412 ENGLISH FIELD SYSTEMS 



as in the midlands, the invaders impressed upon their new con- 

 quest an open-field system which, according to our earliest 

 evidence, was one of three fields. In another respect, however, 

 they seem to have adopted the habits of their predecessors: 

 their settlements were small and of the hamlet type. Perhaps 

 they assimilated a part of the Briton population itself along with 

 the Celtic type of settlement. Place-names here evince more of a 

 commingling of Celtic and Anglo-Saxon elements than is usual in 

 the midlands — a further indication that there was in these two 

 counties a more equitable balance of Celtic and Germanic forces 

 in matters of settlement and agriculture than appears elsewhere 

 in England, unless it be in Northumberland. 



Other counties of the west and north diverged more sharply 

 from the midland model. Often townships and settlements in 

 them were small, as in Celtic countries. In Cornwall, Devon, Che- 

 shire, and southern Lancashire open arable fields seem never to 

 have been numerous or, in any township, extensive. The same 

 cannot be said of Northumberland, Cumberland, and northern 

 Lancashire, where such fields were relatively frequent in the 

 thirteenth century and comprised the largest part of the tilled 

 land of each township. But whether large or small, numerous or 

 infrequent, the open arable fields of all these counties were not 

 of the midland type. In no instance (with perhaps a reserva- 

 tion relative to Northumberland) were they divided into two or 

 three equal parts to which the strips of each holding were equi- 

 tably assigned. In appearance they were more like Scottish or 

 Irish open fields, in which the strips were said to lie in runrig. 

 The underlying principle of runrig was the assigrmient to each 

 tenant of a share in every kind of soil within a township, when- 

 ever an occasion for distribution arose. Since the several qual- 

 ities of land were likely to lie in various parts of the cultivated 

 area, a scattering of parcels was to some extent the result. Re- 

 course to runrig, therefore, brought about either temporarily or 

 permanently a dispersion of the parcels of a holding. Yet there 

 was no guarantee that these would be as symmetrically located 

 throughout the arable area as they were in two- and three-field 

 townships. There might even occur a segregation of parcels, a 



