Fire Hydrant

Holly's leather ring is both a valve seat and a packing. One part, two sealing functions — note that before building, because it explains the shape.

Fill a rigid plastic tube, cap both ends, and freeze it overnight. It bulges or splits. This is what a wet hydrant does in winter, out of sight, underground.

Make a vertical tube about 300 mm tall standing on a base chamber. This is the barrel that will stand above ground and must end up empty.

Run a narrower pipe up the centre, carrying water from the base chamber to an upper outlet. Holly's arrangement is a pipe within a casing, not a single bore.

Cut a leather ring to sit between the main valve and its box. Leather compresses to fit imperfect cast surfaces — the reason a soft seat beats a machined one here.

Fit a valve at the bottom of the assembly, closing upward onto the leather seat. Everything above it will be empty when it is shut.

Run a long rod from the main valve to a squared head at the top. The operator stands at street level and works a valve a metre below their feet.

Fill, open, close. Water stays in the barrel. Tip it and measure what comes out — that volume is what would freeze.

Make a small opening near the base of the barrel, below where the frost line would be, discharging to the outside.

Arrange the valve stem so that as the main valve closes, it uncovers the drain port; as it opens, it covers it. One motion, two valves, opposite senses.

Open, flow, close. The barrel drains by itself within seconds. Measure how much water is left and compare with step 8.

With the hydrant open and flowing, check that nothing escapes from the drain port. A drain that leaks while in use undermines the whole main.

Put the protective cap over the squared head. Without the matching key it cannot be turned — Holly claims this deliberately, because an open hydrant drops pressure across the district.

Cycle it closed, leave it drained and freeze it. Nothing splits. Repeat step 2's comparison and write both results down.

The patent. US 94,749, "Improvement in hydrants", granted 14 September 1869 to Birdsill Holly of Lockport, New York, who held well over a hundred patents and is also remembered for district steam heating.

Holly's larger contribution was the water system, not the hydrant. Before him, urban firefighting pressure came from gravity — a reservoir on high ground, and whatever head it happened to give. His Holly Fire Protection and Water System, built at Lockport in 1863, used steam and water-turbine pumps to push water into the mains at a controlled pressure directly. That changes what a hydrant is for: with direct pumping, a hydrant is a tap on a permanently pressurised network, and it becomes worth engineering properly. Many American and European towns adopted the approach within a decade.

Why the drain is the whole design. The visible parts of a hydrant — the cap, the outlets, the paint — are trivial. The two decisions that matter are both invisible: put the main valve below the frost line, and make closing it open a drain. Together they mean the exposed barrel contains no water between uses. This is the dry-barrel hydrant, and it is still the standard wherever the ground freezes; wet-barrel hydrants, which hold water right up to the outlets, are used only in warm climates where they are simpler and faster to charge. Steps 2 and 14 are the entire argument, and they take a night each.

The leather ring is a nice piece of nineteenth-century pragmatism. Cast iron of the period could not be machined to a reliable metal-to-metal seal at reasonable cost, so Holly used a compliant material that conforms to whatever the casting actually is — and then got a second seal out of the same part by using it as packing at the pipe joint. One component, two leak paths closed, no precision required.

A footnote about lost records. It is often said that the patent for the original fire hydrant was itself destroyed by fire — the 1836 Patent Office fire, which burned roughly ten thousand patents including, by the usual telling, Frederick Graff's earlier hydrant. The story is repeated widely and is difficult to verify precisely, which is worth saying rather than passing it on flat. What is certain is that the 1836 fire destroyed the early record wholesale; the Otis steam shovel elsewhere in this collection has no surviving drawing for the same reason.