from the following calculation as they contribute comparatively little to the results.) The number of fields in thirty randomly selected squares averaged thirty-seven. (In several squares that fell near villages there were over seventy fields, giving an average area of field of only 1240 square yards, i.e. typically thirty-one yards by forty.) Hence there are about 14,000 small fields, of average area 2300 square yards or just under half an acre. A random sample of twenty individual fields measured on the map suggests that their average perimeter is 233 yards. ( NB : Sampling by such methods as dropping a pencil onto the map will produce a misleadingly large proportion of the comparatively big fields, and it is not easy to avoid this effect.) Since nearly all walls separate two fields, the total length of wall is 928 miles. Adding a bit to allow for irregularities that do not show up on the six-inch scale, and for some nests of small fields in the generally open areas omitted, and rounding off, I think it is not wildly inaccurate to say that there are a thousand miles of wall in Árainn. Inis Meáin and Inis Oírr would add about 300 and 200 miles respectively, giving a total of 1500 miles. So the islanders exaggerate when they say that the walls of Aran, put end to end, would stretch all the way to Boston—but that is understandable, since it is to Boston that so many of them have fled, rather than be sucked dry as dead flies by the economics of this stone web.
As to the general pattern of this field-system, it is clear that something so regular over such large areas, even if riddled with inconsistency and wilfulness, is structured by overpowering natural phenomena. In fact two elemental strengths have laid thefoundations for this labour of generations. First, the cracking of the limestone by earth-forces acting on a continental scale. There are two principal sets of vertical fissure-planes or joints, the more fully developed one being oriented roughly north-south. (The bearing is about fifteen degrees east of north, but I will say simply “north” rather than “approximately north-north-east.” In fact, since the bearing is much more fundamental to Aran geography than true north or magnetic north, I could call it Aran north.) The other set is at right-angles to the first, and on certain levels of the islands a third, oblique set appears, confusing the pattern.
The second elemental force is erosion, which has been guided by the joint system. The small-scale effect of rainwater, widening the cracks into little canyons and dividing the surface into rectangular flags or long strips, has already been mentioned. Vastly greater was that of the huge thicknesses of ice pushing across this region of Ireland for a period of perhaps ten thousand years, ending about fifteen thousand years ago. The glaciers ripped out strips of blocks to create valleys of all sizes, from the sea-ways between the islands and the wide depressions that almost divide the big island in three, to narrow ravines and little glens, all oriented to Aran north. It is difficult even today to override this impressed topography, and in the past the currents of life had to flow with it. An oblique walk across an area of open crag is a continuous struggle with little cliffs and ridges and gullies, with no two successive steps on the same level, whereas if one follows the direction of the jointing, smooth flagged paths seem to unroll like carpets before one. Inevitably, walls accept these natural ways, and many of them spring from the natural pedestals provided by long unbroken ridges of rock. East-west walls often have to clamber down into narrow glens and then clamber out again, and tend to do this by the shortest way, confirmed in this labour-saving intention by the bearing of the minor set of fissures. Hence the fields tend to be rectangular and elongated north-south, especially on the plateau where the schema has room to spread itself.It is as if the walls