1921 REVIEWS OF BOOKS 299 item ' c. 750. Bede. Dunelma (latinized) '. In reality the name does not occur in any form before the eleventh century. What is very strange is that this mistake is copied from Dr. Zachrisson, who is usually very accurate in such matters. The article on Durham seems to have escaped revision, for Gaimar's form Durelme is misprinted Dunelme. The sugges- tion that Dunholm preserves the first syllable of the original British name is new to me, and appears very reasonable. Although hardly any of Mr. Mawer's explanations of place-names appear to me demonstrably wrong, there are several instances in which I think a different solution is either preferable or at least equally probable. The name Featherstone is found in Staffordshire, Yorkshire, and Lancashire as well as in Northumberland. The first element has com- monly been identified with the Scandinavian personal name Fathir. Mr. Mawer rightly objects that this is phonologically unsatisfactory, and that it would be strange indeed if so rare a personal name should appear in four different places in combination with stone ; only if Fathir were a mythic or legendary personage would such a coincidence seem credible. He suggests that a compound feather-stone may conceivably have existed, meaning either ' a stone moved as easily as a feather ', or 1 a stone marked with feather-shaped forms '. The former meaning seems too unlikely ; but according to the Dialect Dictionary the word is current in Devonshire as the name of a kind of marble. Possibly this may be the true explanation of the place-name. I would suggest, however, though rather diffidently, that an OE. *federstdn, ' tetralith ', would be an appro- priate term for what is popularly called a ' cromlech ' — a monument con- sisting of three upright stones and a headstone. The puzzling name Slaley was Slaveleia in 1166, and the v or u appears in all the forms down to the sixteenth century. The suggestion that the first element is the common word slave is surely inadmissible. The author himself points out that the word is not found in English before 1290 ; he might have added that it has not been found in French earlier than the thirteenth century, and that in English it was always spelt with scl- down to the sixteenth century. Further, that the word slave was ever applied in the Middle Ages to any class of persons existing in England is historically incredible. What the name really means I do not venture to guess. Barmoor, pronounced ' Baremoor ', and in the thirteenth century spelt Beiremore, Beigermore, is explained as containing the hypothetical personal name *Beaghere. This is formally possible, as the -es of the genitive some- times disappears. But why may not the name be beger-mor, * bilberry- moor ' ? Ousterley appears three times in parish registers (1369, 1391, 1429) as Houstre, Hustre, and the author takes the name to be ' house-tree ', for which he offers very unconvincing explanations. But the spelling of parish registers is often erratic, and Oustrefeld is found in a more trust- worthy source of 1382. As Ouster- or Auster- is a common initial element in place-names, the initial H may be disregarded. I would suggest eowestre (eawestre), ' sheepfold '. The curious place-name Unthank is found in many parts of England.
