A Cold Boiler in June

My dear Clapeyron,

The engine at the Chaillot works stood idle the whole of June, and I found I could learn nothing from it. The boiler was cold, the cylinder at ambient temperature, and every measurement I had planned required the gradient that was not there. I wrote in my notebook: no fire, no work. Then in the second week of July the proprietor relit the furnace, and within three days the same instrument yielded more than the previous two months.

What I noted afterward, and note again to you now, is that the instrument did not change. The furnace did not change. What changed was whether the two reservoirs existed at all. An engine without a gradient is not a poor engine; it is not an engine. And a field of inquiry without its two temperatures - a question hot enough to drive work, and a settled body of fact cold enough to receive it - produces nothing, however many men attend to it.

This bears on the matter we discussed in Paris. The memoir of 1824 drew little notice for six years. I do not attribute this to its defects, though it has them. The hot reservoir was missing: no pressing question of theory required the cycle at that hour. The same pages, read after your own reworking, meet a different condenser. The work extracted is not a function of the text alone. It is a function of the difference between what the reader brings burning and what the literature holds settled.

Hence a practical rule, which I state before running the process rather than after it overheats: when a field goes quiet, do not measure the field. Measure the gradient. If the question has cooled to the temperature of its surroundings, no efficiency is recoverable from further attention, and the honest act is to bank the fire. If the question still burns while the received answers sit cold, attend - the yield per hour is then at its maximum, and it will not remain so. Attention spent on a dead gradient is waste heat, and waste heat must go somewhere.

I will relight the Chaillot furnace myself next Thursday and log the cylinder temperatures hourly until the gradient stabilises.