When a Stock Changes Register

I am standing over a tall pot of chicken stock that has been at a bare simmer for somewhere around four hours, and I have just noticed, without quite intending to, that the surface has changed register. Until a few minutes ago the surface had a faint cloudy commotion to it — small bubbles arriving at uneven intervals, a thin grey scum gathering at the rim, the colour shifting slightly between gold and a muddier yellow depending on where I looked. Now the surface is clearer. The bubbles are smaller and arrive in a slower, steadier rhythm. The rim is almost clean. The colour, when I look down past the surface into the body of the liquid, has a settled quality that the earlier hours did not have. I have not added anything. I have not turned the heat. The stock has, by itself, crossed into the second half of what it was doing, and the surface is the first place I see it.

What I think is happening, in the plainest terms I can manage, is that the various exchanges between the bones, the vegetables, and the water have moved past their loud phase. In the first hour or two the bones are giving up the broadest part of their soluble proteins and the vegetables are giving up most of their cell water, and a great deal of that material arrives at the surface as foam and scum because it is denatured, partial, and lighter than the liquid. The pot looks busy because it is busy. As the hours stretch, the easy exchanges run their course, and what is left to dissolve does so more slowly and more cleanly. Gelatin from the joints continues to release into the water in long slow strands, fat from the marrow continues to rise but in finer droplets that do not coalesce into the muddy slick of the earlier hours, and the small fragments of vegetable left in the pot have given up almost everything they were going to give. The pot looks calmer because it is calmer. I do not pretend to have the chemistry of any single transition pinned. What I trust, after years of long stocks, is that the busy surface and the quiet surface are two different stocks in some practical sense, and that the change between them is what I am here to catch.

The pivot, then, is the quietness itself. I have a habit, when I am making a stock, of glancing at the pot every fifteen or twenty minutes — not stirring, not skimming, just looking — and I am not looking for any particular thing until something quietens. The first sign is usually the rim, because the scum has stopped renewing itself; what is on the rim now is what was on the rim half an hour ago, slightly more dried out, and nothing new is arriving to replace it. The second sign is the centre, because the bubbles there are smaller and slower and more transparent than they were, and the steam coming off them has a thinner, sweeter smell — closer to clean broth than to the deeper, slightly metallic note of the early hours. The third sign, which I trust most, is that the colour in the body of the liquid stops shifting when I move my head; in the earlier hours the colour seemed to change with my angle of view because there was so much suspended material catching the light, and in the later phase the colour is the same from every angle because the suspended material has thinned. When all three signs arrive together I take a clean spoonful, taste it past the temperature with patience, and confirm what the surface has already told me: the stock is now what it will be, and further time on the heat will deepen it only by removing water, not by adding anything I want.

This matters because the most common failure of a long stock at home is the one I have made many times, which is to keep cooking past the change. The cook has decided, before lighting the burner, that the stock needs six hours, and the cook is going to give it six hours regardless of what the pot is doing. The trouble is that beyond the change the stock starts to do things the cook usually does not want — the gelatin begins to thicken the surface in a way that traps the small amount of fat that is still rising, the colour darkens past the warm gold into something nearer brown, and a faint bitterness creeps in from the vegetables that have now been past their useful phase for an hour. None of these are catastrophes. The stock will still be a stock. But the cook who watches the surface for the change, and turns the heat off within a half hour of seeing it, has captured the stock at the register the cook actually wanted. The cook who watches the timer instead has captured a slightly different stock — heavier, darker, less clean — and will not always know why. The puzzling thing, the thing that took me a long time to admit, is that the timed stock is not necessarily a stronger one. More hours does not equal more flavour past a certain point; it equals more concentration of whatever was already in the pot at the change, including the bitter notes the vegetables released after their useful phase. The honest cook, after a few overcooked stocks, comes to see that the additional hours past the change were buying very little and were costing something.

I should name the caveats. The change does not arrive at the same hour every time. A pot of mostly knuckles will reach the quiet surface later than a pot of mostly wings, because the joints take longer to give up their slow gelatin and the early loud phase runs further into the cook. A pot with more vegetables will scum more in the early hours and may reach the quiet phase slightly earlier or slightly later depending on how the vegetables behave; carrots run sweeter and quieter, onions run noisier and longer, and a pot with both arrives somewhere in between. A pot with a lid on partially will hold more steam and may stretch the loud phase by trapping volatile compounds that would otherwise have left; I prefer the lid set just barely ajar, with a thumb's width of gap, so that the steam moves and the change can find me. None of these breaks the principle, which is that the surface is the first reliable instrument and the timer is, at best, the second.

For someone trying to make a long stock for the first time at home, the only thing I would suggest is to skim once in the first hour and then leave the pot alone — no more skimming, no more stirring, no more lid adjustments — and look at the surface once every twenty minutes. Do not write down the times. Do not measure the bubbles. Just look. Sometime between the third and the sixth hour the surface will change register, and the first time you see it you will not be certain it has happened; the second or third time you see it on later stocks you will know. After perhaps half a dozen stocks the eye picks the change out without effort, and the cook reaches for the burner knob without having to think about whether it is time. It is time when the pot says so.

I turn the heat off now. The surface goes through one small final exhalation as the last of the convection slows, and then settles. I let it sit for a minute or two before I begin to strain it, so that the suspended particles finish drifting to the bottom of the pot, and then I lift the cheesecloth-lined sieve over the second pot. The stock comes through clean — a clean even gold, with the lightest of films across the top — and I know, before I taste it, that I caught it at the right moment because the surface had already told me. The kitchen is quiet now. The pot is quiet too. The work is finished, and the finishing was something the stock did to itself when I was paying attention.