¹ In this article the volumes of the Place-Name Society are abbreviated by P-N and the name of the county. Other abbreviations are Collingwood and Myres = Collingwood, R. G. and Myres, J. N. L., Roman Britain and the English Settlements (Oxford 1936); Hodgkin = Hodgkin, R. H., A History of the Anglo-Saxons (Oxford 1935); Stenton = Stenton, F. M., Anglo-Saxon England (Oxford 1943).Google Scholar

² Maitland, F. W., Domesday Book and Beyond (Cambridge 1897) 506–15.Google Scholar

³ Corbett, W. J., ‘The Tribal Hidage,’ Transactions of the Royal Historical Society, new series 9 (1900) 187–230. Evidence exists of the relation of soldiers to hides in a later period. Stenton 288. For a good study of Anglo-Saxon taxes see ibid. 284–9.Google Scholar

⁴ Brownbill, J., ‘The Tribal Hidage,’ English Historical Review 27 (1912) 625–48; 40 (1925) 497–503. The problem of the grand total is discussed on p. 498.Google Scholar

⁵ Chadwick, H. M., Studies in Anglo-Saxon Institutions (Cambridge 1905) 263–8.Google Scholar

⁶ The manuscripts have been described by Brownbill, , op. cit. 625–7. A facsimile of the one manuscript of E appears in ibid. 497 and in Hodgkin, II, 389.Google Scholar

⁷ Possibly the mistake occurred as follows. The original in a Latin copy would have been ‘mdccc.’ A careless copyist may have assumed that the initial m was actually on the end of the preceding word, Spalda, and by placing a line over the last a of Spalda would have lost the m from the number, thus: Spalda mdccc Spaldā dccc The chief objection to assigning 1,800 as the number would be that it is the only instance of this number in the catalogue. However, it fits into the pattern of numbers as half of 3,500 or three times 600. See note 58 below.Google Scholar

⁸ The longest treatment is by Professor Stenton (pp. 292–4) who believes that the Tribal Hidage is ‘an attempt to guide a king's minister in the exaction of his dues from subject provinces.’ No evidence, however, is given in support of this thesis. Hodgkin II, 401; Collingwood and Myres 383.Google Scholar

⁹ Stenton 294.Google Scholar

¹⁰ Stenton 213.Google Scholar

¹¹ Stenton 178.Google Scholar

¹² The dating of Beowulf about 823–50 is given by Bond, George, ‘Links between Beowulf and Mercian History,’ Studies in Philology 40 (1943) 481–93.Google Scholar

¹³ The date of Widsith is given by Kemp Malone (ed. Widsith, , London 1936, p. 54) as of the ‘latter part of the seventh century.’ Widsith's emphasis upon Offa I and his famous boundary (ibid. 71–2) leads one to suspect that the poem in one form was prepared for that second Offa who made Offa's Dyke, the boundary of Mercia. The references to tribes are: Gefflegen, 147; Heathobards, 155; Herefaran, 162; Sweordoras, 191; Wicingas, 155. Note also the resemblance of the Wederas of Beowulf (Chambers, , Beowulf 342) to the Widerigas of the Tribal Hidage.Google Scholar

¹⁴ Stenton 292.Google Scholar

¹⁵ See, for instance, Thorpe, B., Diplomatarium Anglicum Aevi Saxonici (London 1865) 145, 149, 480; Kemble, J. M., Codex Diplomaticus Aevi Saxonici (London 1847) I, 1, 2, 3, 109, 128, 149, 183, 253, 257.Google Scholar

¹⁶ Thorpe, , op. cit. 95, 107, 111; Kemble, , op. cit, I, 297; II, 28, 53.Google Scholar

¹⁷ Kemble I, 128, 183, 272.Google Scholar

¹⁸ The Latin version (‘Ab eo loco ubi primum Mircheneland nominatur’) seems to confirm the reading.Google Scholar

¹⁹ The Mercian letters come from Speculum 18 (1943) 460–1. The Kentish letters are from Cambridge Historical Journal 7 (1942) 114.Google Scholar

²⁰ Stenton 42, 292, 294.Google Scholar

²¹ Hodgkin II, 391; Kemble, , Codex I, xxiv.Google Scholar

²² Stenton 262; Hodgkin II, 602.Google Scholar

²³ Cambridge Historical Journal 7 (1942) 111. Some charters of Kent attributed to Aethelbert of Kent give termini of boundaries (Kemble, , Codex I, 1, 2, 3).Google Scholar

²⁴ Bede, , Hist. Eccles. 4, 13; 4, 26; 3, 24 respectively.Google Scholar

²⁵ An objection may be raised that hidage became static and thus does not represent population accurately. This is true but it does not preclude the idea that an increase of hides in an age of light settlement means an increase of families and thus of population.Google Scholar

²⁶ See below note 34 for revision of East Anglia's hidage to 15,000.Google Scholar

²⁷ The assumption that this was a Mercian document made Wight a difficult problem. Some thought it an unidentified place in central England (Hodgkin I, 401; Stenton 293), while another thought that the Cilternsaetan were West Saxons (Brownbill 498–9). The latter believed that the 1,200 of Bede's Wight included the Gifla and the Hicca (the Yte across the water from Wight).Google Scholar

²⁸ At one time it seemed to me that this document might have come from the time of Raedwald. In this case East Anglia's 30,000 might be considered an imperial total which would have included 15,000 of the East Anglia together with the tribes from ‘Hendrica’ to ‘East Sexena’. The difficulty of reconstructing an original based upon this hypothesis makes this choice less attractive than the other even without the presence of Wight.Google Scholar

²⁹ Stubbs, W., ed. Willelmi Malmesbiriensis monachi de gestis regum Anglorum (London 1887) I, 21; Annals of Winchester, Annales Monaslici (Rolls Series) II, 3. Collingwood and Myres 424, 459. This was about 593 A. D. Google Scholar

³⁰ The root is assumed to be Gag, and thus does not fall into the series of terms ending in ga which is discussed below.Google Scholar

³¹ Hodgkin I, 189.Google Scholar

^31a (Thanet) ‘id est magnitudinis iuxta consuetudinem aestimationis Anglorum familiarum sexcentarum.’ Bede, , Hist. Eccl. 1, 25. (Mevanian isles) ‘Nongentarum sexaginta familiarum mensuram iuxta aestimationem Anglorum, secunda trecentarum et ultra spatium tenet.’ Ibid. 2, 9. (Report of gift) ‘Donavit terram quinquaginta familiarum.’ Ibid. 4, 3. (Isle of Wight) ‘Est autem mensura eiusdem insulae, iuxta aetimationem Anglorum mille ducentarum familiarum.’ Ibid. 4, 16. (South Saxons, Sussex) ‘habens terram familiarum septem mille.’ Ibid. 4, 13. Bede certainly had no practical purpose in reporting these names and their area in hides.Google Scholar

³² Hist. Eccl. 5, 23, as translated by King, J. E. (Loeb Classical Lib.) II, 371.Google Scholar

³³ An objection to this argument is that the hundreds are late institutions and therefore have no necessary relationship to the number of hides of so early a date as the Tribal Hidage. However, territorial divisions are often very permanent, and the Tribal Hidage is based upon hundred hide or three hundred hide units. Certainly no one has yet proved that these units did not persist. The evidence that four areas (Kent, Surrey, Essex and East Anglia) did maintain a positive relationship over the centuries can hardly be questioned by lack of positive evidence of continuity.Google Scholar

³⁴ The East Saxon hundreds were located as follows: 20 in Essex, 5 in Middlesex, and 2 in Hertfordshire. If we have identified Hendrica with Surrey correctly, its hundred-hidage relationship is correct: 14 hundreds and 3,500 hides. The Sussex relationship shows 600 hundreds but only 7,000 hides.Google Scholar

³⁵ Collingwood and Myres, map opposite p. 383; Hodgkin, maps on pp. 284, 285; Stenton, map on p. 200.Google Scholar

³⁶ Bede, , Hist. Eccl. 3, 24. For a statement of recent work on the Wrekin and pictures which illustrate the commanding position of the heights see Kenyon, K. M., ‘Excavations on the Wrekin, Shropshire, 1939,’ The Archaeological Journal 99 (1942) 99–109.Google Scholar

³⁷ Chadwick, , Studies 291, note 1. See also Thorpe, , Diplomatarium 61.Google Scholar

³⁸ Gildas is edited by Mommsen in MGH, Auctores antiquissimi 13, i (1894) 1–85; Stevenson, J., Gildas, De excidio Britanniae (London 1838) 37–42. F. Liebermann (‘Nennius’, Essays in Medieval History presented to Thomas Frederick Tout, Manchester 1925, p. 40) believes that Vortigern's son held Buellt of Aurelius Ambrosius. This is based upon a statement in Nennius (Lot, F., Nennius et l'Historia Brittonum, Paris 1934, p. 188). If this Aurelius Ambrosius was a predecessor of Aurelius Caninus it would tend to identify the territory to the south of Buellt as the area held by these Aurelii. Nennius links with Buellt the territory named for Vortigern to the northeast.Google Scholar

³⁹ See the appendix by Professor U. T. Holmes, University of North Carolina, upon this point.Google Scholar

⁴⁰ See Calendars of Inquisitions post Mortem V, 334–5 and IX, 333. See also Lloyd, J. E., A History of Wales from the Earliest Times (London 1939) I, 277.Google Scholar

⁴¹ For the location of the river see Rees, William, South Wales and the Border in the Fourteenth Century (Ordnance Survey Office, Southampton S. E.). The river appears near (just west of) Gelligaer. Two routes from the upland into the lowland apparently cross this tiny crooked stream. For references to this stream see Thomas, R. J., Enwau Afonydd a Nentydd Cymru (Caerydd 1938) 25. See appendix. These items about Caeach were given me by J. G. Edwards, Jesus College, Oxford, England. Another argument for placing the Tribal Hidage in the time of Aethelbert instead of a century or more later is that the rulers of this part of Wales (Morgannwg) of the later period are known, and they do not include a place for the Gagas.Google Scholar

⁴² Phepson would seem to be a case of movement of the Feppinga to Worcestershire. P-N Worcestershire xvii-xix.Google Scholar

⁴³ This may well explain the series of places which seem to have the Widerigga in their names, see below note 56.Google Scholar

⁴⁴ That is, it is no argument against an identification of a tribal name as an element in a place name if the Place-Name Society does not consider the possibility of its being the source. The existence of several places with elements possibly deriving from a tribe may be used as evidence that the tribe was in the area.Google Scholar

⁴⁵ The name, Surrey, and a subregulus of the area can be carried back as far as 675 A. D. Early forms of the name were Suthrige and Sudergeona. P-N Surrey xvi.Google Scholar

⁴⁶ Appendix. Florence of Worcester (English Hist. Soc.) I, 234. Stenton 291, 298. Modern names derived are Sonning, Sunningwell, Sunningdale, and Sunninghill.Google Scholar

⁴⁷ Bede (Hist. Eccl. 4, 6) says that Peterborough was among the Gyrwas. Their name would appear in such places as Girton, Glitton and Guyerne in Cambridgeshire (P-N Cambridgeshire 177, 64 and 293 respectively); and Girton, Nottinghamshire (P-N Nottinghamshire 204).Google Scholar

⁴⁸ Both hundreds of Rutland are called Wiceslea. Notice also Wixamtree, Bedfordshire (P-N Bedfordshire 27); Wixenabroc (P-N Worcestershire 16).Google Scholar

⁴⁹ Spalding, Lincolnshire, is the most obvious place. A Spalding River was in the area. Notice also a Spaldwick (P-N Huntingdonshire 247) and Spalford (P-N Nottinghamshire 207).Google Scholar

⁵⁰ Unfortunately no Place-Name Society volume for Leicestershire has appeared yet.Google Scholar

⁵¹ The mistake of an ‘e’ for an ‘s’ was made in the case of Sudrica-Hendrica. For Hurstingstone see P-N Bedfordshire–Huntingdonshire 203; Anderson, O. S., The English Hundred Names (Lund 1934) 109.Google Scholar

⁵² P-N Bedfordshire–Huntingdonshire 190. Note also Swordersland (P-N Essex 483).Google Scholar

⁵³ For Gifla, P-N Hertfordshire xvii; Bedfordshire–Huntingdonshire xviii-xix. Note also Yelden (ibid. 20) in Bedfordshire; Ivinghoe and Iver in Buckinghamshire (P-N Buckinghamshire 96 and 239); Ivel River (P-N Hertfordshire 3); Willingham (P-N Cambridgeshire 173). For Hicca, P-N Hertfordshire xvii. Notice also Hinchinbroke (P-N Huntingdonshire 261); Hitchin (P-N Hertfordshire 8); Ixhill and Ixhill Hundred (P-N Buckinghamshire 127 and 114).Google Scholar

⁵⁴ Harrow is usually Herghe or some similar form in the Middle Ages. (P-N Middlesex 52, 53). The hundred in which Harrow lies is Gara ( ibid. 50).Google Scholar

⁵⁵ The Feppinga are said by Bede ( Hist. Eccl. 3, 21) to be among the Middle Angles. Bede apparently never mentions the Middle Saxons so this cannot be used to exclude the Feppinga from Middlesex. The Tribal Hidage has the statement inserted as an interpolation: it probably comes from Bede. Possibly Epping comes from the Feppinga. Or Fobbing (P-N Essex 156) or Punge Wood (Brownbill 499). There are many places beginning with Billing which might possibly come from this tribal name. One of the most tempting is Belymyle (P-N Cambridgeshire 292). See also P-N Northamptonshire xiv.Google Scholar

⁵⁶ (P-N Cambridgeshire 19.) Perhaps also Widdington (P-N Essex 545); Wethersfield (ibid. 465); Widford, Windridge (P-N Hertfordshire 211 and 92); Wittering, Werrington (P-N Northamptonshire xiv, 247).Google Scholar

⁵⁷ The present Willey Hundred (P-N Bedfordshire 25). See also Welwyn (early form Wilga) (P-N Hertfordshire 144). Possibly also Widford, Willingale (P-N Essex 275, 500); Willen (P-N Buckinghamshire 26); Wilbury (P-N Hertfordshire 115); Wilden, Willington (P-N Bedfordshire 66, 99); Willingham (P-N Cambridgeshire 117); Willesden (P-N Middlesex 160).Google Scholar

⁵⁸ Chadwick, , Studies 264–5.Google Scholar

⁵⁹ Stenton 55–7; Hodgkin I, 137–47.Google Scholar

⁶⁰ Gray, H. L., English Field Systems (Cambridge, Mass. 1915) ch. ix.Google Scholar

⁶¹ Siebs, T., Geschichte der englisch-friesischen Sprache (Halle 1886) 9.Google Scholar

⁶² Ibid . 10. The Frisians used geographical directions with their ga's but that was common enough. Fivelga resembles Gifla, to some extent. The Aro saetan, if we have located them properly near Harrow, were also about the sanctuary which was called the Gumeninga Hergh (Stenton 54). Possibly Gumeninga is the same name as that of the great Frisian family of the Cammingha (Encyclopaedia Britannica under ‘Frisian Isles’).Google Scholar

⁶³ Stenton 56, 219.Google Scholar

⁶⁴ Lyons, J. E., ‘Frisian Place-Names in England,’ Publications of the Modern Language Association 33 (1918) 644–55. He finds Frisian very closely related to the Northern and Midland dialects of Middle English, just the areas of the small tribes.Google Scholar

⁶⁵ Stenton 5.Google Scholar

⁶⁶ Little is known of the West Hecana group which Stenton calls the Magonsaetan (Stenton 46–7). Ammianus Marcellinus tells of the disbanding of the Arcani in the fourth-century England because of their treachery along the frontier. One would like to think that they might have been the founders of Archenfeld and of the West Hecana, but they seem to have been a group of messengers rather than a tribal group. Collingwood and Myres 284–5.Google Scholar

⁶⁷ The situation in Bernicia is described by Hodgkin I, 147–53.Google Scholar

⁶⁸ Collingwood and Myres 287. There was also a cohors I Frisiavonum in England. Enciclopedia Italiana under ‘Frisia’.Google Scholar

⁶⁹ Stenton 4–8.Google Scholar