without foundation. Whereas Björkman in his Nordische personennamen and Zur englischen namenkunde derived Algod, Sigod, and Wigod from Scandinavian names in gautr, -got, looked upon Winegod as eine hybride bildung« etc. (see Nord. pers. 177), hesitatingly connected Heregod with Scand. Hergautr (-god scheint sonst immer aus nord. -gautr, -gotr zu stammen (!), Nord. pers. 68), and did not consider it excluded that Is(e)god might be connected with an unrecorded Scand. *Ísgautr1), he now derives Algod, Sigod, Wigod, Winegod, Heregod, and Isgod from the corresponding Continental names in gaud. The Scand. origin of these names has now been reduced to >eine schwache hauptsächliche theoretische möglichkeit (p. 175; cf. also p. 179). I also considered a derivation from Cont. -gaud possible for all these names 2), with the exception of Wigod, which to judge by the references given by Förstemann was not very usual on the Continent. Considering the frequency of Continental names in OE. there need no longer be any hesitation on this account.

Nevertheless there are serious phonological difficulties in deriving the names in god from Germ. gaut. B. (p. 167) gives a survey of the early spellings in which gaut appears on Frankish territory (chiefly after Förstemann and Franck). These spellings are, of course, to a great extent archaic and therefore no safe guide to the pronunciation. What we want to know. is how Germ. gaut was pronounced on Frankish (esp. West Frankish) territory from about the 10th cent. (the date when such names first appear in English) and onward. Such forms. as French gaus (< gaudus) and Frankish -goz (see Franck, $32, n. 3) presupposing a French 3) pronunciation gos and a

1) I cannot see any other meaning in B.'s remarks on this name, Nord. pers. 194 and n.

2) In his account of my views (p. 162) B. tries to disguise this fact by laying the main stress on my assumption of the existence of an OE. inpedendent terminal god. He does not even mention that I derive Algod and Heregod not only from OE. forms in god but also from Adalgaudus and Hairgaudus, and he adopts similar tactics when dealing with the names separately, pp. 171-175.

3) B. objects to my calling names found on coins about 900 Continental French. Frankish was spoken to a great extent in the old Kingdom of the Franks. Hence names which show no traces of French sound-laws, should be called Frankish not French. I do not deny that »viel fränkisch gesprochen

Frankish pron. goz, are all but missing in English. The only traces of such forms are Wengos, Wynegos and Wngos on

coins.

Such 10th cent. forms as gaud are evidently archaic spellings for god, got, which occurred, though more sparingly, by the side of -gaud (Franck, ibid.). Both in French and Frankish the vowel was probably pronounced as a long ō, which was identified with OE. long ō and consequently appeared as (u:) in mod. Engl. (cf. Allgood < Adalgaudus, Hargood< Hairgaudus). The dental must have been pronounced differently in French and in Frankish. In French it was pronounced either as (p) (d) (Nyrop, §§ 387, 395) or (t) (Kalbow, 132 and refs). A French pron. with (d) is hardly conceivable, for in the case of late loans the final (d) (in god) was no doubt unvoiced in French. When final, d in English names was treated in the same way; cf. Stolze, § 37; Rathmann, § 33. Nor are any forms in god, as far as I know, found in OFr., e. g. in the sources examined by Langlois and Kalbow. Here gaut appears invariably as -got1). In Frankish the dental was evidently pronounced either as (t) or (d) (Franck, § 90). The origin of such forms as -gaud <gaut is obscure. Förstemann thinks -gaut has been influenced by Latin gaudere, gaudium. In all probability the Frankish forms in (d) were in course of time superseded by those in (t), the natural tendency of the various German dialects being to unvoice the final stops.

wurde, but at the same time French seems to have been in common use as early as the beginning of the 9th cent. (Nyrop, I p. 11), and consequently it cannot be incorrect to call names which were introduced into England from France in the 10th and 11th cents Continental French, so much the more as the Roman nomenclature in France was rapidly superseded by the Germanic names (Forssner p. XXXVII). Forssner (p. LXI) also points out that many of the moneyer's names exhibit unmistakeable Romance features. Such features are also found in some Anglo-Saxon names on coins. See my paper, E. St. 50, 357. Hence Aegel on coins (for references see Forssner p. 13 f may sometimes be due to the Anglo-French development epel > æel > ail which I have dealt with in A.-N. Infl. p. 101 ff.

1) Instances of OFr. names in got are: Eliot <*Eligaut, Goniot < Wungaud (better: Wunigodo, Förstemann), Guinchot < *Winigaud, Maingot < Magingaut, Helgot < Hildgaut, Manigǝt < Manegaud, Saligot <? *Saligaut (cf. the names under salva, Förstemann) (all noted by Kalbow), Anget (mod. Fr. Angot, Ango) < Scand. Asgautr and Cont. Fr. Ansegaudus, Turgot (mod. Fr. Turgot) < Scand, þorgautr (Fabricius 220, 223, 228),

The pronunciations of Germanic gaut which were likely to occur on Frankish territory in the 10th cent. were consequently gōs, gōz, göp (gōd), got and god, which would have appeared in English as gōs [on English (s) for (ts), see my Anglo-Norman Infl. p. 37]. gōp, gōt, gōd. Of these forms the first is almost missing (cf. above p. 195, n. 2)), the second is missing, the third is comparatively rare, especially when we consider that Engl. got in certain names (e. g. Thurgot, Osgot) is sometimes due to Scand. -got, and that got in some records, mainly those of post-Conquest date, may be an Anglo-French spelling for god. How are we to account for the predominance of god? If we do not care to reckon with an independent OE. terminal god, we are bound to assume that the Continental name-element gos, got, god was often interpreted as and identified with the OE, adj. god, which was usual in the first element of many OE. personal names (see my paper E. St. 50, 353). To the English linguistic feeling these names were actual compounds with the adj. god.

This also accounts for the comparative scarcity of got in later ME. And with one exception (Wigget < Wygot) all the names which have been kept to the present day, exhibit -good in the second element (see E. St. 50. 346, 351) 2). It is only on the assumption of such a transformation of the Cont. suffix that I can agree with Björkman that the majority of the early English names in god, got are to be derived from Cont. names in gaud etc. Nor is it in my opinion excluded that such of these names as are rare or missing on the Continent (esp. in West Frankish) may be hybrid English formations (e. g. Leofgod, Wigod < OE. Leof, Wig + gōd). There is at least one name which perhaps is best explained thus, viz. Godgod. Considering the occurrence of similar reduplicated names on the Continent (Berdbert, Herhere, Wololf, Socin 198), there can be little doubt that Godgod is an actual English name composed of gōd + gōd. Cf. OE. Wulfwulf formed on the same principle.

1) In other English names of Cont. origin z, s, c<t are often found (Forssner p. 272).

2) Elgood and Elegood should be kept distinct from Allgood < Adaigaudus. They are due either to ME. Helgot, OFr. Helgot < Hildegaud (Forssner 146) or to Eligaud, Pol. Irm.

According to B. (p. 175) the names Thurgod and Osgod can only be of Scand. origin. The suffix got was changed into god on the analogy of the Continental names in which -god interchanged with -got. In my opinion the Continental suffix god, which as we have already seen had been identified with the OE. adj. -gōd, was substituted for Scand. -got1).

There seems to be no foundation whatever for B.'s theory that Thurgot was changed into Thurgod in Normandy (p. 176). As has been already shown (p. 198), no forms with -d are likely to have existed in spoken French. Nor is there as far as I have been able to ascertain any documentary evidence of such forms.

Fabricius (pp. 228, 233, 234) gives no other early spellings than Turgot, Ansgotus, and Angotus for the Norman names Turgot, Angot, Ango. Angoville (Eure, Calvados etc.) appears as Ansgotivilla, Ansgothivilla 11th cent., Angotivilla, c. 1110, Dict. Topogr. du département de l'Eure, p. 5, Dictionnaire Topogr. du département du Calvados, p. 5, etc., and Mesnil. Angot is Mesnillum Ansgot in the Black Book (1251) (cf. Fabricius p. 245). Cf. also the OFr. instances of names in -got noted on p. 198 n. If sporadic forms with d occur, they are to be interpreted as inverted spellings, where d stands for (t). Cf. E. St. 50, 351. The transformation of got into -god was exclusively an English phenomenon, and would not have taken place, had it not been for folk-etymological associations with the adj. god. OE. Thurgot and Thurgod may of course in some instances go back to the Norman name Turgot. Neither do I think B. is right in claiming an exclusively Scandinavian origin for Osgod, Osgot). Considering the popularity of the name Ansgot < Continental Ansegaudus and Scand. Ásgautr 3) on the Continent and especially in Nor

1) The Continental suffix -göt, -gōd was more readily identified with OE. god than the Scand. terminal -gỗt. Skand. ursprung ist mir sicher. Die aus dem Kontinentalgerm. stammenden Ans(e)gaud, Ans(e)gaus kommen hier natürlich nicht in betracht.<<

2) B. says:

3) To account for Ans- < Ás- it is not necessary to assume that the vowel in As was nasalized. An analogical transformation has taken place of the same character as in the Continental names Anschericus (OFr. Anchier), Ansculfus (mod. Fr. Ancou, Ancoul) for Ascaricus, Asculjus and perhaps

mandy, there can be little doubt that Osgod is sometimes an anglicized form of this, name. It has been shown (cf. E. St. 50, 356) that OE. Osgod Scand. Asgot in northern English place-names was often replaced by the French name An(s)got, and it goes without saying that this process might sometimes be reversed. Hence Osgod is one of the most remarkable names in OE. Sometimes it is an anglicized form of French An(s)got, in other cases its ultimate source is Scand. Ásgautr, Asgot, which was first anglicized into Osgot and then turned into Osgod by analogy of the many names in -god introduced from the Continent.

Like Björkman I am now of the opinion that the majority of early English names in got, god are to be derived from Continental names in gaud, -god-, -got etc., with the important addition, however, that this Cont. terminal was influenced by the OE. adj. god. Consequently Algod(us) < Adalgaudus, Fredegod < Fredegaudus, Heregod < Hairgaudus, Is(en)god< Isengaudus, Leofgod < Liubgoz or an English hybrid formation (cf. above p. 199), Mangod etc. < Manegaud, Sigod < Sigaud, Sunegod Sunigaut (Forssner 226), Waringod < Warengaudus, Wigod < Wilgoz or an English hybrid formation, Winegod Winegaudus. In my previous article (pp. 341-349) this derivation was suggested or taken into consideration for all these names (with the exception of Sunegod), although only as an alternative. Some forms of Thurgot, Thurgod, Osgot, Osgod may likewise be of Continental origin, but the majority of them are Scandinavian.

The critical remarks B. bestows on my article with regard to other points than the god from -*gōđa theory are for the most part not well founded.

B. considers my explanation of gote in earcongote, Bede, erroneous. He gives a reference to Forssner (75, n), who likewise derives the name from got related by gradation to -gaut. Only, Forssner thinks the name is Continental, whereas I look upon it as a native correspondent to the Continental names in got. It is not certain that "the name-element -got is strange to the OE. pers. nomenclature". Cf. my remarks on Gota, p. 350. I fail to see what is wrong in my explanation of Hrodear,

Anselin for Ascelin. Cf. my Notes on Early English Personal Names p. 283 f. and the references given there.