ANGLO-SAXON REMAINS the surface by the characteristic sand of the county. The geological aspect of the various localities in which Anglo-Saxon remains have been discovered is considered above, and is now referred to in order to empha- size the uniformity in more than one respect of the settlements north of the presumed line of the Tove valley and Watling Street. It is an easy step to the conclusion that these settlements were made by men of one race, with similar traditions and similar aspirations, and as conservative in their disposal of the dead as in their choice of a home for the living. That these men were Gyrwa is a conjecture that is not unreasonable in itself and would explain the apparent connection with East Anglia. It will thus be seen that there is some archasological warrant for dividing the county into three sections : the southern portion being characterized by burials of the entire body, with traces of West-Saxon influence ; a central area marked by cemeteries where the dead were buried entire or their cremated remains deposited in urns ; and the north-eastern extremity in the neighbourhood of the Ermine Street, where the few burials that have come to light present a want of uniformity that contrasts with the regular interments beyond Watling Street. It is a remarkable and perhaps a far-reaching coincidence that these divisions correspond closely with the areas of dialectical varieties within the county. This can be clearly seen by a reference to the map of English dialects prefixed to the standard work on this subject by a former president of the Philological Society.' Northamptonshire is divided between two main districts, the southern and the eastern, the dividing line roughly coinciding with the Watling Street in its passage through the county. Along the northern boundary of both Northamptonshire and Rutland runs the line between the eastern and the midland linguistic areas ; and run- ning parallel to this till it strikes across Huntingdonshire and Cambridgeshire to the Wash, is the line south of which the peculiar northern pronunciation of a test-word does not occur. The greater part of the county is therefore included in the debatable area in which the pronunciation of the test-word is variable ; and it is reasonable to suppose that this mingling of dialects is due to the absorption of West-Saxon districts north of the Thames by the Mercians during the seventh and eighth centuries. The area includes in any case east Gloucestershire, east Worcestershire and north Oxfordshire, all districts in which character- istic West-Saxon and Anglian remains have been met with in the cemeteries. And enough has already been said with regard to the apparent coalition of races in the western part of Northamptonshire, where the mingling of midland and southern dialects affords an exact parallel. All east of Watling Street, about three-quarters of the county, is included in the eastern dialect area, and a sub-dividing line from Rockingham to Fotheringhay separates the neighbourhood of Peter- borough from the central portion of the county, where cemeteries have been found exhibiting both methods of interment and suggesting a mixed ' English Dialects : their SouaJt and Hornet, by Dr. A. J. Ellis, 1890 (English Dialect Society). 251
