I yield to no one in admiration for Mr Seebohm's work, but the question raised is so important that accuracy as to the fact is here essential. (1) Sullung is nowhere found in Essex, but only solanda; (2) Solanda does not occur 'in the Manors' referred to, but at Tillingham alone; (3) In Essex it nowhere 'takes the places of the hide', as it does in Kent; (4) The Essex instance adduced by Professor Vinogradoff is taken from a Manor where solanda does not occur.
Two issues—quite distinct—are involved. In the first place, Mr Seebohm contends that Professor Vinogradoff must not argue from 'the custom of Kent' to the rest of England, because (inter alia) Kent, unlike the rest of England, was divided into sulungs, which points to some difference in its organization.[197] This contention is sound, and is actually strengthened if we reject the identity of sulung and solanda. But, in the second place, he endeavours to explain away the Essex case of subdivision at Eadwulfsness, to which the Professor appeals, by connecting it with the Kentish system through the term solanda. This, as I have shown above, is based on a misreading of the evidence, and is contrary to the facts of the case.
Let us then look more closely at the Essex instance of subdivision. It is taken from one Manor alone, the great 'soke' of Eadwulfsness, in the north-east corner of the county. This 'soke' comprised the townships of Thorpe 'le soken', Kirby 'le soken', and Walton 'le soken' (better known as Walton-on-the-Naze). Such names proclaim the Danish origin of the community, and it is noteworthy that the 'hidarii', on whom the argument turns, are found only at Thorpe and Kirby, the very two townships which bear Danish names. This circumstance points to quite another track. That the system in this little corner of Essex was wholly peculiar had been pointed out by Hale, and it might perhaps have originated in the superimposition of hides on a previous system, instead of in the breaking up of the hide and virgate system. But this is only a conjecture. The two facts on which I would lay stress are that at Thorpe, according to Hale, 'the holders of the nine hides (in 1279) possessed also among them seventy-two messuages', which, by its proportion of eight to the hide, favours Mr Seebohm's views; and that the holdings of the 'hidarii' were rigidly formed on the decimal system (such as 60, 30, 15, 7½ acres, or 40, 20, 10, 5 acres),[198] unlike the holdings of an odd number of acres on the Kentish Manors of St. Augustine's. The reason for the Essex system was clearly the necessity of keeping the holdings in a fixed relation to the hide, that their proportion of the hide's service might be easily determined. These two points have, perhaps, I think, been overlooked by both of the eminent scholars in their controversy.
Before leaving the subject of the sulung, one should mention perhaps that it was divided (as Mr Seebohm has explained) into four quarters known as juga, just as the hide was divided into four virgates. Mr Seebohm bases this statement on Anglo-Saxon evidence,[199] but it is abundantly confirmed by Domesday, where we read of Eastwell (in Kent): 'pro uno solin se defendit. Tria juga sunt infra divisionem Hugonis, et quartum jugum est extra' (i. 13). So far all is clear; but Professor Vinogradoff, on the contrary, asserts that 'the yokes (juga) of Battle Abbey (in Kent) are not virgates, but carucates, full ploughlands' (p. 225). This assertion is based on a very natural misapprehension. In the Battle Manor of Wye (Kent) we find that the jugum itself was divided into four quarters, called 'virgates' which were each, consequently, the sixteenth, not, as in the hidated district, the fourth of a ploughland. Professor Vinogradoff, naturally assuming that the 'virgate' meant the same here as elsewhere, inferred that four 'virgates' (that is, a jugum) must constitute a full ploughland. But this change of denotation goes further still. The Battle Cartulary records yet another 'virgate', namely, the fourth (not of a ploughland, but) of an acre! This led me, on its publication, to wonder whether we have here the clue to the origin of the somewhat mysterious term 'virgate'. Starting from the acre, we should have in the virgata (rood) its quarter, with a name derived from the virga (rod) which formed its base in mensuration. The sense of 'quarter' once established, it might be transferred to the quarter of a jugum, or the quarter of a hide. This is a suggestion which, of course, I advance with all diffidence, but which would solve an otherwise insoluble problem. The relation of the bovate to the carucate, and of the jugum to the sulung, are both so obviously based upon the unit of the plough-team that they raise no difficulty. But the term 'virgate' does not, like them, speak for itself. If we might take it to denote merely a 'quarter' of the hide, it would become a term of relation only, leaving the 'hide' as the original unit. Should this suggestion meet with acceptance, it might obviously lead to rather important results.
Mr Elton, in his well-known Tenures of Kent, attaches considerable importance to a list, 'De Suylingis Comitatus Kanciæ et qui eas tenent', in the Cottonian MS., Claud. C. IV, which he placed little subsequent to Domesday. Having transcribed it for collation with the Survey, I came to the conclusion that it was not sufficiently trustworthy for publication, for the names, in my opinion, involve some anachronism. The feature of the list is that it shows us as tenants-in-chief, the leading tenants of Bishop Odo; and the change of most interest to genealogists is the succession of Patrick 'de Caurcio' to the holding of Ernulf de Hesdin.
XIV. THE 'FIRMA UNIUS NOCTIS'
The curious and evidently archaic institution of the firma unius noctis was clearly connected with the problem of hidation. In Somerset the formula for a Manor contributing to this firma was: