STONE TILES. 485 
<dry before the frost comes, so that the " sap " is dried up, no frost will 
move the layers, and the rock is then "bound." Formerly it was the 
custom to begin digging the stone about Michaelmas time or before ; now 
the men commence about the end of October, and work till the end of 
January. When dug and raised, the stone is banked up and covered with 
earth, to keep the moisture in, until a frost comes. In dry seasons, the 
quarrymen even water the blocks to keep them moist. It requires three 
or four good frosts to act upon the stone ; hence a mild winter is disastrous 
to these toilers. Sometimes the stone is put out and then covered up, 
again and again, until a sufficiently long frost is experienced. As a rule 
a week's frost is needed. 
The stone can be split up any time after it has been "frosted," and of 
course it can be dressed any time after it has been split. Splitting is 
usually commenced about the middle of February. 
The only qualities in the slates are the first and second, the former 
being thinner and the latter thicker, though both may be composed of 
the same kind of stone. The more oolitic slates as a rule are a trifle 
thicker than the eandy slates, but no particular difference in quality and 
durability is recognized. 
It is thought that the slates do not last well in cities. Mr. Barrett in- 
formed me that Exeter College Chapel at Oxford was roofed with Stones- 
field slates, but after 20 years had elapsed the stone had become so 
41 bent, broken, and perished," that it had to be removed. 
Deal pegs are used for fixing the slates. Oak pegs were formerly used, 
but it is found that deal forms a better head when driven in than oak, 
and not so many slates are broken. Copper and iron nails have also been 
used. In the district around Stonesfield, not only the houses but the 
porches are roofed with the stone-tiles. 
At Througham Field, where Stonesfield Slate is worked, the blocks are 
improved by lying out all the winter the second or third frost breaks 
them up. Slabs 6 feet square or even more are obtained. They furnish 
material for cow-sheds, mangers, bordering for gardens, and pitching for 
stables. There is not more than 18 inches of good tile-stone at this 
locality, but it occurs at slightly different horizons. I obtained some 
examples of the tools used at Througham Field, and these are placed in 
the Museum of Practical Geology. 
Forest Marble. 
Over a large area of Wiltshire, Gloucestershire, and Oxfordshire, where 
the Forest Marble is exposed, the beds have been worked for " slates." 
Some of the pits are known as " Slate and plank quarries " ; and the 
material is obtained at or near the surface. 
These stone- tiles do not require to be split : hence no exposure to frost 
i's necessary. They occur as thin flags or leaves of limestone (or occasion- 
ally calcareous sandstone), often ranged in oblique bands separated by 
horizontal layers of clay and shale. All that is necessary is to dress them 
to the required shape. They resist frost and are considered to be moro 
durable than the Stonesfield Slate, but they are heavier and coarser.* 
They have been obtained at Atford, in Wiltshire ; and many of the 
old buildings at Bradford- on- A von are roofed with them. J. Woodward 
mentions " Plates, or thin Strata of Stone, used as Slate for covering the 
Houses at and about Bath. These were taken out of the Slate-Quarries, 
in Charlwood, in the Parish of Box, about five miles from Bath, and about 
.a Mile from London Road." This slate, as in other cases, often has ". a thin 
Crust -of reddish Spar on its Surface. "f (See p. 355.) 
Other localities for these Forest Marble Sla,tes, are Beverstone and 
Charlton near Tetbury, Avening, Chavenage, Cirencester, Aldsworth,J 
Burford, between Holwell and Shilton, Poulton, and Fairford. (See p. 369.) 
In addition to slates, large slabs 5 X 3 or 4 feet are obtained from the 
Forest Marble for piggeries, garden-edgings, planks, pitching, &c. 
* See Lycett, Quart. Journ. Geol. Soc., vol. iv. p. 185. 
t J. Woodward, Nat. Hist. Foss. Eng., Tome II., 1729, p. 101 
j See Hull, Geol. Cheltenham, p. 70. 
