How Do You Get a Block Out of a Cliff? Splitting Rock Along Lines You Choose

Rock contains natural planes of weakness, including bedding between layers and joints running through it, and a quarryman works with those rather than against them wherever possible, since splitting along an existing plane takes a fraction of the effort of cutting across one. Assessing where those planes lie and how the rock will behave is the first skill, and it determines where a quarry opens and what sizes of block it can supply. Cutting across the grain is done only where necessary. That dependence explains why particular quarries produced particular products, with some yielding large blocks and others only rubble, and why a building's stone can frequently be traced to a source by matching its characteristics.

Several techniques were used and each leaves its own evidence:

• •Wedges driven into a line of cut slots, splitting the rock along the line between them
• •Wooden wedges soaked so they swell, used where iron was scarce
• •Feather and tare, where two shims and a wedge in a drilled hole exert force sideways
• •Channelling, cutting a groove around a block with a pick, which is slow and leaves distinctive tool marks
• •Fire setting, heating the rock and quenching it to crack the surface
• •Explosives from the seventeenth century, which are fast and shatter what they are meant to free

Extraction leaves traces that survive on the quarry face and on the finished stone, and they are dateable. Rows of rectangular slots cut for wedges are characteristic of ancient and medieval work and their size and spacing vary by period and region. Rows of round drilled holes indicate the later technique using shims and a wedge, which arrived at a documented date in several countries and therefore gives a firm earliest date for the work. Continuous grooves indicate channelling. Saw marks indicate mechanised cutting. On dressed stone, the tooling of the surface is equally diagnostic, with different chisels and different finishes belonging to identifiable periods, which is how a mason dates masonry without any documents.

How a stone performs in a building depends on properties a quarry assesses before extraction. Porosity determines how much water it absorbs and therefore how it behaves in frost, since water expanding on freezing inside pores splits the surface away, which is why some stones flake and others do not. Bedding orientation matters enormously, since a block laid with its bedding planes vertical rather than horizontal delaminates within decades while the same stone laid correctly lasts centuries, and face-bedded stone is a common and expensive mistake. Solubility matters where acid rain falls on limestone. Salt crystallising within pores near the sea causes severe decay. And stone freshly quarried contains moisture that must dry before working, which is why it is cut easily when fresh and hardens afterwards.

Extraction is the easier half of the problem and transport determined where stone buildings could be built. A cubic metre of stone weighs well over two tonnes, which makes land transport enormously expensive with pre-industrial technology and water transport comparatively cheap, so stone travelled by river and sea over long distances while moving it ten miles overland could cost more than the quarrying. That is why buildings near navigable water use stone from far away and buildings inland use whatever is local, and why the geology of a region is legible in its vernacular architecture. Roads, sledges, rollers, cranes and inclined planes are all documented. Reusing stone from earlier structures was standard practice and is visible in a great many buildings.

Splitting along existing bedding and joints takes a fraction of the effort of cutting across them, which determines what a quarry can supply. Wedge slots, drilled holes for shims and continuous grooves each belong to identifiable periods, so extraction marks date the work. Stone weighs over two tonnes per cubic metre, so it travelled by water and rarely far overland.