Button.

But the names exemplified above are very thinly represented in early
records, and, though their existence in surnames derived from
place-names (Dodsley, Bagshaw, Bensted, Bedworth, Cobham, Ebbsworth,
etc.) would vouch for them even if they were not recorded, their
comparative insignificance is attested by the fact that they form very
few derivatives.

Compare, for instance, the multitudinous surnames which go back to
monosyllables of the later type of name, such as John and Hugh, with
the complete sterility of the names above. Therefore, when an
alternative derivation for a surname is possible, it is usually ten to
one that this alternative is right. Dodson is a simplified Dodgson,
from Roger (Chapter VI); Benson belongs to Benedict, sometimes to
Benjamin; Cobbett is a disguised Cuthbert or Cobbold (cf. Garrett,
Chapter II); Down is usually local, at the down or dune; Dunn is
medieval le dun, a colour nickname; names in Ead-, Ed-, are usually
from the medieval female name Eda (Chapter VI); Sibbs generally
belongs to Sibilla or Sebastian; Tait must sometimes be for Fr. Tete,
with which cf. Eng. Head; Tidd is an old pet form of Theodore; and
Wade is more frequently atte wade, i.e. ford. Even Ebbs and Epps are
more likely to be shortened forms of Isabella, usually reduced to Ib,
or Ibbot (Chapter X), or of the once popular Euphemia.

To sum up, we may say that the Anglo-Saxon element in our surnames is
much larger than one would imagine from Bardsley's Dictionary, and
that it accounts, not only for names which have a distinctly
Anglo-Saxon suffix or a disguised form of one, but also for a very
large number of monosyllabic names which survive in isolation and
without kindred. In this chapter I have only given sets of
characteristic examples, to which many more might be added. It would
be comparatively easy, with some imagination and a conscientious
neglect of evidence, to connect the greater number of our surnames
with the Anglo-Saxons.

Thus Honeyball might very well represent the Anglo-Sax. Hunbeald, but,
in the absence of links, it is better to regard it as a popular
perversion of Hannibal (Chapter VIII). In dealing with this subject,
the via media is the safe one, and one cannot pass in one stride from
Hengist and Horsa to the Reformation period.

"HIDEOUS NAMES"

Matthew Arnold, in his essay on the Function of Criticism at the
Present Time, is moved by the case of Poor Wragg, who was "in
cus