Bedding and bearing

The whole game is contact area

Granite in compression is nearly indestructible. Granite under a point load is not. Set one dressed block on another and the load travels only where the surfaces actually meet. If those surfaces are convex, or grit-contaminated, or dressed to a crown, contact reduces to a few high spots — and the same weight that a full bed distributes harmlessly becomes a line or point load that can split a stone along its length.

This is why bed dressing matters more than face dressing in a loadbearing course. The face is visible; the bed is structural. Masons working in the South Indian granite tradition worked the bed with a flat chisel across its full extent, testing it against a straight-edge rather than by eye. The target was not polish — it was plane. A slight hollow, imperceptible to a visitor looking at the joint, is actually preferable to a crown, because a hollow ensures that load is taken at the perimeter of the bed rather than at a central high point that might rock or crack under settlement.

Dry-laid construction — no mortar to fill gaps and redistribute load — makes the requirement stricter than in mortared work. Mortar is forgiving of minor surface irregularities; it flows under pressure and evens out contact. Without it, what you dress is what you get. A course that looks tight from the outside can be bridging across a void at centre, concentrating load at two narrow edges. The stone above acts as a beam across that void. If the void is wide enough and the course thin enough, it flexes until it cracks.

The check is simple: a thin steel feeler gauge slid along the joint after laying. If it passes, the surface needs reworking. This is not a modern quality-control procedure imposed on ancient practice; it is a restatement of what the dressed-bed requirement was always enforcing, because the alternative shows up as cracked headers in courses that have been standing for centuries.

Joint thickness in well-laid dry granite courses at major Dravidian temples is often under two millimetres and sometimes difficult to locate by eye at all. This is not aesthetics — it is evidence that bed contact is high. The tighter the joint, the more of the bed area is working. The more of the bed area is working, the lower the bearing stress at any one point.

The elevation looks monumental. The load path is arithmetic: force divided by area. Maximise the area, and the stone does not care how heavy the tier above it is.