The Ether Inhaler

Set the chemistry down before touching glass, because it explains every later choice. Diethyl ether at 20 degrees Celsius has a vapour pressure of about 442 mmHg against an atmosphere of 760 mmHg. Divide: the gas immediately above liquid ether is roughly 58 percent ether by volume. A patient breathing that dies. Surgical anaesthesia sits near 2 percent. So the globe has three jobs and only three. First, present a large liquid surface — Morton used a sponge precisely to multiply surface area, so the vapour space stays saturated even while gas is being drawn off. Second, split the inspired flow so that only a fraction passes over that surface and the rest bypasses it. Third, make the patient's exhaled breath leave without going back through the ether, or the circuit accumulates. Work out the splitting ratio you need. If saturated vapour is 58 percent and you want 2 percent delivered, the fraction of total flow that may pass through the vaporising chamber is 2/58, about 3.4 percent. Thirty parts of bypass to one part of carrier. Hold that number — it is why every draw-over vaporiser ever built has a large, deliberately leaky bypass channel, and why the ones that killed people were the ones where the bypass blocked. Record this in your build notes as the design target before you cut any glass.

Work the borosilicate on the bench lathe. The historic form is a globe of roughly 200 to 300 millilitres with two ports on the same hemisphere, and there is a reason for the geometry: a sphere gives the largest vapour volume for the smallest surface losing heat, and vaporisation chills the liquid fast. That cooling is not a detail. Pulling ether vapour off the surface removes its enthalpy of vaporisation from the remaining liquid, the liquid temperature drops, the vapour pressure drops with it, and the delivered concentration falls during use. Nineteenth-century practice was to stand the globe in a bowl of warm water for exactly this reason. Your build should reproduce the effect rather than hide it, so keep the glass thin and do not lag it. Set the inlet port so its tube ends well above the liquid line, and the outlet port so it draws from the top of the vapour space. Anneal properly after working — an unannealed globe under thermal cycling will craze and, with a flammable charge inside, that is not a recoverable failure. Fit the sponge through the wider port before you constrict it. Cut the bypass as a separate T in the silicone line downstream, not into the glass. Making it external means it is inspectable, and after the last notebook you know that inspectability is the safety case.

Charge the chamber with ethanol, not ether. Ethanol has a vapour pressure near 44 mmHg at 20 degrees — about a tenth of ether's — so the circuit runs at a concentration you can measure safely and the arithmetic from step 2 still applies with one substitution. Draw a known flow through the circuit with a hand bellows or a low-flow pump and measure delivered concentration. The cheap and honest method is gravimetric: weigh the whole vaporiser before and after a timed run, and divide the mass lost by the volume of air drawn through. That gives you grams per litre, which converts to percent by volume through the molar mass. Repeat at three chamber temperatures by standing the globe in water baths at 15, 20 and 25 degrees, and plot your measurements on top of the curve the notebook predicted. Expect your delivered concentration to fall during each run. That is the vaporisation cooling from step 3 showing up as data, and if you do not see it your run was too short. Where this stops: you have built and calibrated a draw-over vapour circuit. That is a genuine and complete piece of instrument physics. It is not an anaesthetic machine. Modern vaporisers are temperature-compensated, flow-compensated, agent-specific and keyed so the wrong agent cannot be filled, and they are verified against a gas analyser before every use. The gap between this bench circuit and that is the whole of twentieth-century anaesthetic engineering, and it is a gap you do not close with a sponge.