606 REVIEWS OF BOOKS October It is Dr. Ekwall's merit that he has recognized, as few of his predecessors have done, that for the interpretation of place-names linguistic methods are inadequate, and may often be misleading unless supplemented by topographical and historical research. He very frequently appeals to the physical character or situation of a place as a means of deciding between two alternative etymologies of its name. The names of rivers or streams and hills, which writers on place-names usually neglect, are carefully enumerated, and such etymological light as is available is brought to bear on them. One advantage that Dr. Ekwall has over those who have previously written on Lancashire names is that he has sufficient knowledge of Celtic philology to enable him to discuss with some profit the British element in the nomenclature of the county, which has hitherto been almost wholly ignored. The names treated of in the book are arranged not alphabetically but in geographical order, according to the hundreds and the (original) parishes. The advantage of this method may be seen from the concluding chapter entitled ' Summary of Results ', which shows the relative frequency of Anglian, Scandinavian, and British names in the various parts of the county. Dr. Ekwall recognizes the need for caution in drawing historical inferences from the etymological evidence ; but there can be no doubt that this evidence is valuable if wisely used. The geographical method of arrangement has sometimes a critical value ; when on formal grounds a name seems to admit of equally probable explanations from more than one language, it may be of consequence to know what is the linguistic character of the names in the immediate neighbourhood. As Lancashire has very few early records, and there is no likelihood that many more will ever be discovered, the etymology of a large pro- portion of the names of the county will probably always remain uncertain. Hardly any of the names are mentioned in pre-Conquest writings ; a few occur in Domesday Book, but even there the spelling is often phonetically ambiguous ; the great majority are traceable no further back than the thirteenth or fourteenth century. Further difficulty arises from the mixed character of the Lancashire local nomenclature, to which three languages Anglian, Scandinavian, and British have contributed. It is therefore not Dr. Ekwall's fault that he often hesitates between two or more alter- native etymologies, even where previous investigators have seen no reason for doubt. Scholars who are aware of the difficulties of the inquiry will probably be surprised at the large amount of secure positive result that he has been able to attain by the skilful use of topographical and historical data, and of the analogies of local nomenclature in Britain and Scandinavia. Even where the precise etymology of a name remains uncertain, he has often succeeded in determining with a high degree of probability the language to which it belongs ; and this is very frequently the most im- portant matter from the historical point of view. As a model of scientific method and thoroughness of research this admirable volume cannot be too strongly commended to all future in- vestigators of English local etymology. HENRY BRADLEY.
