205. Ibid. 162 a et seq.
206. 185b. See also Freeman's Norman Conquest, ii. App. SS, p. 685.
207. Domesday, i. 179 a.
208. See Leges Wallice, p. 812. 'De qualibet villa rusticana debet habere ovem fetam vel 4 denarios in cibos accipitrum.' The 54 villæ at 4d. each would make xviii s. (? whether xxviii. by an extra x. in error).
209. Domesday, i. 162 a.
210. Domesday, i. 162 a (last entry).
211. F. 179 a.
212. F. 181 a.
213. So f. 185b: 'In Castellaria de Carlion . . . iii. Walenses lege Walensi viventes cum iii. car. et ii. bord. cum dim. car. et reddunt iiii. sextar. mellis.'
214. Ancient Laws of Wales, p. 373.
215. Description of Wales, chap. cxvii.
216. C. x.
217. C. viii. The district of Snowdon afforded the best pasturage and Anglesey the best corn-growing land.
218. C. viii. and xvii. In the Isle of Man four oxen were yoked abreast to the plough, Train's Isle of Man, ii. p. 241.
219. C. viii.
220. C. xvii.
221. Ancient Laws and Institutes of Wales. Record Commission, 1841. See preface by Aneurin Owen.
222. Venedotian Code. Ancient Laws of Wales, pp. 81–2, and see pp. 644–6 (Welsh Laws). Mr. Skene, in his chapter on The Tribe in Wales in his Celtic Scotland, iii. pp. 200, 201, does not seem to have grasped fully the distinction between the free tribesmen and their family land on the one hand and the Aillts and Taeogs with their geldable or register land on the other. Everything, however, turns upon this. Compare Welsh Laws, xiv. s. 31 and s. 32 (pp. 739–741), where the distinction is again clearly stated.
223. Ancient Laws of Wales, p. 638 (s. 45).
224. Id. 651 (s. 83).
225. P. 639 (s. 51).
226. Pp. 81–2.
227. See the surveys in the Record of Carnarvon (14th century), where the holdings are sometimes called 'Weles,' thus:—'In eadem villa sunt tria Wele libera, viz. Wele Yarthur ap Ruwon Wele Joz. ap Ruwon and Wele Keneth ap Ruwon. Et sunt heredes predicte Wele de Yarthur ap Ruwon, Eign. ap Griffiri and Hoell. ap Griffri et alii coheredes sui;' and so on of the other Weles (p. 11). This is the common form of the survey passim.
228. Ancient Laws, &c., of Wales, p. 741.
229. Id. pp. 82 and 740.
230. The fullest description of the rules of 'family land' are those in the Venedotian Code, c. xii., The Law of Brothers for Land, pp. 81 et seq. See also Welsh Laws, Book IX. xxxi. p. 536; also Book XIV. xxxi. pp. 739 et seq.
231. Ancient Laws, &c., of Wales, Glossary, p. 1001.
232. The Record of Carnarvon, passim. Thus 'the Wele of So-and-so, the son of So-and-so, and the heirs of this Wele are So-and-so.'
233. This was not payable if an investiture fee had been paid by the person dying.
234. Ancient Laws, &c., p. 92 and 93.
235. Id. p. 375.
236. Book of Carnarvon, passim.
237. Sometimes an 'uchelwr' or tribesman had taeogs under him. Ancient Laws, &c., pp. 88, 339, and 573. See also Id. p. 646. Welsh Laws.
238. Id. pp. 82 and 536. Welsh Laws, s. xxxii.
239. Id. p. 376.
240. Id. p. 82.
241. Id. p. 82.
242. It was sometimes called 'tir kyllidin,' or geldable land, as before stated.
243. Ancient Laws, &c., p. 673.
244. Pp. 36–7 and 212–13.
245. Id. pp. 88 and 646.
246. Pp. 93 and 376.
247. P. 375–6. Gwentian Code, 11, xxxv.
248. Ancient Laws, &c., p. 697.
249. P. 294 (Dimetian Code). 'The caeth—there is no galanas (death-fine) for him, only payment of his "werth" to his master like the "werth" of a beast.'
250. Ancient Laws, &c., pp. 90–1.
251. Id. p. 91, s. 14.
252. Id. p 91, s. 15. In Leges Wallice, p. 825, 'score pence' or 'score of silver' is translated 'uncia argenti;' ∴ 3 uncie agri should equal a 'trev.' See Liber Landavensis, pp. 70 and 317.
254. The word Gabail still in Scotch Gaelic retains its meaning of a farm. The word is pronounced 'gāv´-ul.'
255. Ancient Laws, pp. 261. 'Four randirs are to be in the trev from which the king's gwestva is to be paid' (s. 5).
256. In upland districts there were 13 trevs in the maenol, p. 375.
257. There were seven taeog-trevs in taeog-maenols, and each contained three randirs, in two of which there were three taeog-tyddyns to each, the third being pasture for the other two. There were therefore six taeog holdings in each taeog-trev. Ancient Laws, &c., pp. 375 and 829.
258. Pp. 374–5.
259. P. 829. 'In randir continentur ccc. et xii. acre: ut in ccc. acris, araturam, et pascua et focalia possessor habeat; inde xii. domicilia.' See also p. 790. 'Id est xii. domicilia.' The Dimetian Code has it 'space for buildings on the 12 erws' (p. 263).
260. 'There are to be thirteen trevs in every maenol, and the thirteenth of these is the supernumerary trev.' Gwentian Code, p. 375.
261. Leges Wallice, Ancient Laws, &c., p. 771 et seq.
262. Ancient Laws, &c., p. 360.
263. Ancient Laws, &c., p. 325.
264. Id. p. 213.
265. Id. p. 92 (s. 5).
266. Liber Landavensis, p. 271, App., and p. 615.
267. For the translation see p. 616. For the original, p. 272, as follows: 'Ostendit ista scriptio quod dederunt Ris et luith Grethi Treb guidauc i malitiduck Cimarguich, et hic est census ejus, douceint torth hamaharuin in irham, haduceint torth in irgaem, ha huch, ha douceint mannudenn deo et sancto elindo. . . .'
268. For the translation see p. 617; for the original, p. 272.
269. See Leges Wallice, ii. 14, 'De Daunbwyt' [Dono Cibi]. Ancient Laws, &c., of Wales, p. 790.
270. Fol. 162 b. 'In Cirencester hundred King Edward had five hides of land. In demesne v. ploughs and xxxi. villani with x. ploughs, xiii. servi and x. bordarii, &c. The Queen has the wool of the sheep. T. R. E.: this manor rendered iii.12 modii of corn, and of barley iii. modii, and of honey vi.12 sextars, and ix.l. and v.s., and 3,000 loaves for dogs.'
This is very much like a survival of the Welsh food-rents at one of the cities conquered by the Saxons in 577.
In some other places out of Archenfield there was a mixture of Welsh and English customs.
The manor of Westwode (f. 181) was held by St. Peter of Gloucester. It contained vi. hides, 'one of which had Welsh custom, the others English.' A Welshman in this manor had half a carucate, and rendered i. sextar of honey.
And at Clive (f. 179 b), 8 Welshmen had 8 teams, and rendered x.12 sextars of honey and vi.s. v.d., and in the forest of the king was land of this manor, which T. R. E. had rendered vi. sextars of honey, and vi. sheep with lambs.
These instances are sufficient to show that in Herefordshire, as in Gloucestershire, in the newly conquered districts, the old Welsh dues of honey, sheep, &c., remained undisturbed; while in the districts which had long been under Saxon rule, in some few cases there was a mixture of services, and in others the Saxon services of ploughing on the lord's demesne had become general.
It may be assumed that when the services were thus described contrary to the usual routine of the Domesday surveyors, it was because there was something unusual about them; and that in the majority of instances where Saxon customs prevailed, no description was deemed needful. Compare the Domesday survey of Dorsetshire—a portion of the 'West Wales'—where the manors in the royal demesne are grouped so that each group renders a 'firma unius noctis,' or a 'firma dimidiæ noctis.'
271. Laws of Ine, No. 67. Thorpe, p. 63.
272. Id. No. 70. Thorpe, p. 63.
273. Id. p. 504.
274. For much curious information respecting the Welsh system of tenures, see Taylor's History of Gavel-kind. London, 1663.