Apparatus for Treating Air

The patent is titled for treating AIR, not for cooling it. Humidity is the controlled variable; temperature is the lever.

Measure a long paper strip, leave it over a steaming kettle for an hour, and measure again. It grows. That growth is the printing fault Carrier was hired to fix.

Mount two thermometers side by side and wrap one bulb in wet muslin. The difference between them is your humidity measurement.

Evaporation removes latent heat. Dry air evaporates faster and cools the wet bulb more, so a big gap means dry air and no gap means saturated air.

Add ice to a shiny metal can while stirring and watch for the instant mist appears on the outside. Record that temperature — it is the dew point of the room.

The dew point tells you how much water the air actually holds, regardless of its current temperature. Carrier's whole method is choosing a dew point.

Make a simple rectangular duct and mount a fan to draw air through steadily. Everything else sits inside this airflow.

Run copper tube through the duct and pass ice water through it. This is the cooling coil, and its surface must be BELOW the dew point to do anything.

Condensate must be collected and drained. The water in that tray is the humidity you removed — it is the product, not a nuisance.

Take wet and dry bulb readings at both ends of the duct. Downstream air should be colder and, crucially, holding less water.

Warm the chilled air back up with a small heater. Its water content is unchanged, so RELATIVE humidity falls — you now have warm dry air on demand.

The coil sets the moisture; the reheater sets the temperature. Independent control of both is exactly what a printing works needs and what a single cooler cannot give.

Leave the measured paper strip in the treated stream. Its length should now hold steady — the actual result Carrier was paid for.

Mark before, after-coil and after-reheat. The path drops to saturation, runs along it, then moves right — Carrier drew this chart himself in 1911.

The patent. US 808,897, "Apparatus for Treating Air", granted 2 January 1906 to Willis H. Carrier. He designed the first system in 1902 for the Sackett-Wilhelms Lithographing and Publishing Company in Brooklyn, whose multi-colour printing was misregistering because paper changed dimension with humidity between passes. The brief was dimensional stability of paper. Human comfort was not mentioned.

The dew-point control principle. Air holds a maximum amount of water vapour that depends strongly on temperature. Cool air below its dew point and the excess condenses on the cold surface, leaving the remaining air saturated at that lower temperature — which means it now carries a known, fixed amount of water. Reheat it and the absolute moisture content stays put while relative humidity falls. By choosing the coil temperature you choose the moisture; by choosing the reheat you choose the temperature. Carrier called it "dew-point control", and it is why every air conditioner produces a stream of condensate.

The chart is arguably the bigger contribution. In 1911 Carrier published the psychrometric chart, which plots dry-bulb temperature, humidity ratio, wet-bulb temperature, relative humidity and enthalpy on a single diagram. It turned air conditioning from craft into engineering: any process can be drawn as a path on the chart and its energy cost read off. Engineers still use it, and the version on the wall of a modern plant room is recognisably his.

How the name drifted. Early installations went to industry — printing, textiles, tobacco, pharmaceuticals, munitions — wherever material properties depend on moisture. Comfort cooling arrived through cinemas in the 1920s, which used it as a summer attraction, and public expectation reversed the emphasis entirely. The term "air conditioning" was in fact coined by Stuart Cramer in 1906, for a textile-mill process of adding moisture to yarn, the opposite of what most people now assume it means. The device you know as a cooler is a humidity controller whose side effect became its purpose.