some classic recipes appear paradoxical. to make a salmis de canard one removes the breasts and thighs and roasts them. the cook then makes a sauce by cooking the scraps and bones of the duck in water with vegetables and aromatic herbs. isn't the second step redundant? no, because in the final dish the odorant and taste molecules contributed by the duck are retained for different lengths of time by the meat and the sauce before being released. the flavor lingers in the mouth longer. how can odorant molecules be trapped? in the sauce that accompanies the roast duck the long cooking of the scraps and bones in water with vegetables has the effect of extracting gelatin present in the skin, tendons, and bones, producing a decoction that extracts the odorant and taste molecules present in two distinct physicochemical environments: in the meat where they are dissolved for the most part in fats, which are themselves dispersed between the muscle cells; and in the sauce, where they are in a liquid solution. in the mouth these molecules are released in different ways, so that the flavor of the duck lasts longer. how can this release be controlled? in their pure state these molecules are highly volatile and the pressure of saturated vapor increases with temperature. the cook is able to vary the degree of volatility by redistributing the odorant molecules into environments of different temperatures. the molecules can also be placed in solution. in this case their volatility depends on the solvent used (whether water, alcohol, oil) because the molecules bind to a greater or lesser degree with the molecules of the solvent (thus saturated and unsaturated oils differentially retain odoarant molecules in solution). compartmentalization is another, more radical means of retaining molecules. fines herbes (a mixture of fresh chopped herbs, e.g. parsley, chervil, tarragon, chives) release their odorant molecules only when their cells are ruptured by chewing. emulsions, foams, gels, and pasta are systems of the same type. similarly, cooks may soon be able to use liposomes - artificial cells created by the assembly of molecules analogous to those of cell membranes. this question is being studied as part of a european union project devoted to the innovative transfer of technology for culinary purposes h this / molecular gastronomy. the best-known explanation of a cooking method is probably this catchy phrase: "sear the meat to seal in the juices". the chemist von liebig came up with this idea around 1850. it was disproved a few decades later. yet this myth lives on, even among professional cooks. libeig thought that the water-soluble components of meat were nutritionally important, so it was worth minimizing their loss. he explained: when meat is introduced into the boiling water, the albumen immediately coagulates from the surface inwards, and in this state forms a crust or shell which no longer permits the external water to penetrate into the interior of the mass of flesh, which retains its juiciness. and if the crust can keep water out during boiling, it can keep the juices in during roasting, so it's best to sear immediately. but simple experiments in the 1930's showed that the crust that forms around the surface of meat is not waterproof: the continuing sizzle of the meat in the pan or oven or on the grill is the sound of moisture continually escaping and vaporizing. in fact moisture loss is proportional to meat temperature, so the high heat of searing actually dries out the meat surface more than moderate heat does. but searing does flavor the meat surface with products of the browning [maillard] reactions, and flavor gets our juices flowing h mcgee / on food and cooking. plants store glucose as starch in two forms: long straight chains (amylose) or small branched shapes (amylopectin). how much amylose a starch contains has a major effect on how it behaves in cooking. when the granules pop and starch rushes out into the sauce or filling, these long amylose molecules are more effective thickeners than small amylopectin. an even greater change takes place as the starch cools. amylose bonds together to form a firm solid gel, while amylopectin does not. high-amylose starches set up into an opaque gel that is firm enough to cut with a knife. starches that are mostly amylopecin form a clear, thick, glossy coating. if you are making a coconut cream pie you definitely need it firm enough to cut, but it doesnt matter if it is opaque. but it would be a shame to have a cloudy covering on bright red cherries in a cherry pie. there are a number of other differences. high-amylose sauces fare badly when frozen and thawed. they become a dry, spongy mass surrounded by liquid. high-amylopectin starches freeze and thaw beautifully. these granules swell and pop at lower temperatures than those of amylose, so their sauces thicken sooner. how can you tell which starches contain more amylose? nature has simplified this for us. the ordinary grain starches like wheat and corn are high in amylose, ~26%, while the root starches like arrowroot and tapioca contain 20% or less. potatoes are a tuber, not a true root, and potato starch falls somewhere in between, ~23%. some types of cereal starches called waxy, e.g. waxy cornstarch, unlike regular cornstarch, are almost pure amylopectin. these and modified starches are the mainstay of the frozen food industy s corriher / cookwise. gluten is formed only when wheat flour is deformed in the presence of water, which partially dissolves - hydrates - proteins present in the flour. hydration is the first step in gluten formation. proteins also need to be partially denatured by mechanical manipulation. then they reform into new structures. there are two components in the gluten complex: gliadin is a sticky, semi-fluid protein that is soluble in alcohol; glutenin is not soluble in alcohol and has a fibrous elastic texture, formed when lots of links are created between many different gliadin and glutenin units. before these can be formed the components need to be well hydrated. about twice as much water by weight of proteins is needed to form gluten. by controlling the rate at which water is added and the extent to which the water can reach all starch granules, the degree of hydration can be controlled. rubbing in fats to coat individual grains of flour, as in pastry-making, will tend to repel water from the surface, decrease the rate of hydration, and help control the amount of gluten formation. once hydrated, different protein molecules can begin to interact with one another and new inter-protein blonds - di-sulphide bridges, hydrogen bonds, etc - can be formed. the details of these processes are not understood but it is well known that stretching of the individual proteins helps break up the initial internal bonds. the proteins between startch granules develop into thin sheets which break into fine fibrils that stick to one another and stretch as the dough is kneaded. if the dough is allowed to relax before the gluten is fully developed, then some of these stretched proteins will relax back without forming new intermolecular bonds that stabilize gluten. this is why pastry recipes call for rest periods between rolling and tell you to avoid overworking the dough p barham / the science of cooking. aquaculture is not new. the chinese invented it 3k years ago using waste products from cultivated silkworms to feed carp in small-scale freshwater-pond farms. as recently as 20 years ago in western countries, with the exception of salmon, acquacultured products were niche items. today dozens of mainstream fish are being domesticated to appear at supermarkets everywhere. yellowtail, halibut, red snapper and even vw-sized bluefin tuna are coming under some kind of human-controlled production. whereas animals like sheep and cattle were adapted to fit the farm over thousands of years, many of the ocean species could be tamed in as little as a decade. in 1937 j.l. lush, an iowa state agriculture prof. wrote "animal breeding plans", a book that for the first time applied the laws of quantitative genetics to animal populations, proving that an animal's body mass acquisition rate could be increased through selective breeding. in the course of just a few decades, the food-conversion ratios of cattle, chickens and pigs were all at least doubled, meaning that the same amount of grain could grow twice as much meat. for salmon, 30% of the variation in growth goes back to the genes. further, because fish produce exponentially [sic] more offspring and reach sexual maturity faster than land animals, the interval of generational improvement is shorter, leading to the rapid creation of faster-growing breed. these factors helped norway increase salmon production to 500,000 tons from just 100 tons 30 years ago. there are other differences. whereas terrestrial livestock expend significant amounts of energy warming their blood and fighting the force of gravity, fish, because they are coldblooded and float all their lives buoyantly, have much more energy to devote to acquiring edible body mass. an "unimproved" cod already has food conversion ratio that is better than the average "improved" cow. but to breed a fish that is more efficient, that will eat less wild fish, risks a different environmental hazard, for selectively bred fish can escape into the open ocean where their interaction with wild fish is not yet entirely understood. cod, unlike salmon, have been known to chew their way to freedom. and unlike land-based animal improvement, which started during the great depression, fish breeding was born in the bright light of environmental activism. over the last decade, the collapse of wild cod has emerged as a narrative for ocean conservationists. cod were once the most important food fish in the world. seemingly inexhaustible, they propelled the development of spain's oceangoing fleets, colonial america's economy, and fed the world's poor for centuries. humans showed their gratitude by wiping it off the face of most of earth p greenberg / "green to the gills" nyt 6.18.06. unlike sugar molecules, which reside in teh stalks of sugar cane or the beets that are used to make sugar, high-fructose corn syrup is artificial because it is not found anywhere in corn. produced in large manufacturing facilities scattered mostly across the flat, golden expanse of the american corn belt - outside of canada, the united states is the only country with a significant consumption of high-fructose corn syrup, largely because other countries have erected successful trade barriers to protect sugar - it is not a product that anyone could cook up at home using a few ears of corn. the process involves stainless steel vats, a dozen mechanical processes and chemical reactionns, including high-velocity spinning and the introduction of enzymes to catalyze molecular rearrangements. these enzymes turn most of the glucose into fructose, which is sweeter. this 90% fructose syrup mixture is then combined with regular corn syrup which is 100% glucose. the final product is a clear, goopy liquid as sweet as sugar. the recent backlash against the ingredient, which has seen 20 years of uninterrupted sales growth, has caused its corporate sponsors to take notice. the president of the corn refiners assn. says her arguments that high-fructose corn syrup is a safe ingredient have gained little traction. her trade group recently entertained the idea of changing the sweetener's name. "it really does have this negative connotation" nyt 7.2.06. in its quest to create ice cream as voluptuous as butter andd as virtuous as broccoli, the ice cream industry has probed the depths of the arctic ocean, studied the intimate structures of algae and foisted numerous failures on the american public. unilever applied to britain's food standards agency for permission to use a new igredient in its frozen desserts - a protein cloned from the blood of an eel-like arctic fish, the ocean pout. instead of extracting the protein from the fish, which is neither sustainable nor economical, the company developed a process for making it by altering the genetic structure of a strain of baker's yeast so that it produces the protein during fermentation. this ingredient, called an ice-structuring protein, has been approved by the fda and is used by unilever to make some productss in the usa, like some popsicles and a new line of ice cream bars. ice-structuring proteins protect the fish, which would otherwise die in freezing temperatures. they also prevent ice crystals from growing. no dna or other material from fish is used in the process. since the 1980s making ice cream lighter and creamier has been largely about adding ingredients. now we have complicated processes but the recipes can be simplified. products produced with the new technology are less affected by partial thawing than traditional ice creams, which become dry, sticky and hard in fluctuating temperatures. ice crystals are the enemy of ice cream, which is an emulsion of air, fat and water. emulsions are always fragile because the elements want to separate. every time ice cream thaws slightly, the emulsion is compromised and the ice crystals combine into larger, jagged pieces that destroy the texture. nyt 7.26.06. an especially memorable dinner at blue hill (greenwich village) began with a shot glass of pale tomato water so concentrated it was like some platonic ideal of nourishment. tomatoes came into play again later on: crimson and rose and yellow tomatoes, bouncy tomatoes and supple tomatoes, cut into wedges and mixed into a salad that charted a whole spectrum of tomato possibility, from modestly tart to immodestly sweet. the salad was dressed with a "tomato cloud" of tomato and cucumber water whipped with gelatin into something with the texture of an ethereal mousse. our server spooned the nimbus into the salad, where it began to dissipate and spread, becoming more of a tomato fog. nyt 8.2.06. the 6k year-old breadmaking method hasn't changed much since pasteur made the commercial production of standardized yeast possible in 1859. jim lahey's method may be the greatest thing since. mr lahey's method requires no kneading or special ingredients equipment or technique. time does almost oall of the work. lahey's dough uses very little yeast and the dough ferments very slowly. he mixes a very wet dough, about 42% water, which is at the extreme high end of the range that professional bakers use to create crisp crust and large, well-structured crumb, both of which are evident in this loaf. it is mixed in less than a minute, then sits for 18 hours. the slow rise does over hours what intensive kneading does in minutes: it brings the bluten into side-by-side alignment to maximise their opportunity to bind to each other and produce a strong, elastic network. the wetness of the dough is an important piece of this because the gluten are more mobile in a high proportion of water, and so can move into alignment easier and faster than if the dough were stiff. the resulting cmbination of great crumb, lightness, flavor, and a crackling crust that most separates amateurs from pros because it requires getting moisture onto the bread as the crust develops. to get that, pros use steam-injected ovens. lahey solves this by putting the dough in a preheated covered pot (e.g. cast iron). by starting this very weet dougin a hot, covered pot, he lets the crust develop in a moist, enclosed environment. the pot is in effect the oven which has plenty of steam in it. once uncovered, half-hour later, the rust has time to harden and brown, still in the pot. nyt 11.8.06.