THE CELTIC SYSTEM 171 



There may be some truth in this conjecture as to the conse- 

 quences of the long persistence of the annual redistribution of 

 strips. Robertson's account of the first steps taken in getting 

 rid of runrig shows that such fluidity made easier the beginnings 

 of a more convenient arrangement.^ Yet in many places the 

 custom of annual re-allotment cannot have persisted so long as 

 cooperative ploughing and the old intermixture. The other re- 

 porters do not speak of it, and Robertson elsewhere is careful to 

 hmit his statement by saying that " these ridges were in some 

 cases frequently exchanged." ^ What generally gave the first 

 impetus toward consohdation was not the practice of annually 

 re-allotting strips, but the falhng-in of the leases and the action 

 of the landlord. Disregarding, however, the effect of annual re- 

 distribution upon the beginnings of consolidation, we can scarcely 

 look upon the usage as the most distinctive feature of Scottish 

 runrig. Had the practice been in vogue under English two- and 

 three-field husbandry as we have come to know it, the latter 

 would still have been very different from the agriculture of 

 Scotland. More characteristic of the latter were the size of the 

 farm or township, its occupation by co-tenants or co-heirs, the 

 manner in which it was tilled, and the distribution of the tenants' 

 acres throughout the arable fields. 



Before considering these features, however, as manifestations 

 of a Celtic type of field system, we shall do well to examine such 

 information touching them as comes from Wales and Ireland. 

 Some of it is earlier and some of it more specific than the Scottish 

 evidence. 



When reports from Wales were made to the Board of Agri- 

 culture in 1794, no open-field arable lying in common was to be 

 found in certain counties.^ Much waste land in the principality* 



^ Cf. above, p. 169. ^ Inverness, p. 334. 



' Brecknock, p. ^j: " There are no common fields in this district." Carmarthen, 

 p. 21: "I do not know of any considerable extent of open common field land in 

 the county." Denbigh, p. 1 1 : " There are no common arable lands in this county." 



* Brecknock, p. 39: " One half of the district, containing on the whole 512,000 

 acres, is waste lands." Cardigan, p. 29: "The greater part of the low lands is 

 pretty well inclosed; but hilly and exposed situations are mostly open." Carmar- 

 then, p. 20: "About two-thirds of the county is inclosed." Glamorgan, p. 42: 

 " The waste land in this county is considerable; computed to amount to upwards 



