THE CELTIC SYSTEM 201 



seldom took place.^ From the limitations of outfield tillage 

 another escape, it seems, was devised and some approach to 

 midland methods was made.^ But this happened so early and 

 the traces of outfield cultivation in England are so slight that 

 the contrast between Celtic and midland tillage, sharp as it was 

 in reaHty, is not very helpful in estimating Celtic influences in 

 England. 



On this account it is desirable to determine for the Celtic system 

 the attribute which we have so often found pertinent in midland 

 England — the grouping of the parcels of a tenant's holding 

 within the arable area. In the fragmentary Welsh fields of the 

 sixteenth century such grouping would tell us little. Where only 

 a few tenants had each a parcel or at best a few parcels in the 

 common arable field, the location of the parcel or parcels imports 

 little, since the tenant's reliance was upon his closes. If the 

 grouping of parcels is to be important, the parcels must constitute 

 the major part of the holding. 



So they did in Scotland and without doubt in Ireland. In the 

 latter country, whenever townlands were subdivided each tenant 

 desired a share of each quahty of land. The location of the 

 parcels of a holding was thus dependent upon the number and 

 location of the different quahties of land to be divided. There 

 can scarcely have been thought of dividing the townland into two 

 or three equal compact fields. Indeed, it would have been impos- 

 sible to do this unless nature had given to the township only two 

 or three quahties of land in compact areas, and there would have 

 been no occasion for doing it unless a fixed two- or three-course 

 rotation of crops was to be estabhshed. The map of the Donegal 

 townland, which has been reproduced above, shows no such divi- 

 sion. The strips there assigned to three tenants were not scattered 

 throughout the arable; in fact, in about one-half of it no one 

 of the three tenants had any strip whatever. The field system 

 evolved by Irish co-heirs and co-tenants in the eighteenth and 

 nineteenth centuries was clearly not that of the EngUsh midlands. 



In Scotland the succession of crops itself prevented the sub- 

 division of the outfield into three equal parts. Only about 



* But cf. below, p. 232. ^ Cf. below, pp. 221, 225-226, 271. 



