CELTIC SYSTEM IN ENGLAND 269 



has determined the localities in which such a usage prevailed and 

 the degree to which Celtic influence is responsible therefor/ no 

 generalizations are possible. From what evidence we have it 

 would seem that in most districts of the north and west the sub- 

 division of socage and villein holdings, if it ever prevailed, early 

 gave place to impartible succession, the custom which from the 

 thirteenth century at least was usual in the midlands. 



Of greater assistance in estimating Celtic influence upon the 

 field system of English counties is a third trait of Celtic agricul- 

 ture. This is the irregular disposition of the scattered parcels of 

 the tenants' holdings throughout the cultivated area of a town- 

 ship; for, if it be assumed that an early subdivision of the land 

 among co-heirs became permanent in the AngHcized border 

 counties, such disposition would be for us the only reminder of 

 the earlier field-history of the region. If it chanced that the 

 dispersed parcels were in certain places reconsolidated (or if in 

 some townships a division had never taken place), we should 

 expect to find there enclosed areas. When, consequently, Devon, 

 Cornwall, and Cheshire appear as counties largely enclosed in the 

 sixteenth century, this phenomenon is explicable as a normal 

 manifestation of the Celtic system. If some traces of arable 

 common fields still remained within their bounds, these too are 

 normal phenomena. Although in the southwest some such fields 

 may have been due to improvement of marsh or down-land, other 

 tracts are not easily so explained. At Brixham in Devon, as 

 well as at several places in Cheshire, there seem to have been 

 ancient arable fields that had long been characterized by inter- 

 mixed holdings. Cheshire terriers of the thirteenth century give 

 details which enable us to see that these fields were not of the 

 midland type. In structure they were, on the contrary, like 



^ Two vague passages in the laws of Cnut which may imply that partible suc- 

 cession was the Anglo-Saxon usage of the eleventh century are as follows: " [If a 

 man die intestate] Ac beo be his dihte seo aeht gescyft swytSe rihte wife 7 cildum 7 

 nehmagum, aelcum be J^aere maeSe, |'e him to gebyrige. . . . And se man, |7e on 

 |)am fyrdunge aetforan his hlaforde fealle, . . . fon j^a erfenuman to lande 7 to 

 aehtan 7 scyftan hit swyt5e rihte " (Cnut II 70, 78 [1027-1034], Liebermann, 

 Gesetze, i. 356, 364). Chapter 34 of the Lets Willelme (1090-1135) is of similar 

 purport: " Si home mort senz devise, si departent les enfans I'erite entre sei per 

 uwel " (ibid, i, 514). Cf. Maitland, Domesday Book and Beyond, p. 145. 



