446
NOTES AND QUERIES. [io s. XIL DEC. 4, im
perhaps, from the thickness of the strokes,
with a brush rather than with a pen. The
decipherment of these protocols is one of
the great difficulties of papyrology, and no
protocol of the Byzantine age has yet been
satisfactorily read. The locus classicus for
the protocol is Justinian's Novell. Ixi.
chap. 2 (so in the Teubner text ; in the
Leipsic edition, ' Corpus Juris Civilis,'
vol. iii., 1849, the edict is numbered xliv.).
This is to the effect that it was illegal to
draw up any contract unless the protocol
was left attached to the papyrus.* The
law was, however, habitually broken.
After the Arab conquest of Egypt, the Arabs kept up the papyrus monopoly and the practice of writing protocols ; but, while retaining the illegible Byzantine script at the beginnings and ends of lines, they introduced slightly more legible formulae in Greek and Arabic. These consist of Mohammedan formulae and the name of the Governor or Khalif, or both ; sometimes also the date. The Arabic name for protocol is tirdz.
It should be emphasized that this sense of the word protocol, is in no degree covered by the definition quoted from the ' N.E.D.' The protocol had no connexion with the document written on the papyrus to which it was attached ; it was written in the papyrus factory at the time of manufacture, and a piece of papyrus might of course be kept for some years before it was used. More- over, protocols are found attached to MSS. other than legal. Thus one of the finest known specimens of a protocol of the Byzantine type is that on the first folio of B.M. Or. MS. 5001, a papyrus codex of Coptic homilies.
For English use of the word reference may be made to H. I. Bell, ' The Aphrodito Papyri' (Journal of Hell. Studies, xxviii. 97-120), p. 109, " It has a protocol appa- rently dated in the governorship of 'Abd- allah," and to W. E. Crum's ' Catalogue of Coptic MSS. in the British Museum,' London, 1905, and ' Catalogue of the Coptic MSS. in
the John Ry lands Library,' Manchester
and London, 1909 (see the indices to those volumes). For discussions of protocols in general see especially J. von Karabacek,
Papyrus Erzherzog Rainer : Fiihrer durch die Ausstellung ' (Vienna, 1894), p. 17f. ;
Arabic Palaeography' (Vienna Oriental
- Of. the passage quoted from ' Schol. Basilic.,'
hb. xxii. p. 94, by Wessely in his ' Prolegomena ad Papyrorum Graecorum Novam Collectionem Eden- dam (Vienna, 1883), p. 5.
Journal, xx. 131-48), p. 139 ff. ; Sitzungs-
berichte d. Kai's. Akad. d. Wissenschaften
in Wien, Phil.-Hist. Klasse, 161, 1, 'Die
arabischen Papyrusprotokolle ' ; C. H.
Becker, ' Arabische Papyri des Aphrodito-
fundes ' (Zeitschrift fur Assyriologie, xx.
68-104), p. 97 ff. ; ' Das Lateinische in den
arabischen Papyrusprotokollen ' (ib,, xxii.
166-93) ; and H. I. Bell, ' Latin in Protocols
of the Arab Period ' (Archiv fiir Papyrus-
forschung, v. 143-55). H. I. B.
WALSH SURNAME : A NEW THEORY. Years ago, when I first heard from Lan- cashire men that the termination -halgh, which occurs in many of their place- and personal names, was pronounced halsh, I wrote to ' N. & Q., J and several readers replied and confirmed this (9 S. i. 345 ; ii. 15, 115). The change of gh to sh puzzled me, but I now think I can solve the enigma.
Every student of German is aware that the ch in that language has two sounds, hard and soft, and that in the combination Ich the latter is correct. Such words as Milch and Dolch sound to an English ear like milsh and dolsh. I feel quite happy in comparing this with the pronunciation of halgh as halsh. It is curious that no mention is made of this sound in Prof. Wright's ' English Dialect Grammar. l It is one of the most interesting phenomena in the whole range of English phonetics.
The question naturally arises, If halgh, which represents A.-S. healh, can be pro- nounced halsh, are there no other examples of the same development ? One thinks of the A.-S. Wealh, a Welshman, which would become Walsh. Walsh is a well- known surname, hitherto looked upon as a variant of the adjective Welsh.
I venture to suggest an altogether new theory, viz., that Walsh is the modern form of the A.-S. noun Wealh, whereas Welsh is the A.-S. adjective Welisc. Perhaps PROF. SKEAT will give us his opinion of the lines in ' Piers Plowman ? :
Godefray of Garlekhithe And Griffyn the Walshe.
On my hypothesis, this would be A.-S. " Griffyn se Wealh. n Unfortunately, it will be some time before the ' N.E.D. * will have to discuss this, under the letter W.
JAS. PLATT, Jun.
DE RAET BARONETCY. At 6 S. v. 267 there was an inquiry as to the existence of this baronetcy. The Sir Dirck de Raet there referred to was a grandson of the original grantee, and d.s.p.m. at Leyden, 16 Aug.,
