380 Corresjjondence — Mr. F. J. Bennett. 



all to start from one straight line drawn about the middle of the 

 slope, and to end, diminishing gradually in width, on the steep side 

 of the slope. Their ground-plan also is alioays the same, and that 

 is the shape of the blade of a scythe, and they end against one 

 gradual curve, viz., the steep side of the slope, and are as nearly 

 one size as the shape of the slope will allow. 



When they occur on the slopes of escarpments and run along 

 combes, these terraces will be seen to be parallel with the sides of 

 the combe ; and yet when they occur on the spur or slope separating 

 two combes they will be found to run almost at right angles to those 

 running up the combe, and right across the slopes, always com- 

 mencing at one straight line and having the curved termination 

 before alluded to. This, I think, shows conclusively that natural 

 causes could not result in two sets of opposing terraces occurring 

 under precisely similar geologic conditions. These artificial terraces, 

 too, have in many cases quite altered the natural shape of the valley 

 or combe, turning a A -shaped valley into a fl one, as at Heddington, 

 near Devizes. Again, the walls of these terraces may often be 

 found, when cut into, to be faced with stones ; flints in some 

 areas, and sarsens in others, and these walls have, from this reason, 

 been resorted to for the sake of the facing material. 



This facing seems to me to have been a gradual process, and is 

 another proof of the artificial origin of the terraces, which go as far 

 back as the Common Field System, as Mr. Seebohm has so well 

 shown in his book on the " English Village Community." The 

 holdings in these fields were divided into strips all of the same 

 length, forty rods long and four rods wide when they were acre 

 strips, and two rods wide when they were half-acre strips. Now 

 a boundary, of course, was necessary to separate strip from strip, and 

 this boundary on the flat land consisted of a strip of turf, or 

 a " balk," or a " mere," as it was termed. 



Now in the case of the hill slopes it would not be so easy to keep 

 a well-defined boundary, owing to the wash of the soil when culti- 

 vation began. No doubt to begin with, the strips were marked by 

 cutting into the side of the hill, and we see that thej' all started 

 from one straight line. To keep the boundary, stones seem to have 

 been used, and fresh marked each year no doubt, as the line became 

 soiled up. In process of time these lines upon lines of stones would 

 result in a bank faced with stones, as we now find, and in flat 

 ledges or terraces on the hillside. 



These terraces, moreover, are often found associated with pit- 

 dwellings, earthworks, and old trackways ; and I have generally 

 found near them many pot-boilers or flints, which when heated 

 were used for cooking purposes. Hence, a very much stronger 

 case can be made out for the artificial than for the natural origin 

 of those terraces, as shown by Scrope, Seebohm, and others. 



F. J. Bennktt. 



Salisbury, July 8, 1897. 



