pl. in s all the more inexcusable, for his MS. annotations of his grammar argue that he does not wish to put himself, like some etymologists, supra grammaticam, although, it is true, he displays some impatience of phonological restraint. His argument that, if a knowledge of the non-existence of an A.-S. gen. pl. in s could have been obtained so easily as I said, it is not “likely that any reasonable person would avoid seeking it,” is more amusing than conclusive.
Finding that the facts do not agree with his views, Mr. Addy attempts to get over them by an assertion that it is difficult to reconcile with any respect for A.-S. grammar. He tells us that it seems clear to him that both Huns-ton and Hun-ton represent an A.-S. *Húna-tún, “town of Huns.” It is manifestly wrong to state that the Domesday Hunes-tune represents an A.-S. *Húna-tún, but Mr. Addy attempts to justify this assertion by saying that the old inflections “were dying out or changing to newer forms” when Domesday was compiled. This is one of those vague, unsupported assertions with which we are only too familiar in local etymology, and, like most of these shadowy generalizations, it is entirely wrong. In the first place, the names in Húnes, &c., do not depend solely upon the testimony of Domesday, for I quoted several A.-S. instances; secondly, even if the gen. pl. in s had been in common use in 1086, it would not support Mr. Addy, for these names were compounded centuries before that date; and, finally, there is not the slightest evidence of the existence of this gen. pl. when Domesday was compiled. There is, therefore, absolutely no reason for holding that the Domesday Hunes-tune represents an A.-S. *Húna-tún; and there is very little more reason to believe that the Yorkshire Hun-tone comes from this *Húna-tún. According to the phonology of Domesday, this latter name would appear as *Hune-tune or *Hune-tone, not as Hun-tone. And even if *Hune-tune existed, it would not benefit Mr. Addy’s case, for such a form would also represent an A.-S. *Húnan-tún, from the personal name Hún-a.[1] This *Hune-tone is precisely the form we should expect *Húnan-tún to assume in Yorkshire, for Northumbrian began to drop the n of the weak declensions so early as Bede’s time. Moreover, the Domesday scribes frequently represented the weak gen. an by e, even in cases where we can prove that the full form still existed at that time. Thus the A.-S. Huntan-dún, Huntingdon, is spelt Hunte-dun in the Survey; the Derbyshire Willington is given as Wille-ton;[2] the Staffordshire Bednall appears as Bede-hala (=*Bédan-heall); and the Lincolnshire Bucknall (=*Buccan-heall) is spelt Buche-hale.[3]
So far from Hun-tone representing *Húna-tún, it is clearly equivalent to *Húnes-tún, and it thus supports my contention. My studies of Domesday phonology soon led me to perceive that the scribes of that work frequently omitted the gen. es. I select, to prove this, a few examples of local names compounded with personal names, since these witnesses are free from doubt. In the following table I have placed the modern name in the first column, the Domesday form in the second, and the personal name in the third:—
| Nottinghamshire. | ||
| Thoro-ton | Toruer-tune | O.N. þor-varð-r. |
| Tor-worth | Turde-worde | O.N. þórð-r. |
| Egman-ton | Agemun-tone | O.N. Og-mund-r. |
| Osuui-torp | A.-S. Os-wíg. | |
| Derbyshire. | ||
| Alkman-ton | Alchemen-tune | A.S. Ealh-mund |
| Lincolnshire. | ||
| Asgar-by | Asgere-bi | O.N. As-geirr. |
| Aslack-by | Aslache-bi[4] | O.N. As-lák-r. |
| Hawer-by | Hauuarde-bi | O.N. Há-varð-r. |
| Thurlby (Bourne) | Tvrolve-bi, Torulf-bi | O.N. þór-ólf-r. |
These names suggest that the English in forming local names followed the old Teutonic (and Aryan) system of using the stem as the compounding form. But it is evident from the A.-S. charters that they invariably used the later system of compounding with the gen. for this purpose, for amongst the hundreds of local names recorded there are only one or two dubious instances where the gen. of the personal name is wanting. Hence we may conclude that the gen. es originally formed part of the names in the above cases, although it is omitted by the Domesday scribes. We have, fortunately, several instances where the Survey gives two forms of the names of certain villages—one with and the other without the gen. sing. Here are a few examples:—
| Nottinghamshire. | ||||
| Thurgar-ton | 
|
Turgars-tone | 
|
O.N. þor-geir-r. |
|
Torgar-tone | 
| ||
| Aslock-ton |
|
Aslaches-tone |
|
O.N. As-lák-r. |
|
|
Aslache-tone |
| ||
| Lincolnshire. | ||||
| Audle-by |
|
Adulfes-bi |
|
A.-S. Eald-wulf. |
|
|
Aldulue-bi |
| ||
| Aud-by |
|
Aluuoldes-bi |
|
A.-S. Ælf-weald. |
|
|
Aluuolde-bi |
| ||
| Osgod-by |
|
Osgotes-bi |
|
O.N. As-gaut-r. |
|
|
Osgote-bi |
| ||
- ↑ Compare, A.D. 943, Húnan-weg (‘Cart. Sax.,’ ii. 524, 9); A.D. 947, Húnan-héafod (‘Cod. Diplom.,’ v. 313, 13); and the Norfolk Hun-worth, which occurs in Domesday as Hune-worda, Hune-uurde, and Huna-worda, representing an A.-S. *Húnan-weorðig.
- ↑ This must, on the analogy of Huntingdon, represent an A.-S. *Willan-tún, from the personal name Will-a.
- ↑ This name occurs as Buken-hale in one of the spurious Croyland charters, dated 1051, in ‘Cod. Dipl.,’ iv. 126, 12, and as Boken-hale, A.D. 806, in ‘Cart. Sax.,’ i, 453, 28—one of the clumsiest forgeries in the collection. The Staffordshire Bucknall is called Bucken-ole in the Survey.
- ↑ Compare Aslaches-hou (now Aslacoe) Hundred in the same county, the Yorkshire Aslaches-bi, and the Nottinghamshire Aslock-ton in the next table.
