On the morning of 8 December 1941, the day after the Japanese attack on Pearl Harbor, every bottle of human albumin solution in Boston was put on a military aircraft and flown to Hawaii. Boston had every bottle in existence. There were not very many.

At the naval hospital at Pearl Harbor a Lieutenant Colonel of the United States Army Medical Corps named Isidor Ravdin took delivery of the bottles and went looking for the seven most severely burned men in the wards. They were so swollen with shock fluid that the veins in their arms could not be found. Ravdin punctured the femoral vein in the groin of each man and ran in the clear yellow albumin, in some cases over hours, in some cases over days. All seven survived.

The albumin had come from a small pilot plant in the basement of a Harvard Medical School laboratory in Boston, where a forty-eight-year-old physical chemist named Edwin Cohn had spent the previous eighteen months trying to break human blood plasma into its component proteins.

This is the story of how, because of one man's stubborn application of ethanol and refrigeration to a centrifuge full of donor blood, a generation of soldiers came home with skulls that had been opened and closed again, and how, last week, an engineer in Montreal made the same kind of object work seventeen times faster.

Edwin Cohn had begun his work for the National Research Council's Committee on Transfusions in the autumn of 1940, with a grant of six and a half thousand dollars. The Committee believed, correctly, that war was coming. The transfusion problem of 1939 was simple enough: whole blood is fragile, has a short shelf life, must be matched, and cannot be carried in a haversack. Plasma was already in use as a substitute, but plasma is a complicated mix of proteins, and most of those proteins are not what is needed at any given moment in a wounded man.

What Cohn invented, between 1940 and 1945, was a way to separate the proteins. Plasma, cooled to just above freezing, was added in stages to a series of ethanol baths at carefully controlled pH and ionic strength. At each stage a different fraction precipitated out: albumin first, then gamma globulins, then fibrinogen, then thrombin, then half a dozen smaller proteins. The fractions could be lyophilised into white powders, packed in cardboard boxes, and shipped to North Africa or the Pacific. The Cohn process is in essence still how plasma fractionation is done today, eighty-five years later.

Albumin was the first product to leave the plant, and it was albumin that flew to Pearl Harbor. Albumin holds water in the bloodstream. A burn victim loses fluid through the cooked skin and into the soft tissues until their veins collapse and their heart cannot fill. Run human albumin into the deep veins of a burned man and the water comes back out of the swelling. The blood pressure climbs. The heart fills. The man wakes up.

By the time the Allies landed in Normandy in June 1944, the Cohn plant had been replicated as five industrial facilities across the United States. Two million of the thirteen million units of plasma that the American Red Cross collected during the war were turned into human albumin. Tens of thousands of soldiers came back from beaches and reefs because of the bottles in the medics' packs.

The plant produced one other thing in 1944 that did not get the press the albumin got. It was a small dehydrated lump of fibrin, the protein that gives a normal blood clot its scaffolding, prepared in an open-pore sponge form. Two of Cohn's younger collaborators, the neurosurgeon Franc Ingraham and the pathologist Orville Bailey, working at the Children's Hospital and the Peter Bent Brigham in Boston, had taken the fibrinogen and thrombin coming off the Harvard pilot plant and developed a method of using them in the operating theatre.

The procedure was simple. The surgeon cut a small piece of the fibrin foam, dipped it briefly in a solution of thrombin (also a Cohn fraction), and pressed it against the bleeding tissue. The thrombin converted the fibrinogen in the foam to fibrin almost instantly. The fibrin polymerised into a clot inside the sponge. Within twenty seconds the bleeding stopped. The fibrin foam, being itself fibrin, was eventually absorbed by the body. Nothing had to be removed.

Ingraham and Bailey reported the technique in the very first issue of the Journal of Neurosurgery, in July 1944. They were specific about the application. The fibrin foam was for brain surgery, where the bleeding tissue cannot be clamped or sutured the way an artery in a leg can be clamped, where electrocautery risks roasting the very tissue you are trying to preserve, where the surgeon may have only seconds before the patient is brain-dead. Time magazine wrote it up as "blood foam and film." A companion product, fibrin film, was used as a dural substitute and to prevent the brain from sticking to the skull as the dura healed.

Bleeding from inside the skull had been, for most of the four millennia of recorded surgery, a death sentence. After 1944 it was a problem with a tool.

Stopping bleeding has always been the first problem of surgery. Egyptian physicians cauterised wounds with red-hot iron in the second millennium BC. Galen, in the second century AD, recommended vinegar and pressure. The young French army surgeon Ambroise Paré in 1536, after he ran out of the boiling oil that was then the standard cautery treatment, dressed a battlefield's worth of musket wounds with egg yolk and turpentine, slept on it, and woke to find that the men he had treated with the oil were screaming and the men he had treated with egg were comfortable. Paré spent the rest of his career arguing for ligatures (silk threads tied around the bleeding artery) instead of cautery. Joseph Lister published his antisepsis paper in 1867. The twentieth century invented the haemostatic clamp, the electrocautery, and the gelatin sponge. Each of those advances was, at the time, considered the final answer to bleeding. Each was overtaken.

Last week, in the issue of Nature published on the twenty-ninth of April twenty twenty-six, a team led by the mechanical engineer Jianyu Li at McGill University in Montreal, with the lead author Shuaibing Jiang, reported a new way to make a clot. They call it click clotting, after the family of bio-orthogonal chemical reactions to which it belongs.

The click-clot is built out of red blood cells. A solution of donor or autologous red cells is mixed with a small molecule that places matched chemical handles on the surface proteins of the cells. A second molecule, added with a syringe, locks the handles together in a fast, biocompatible reaction. The cells crosslink into a solid gel in five seconds. The gel is thirteen times more resistant to fracture than a normal blood clot, and four times more adhesive to bleeding tissue. Tested on rodent liver lacerations, it stopped the bleeding and held. By day fourteen there was minimal inflammatory infiltration. By day twenty-eight the liver had normal tissue architecture and a healed scar.

Allogeneic preparations of the click-clot can be made in about ten minutes. Autologous preparations, which use the patient's own red cells, in about twenty.

The arc closes. Eighty-two years after a fibrin sponge dropped into a thrombin bath stopped bleeding in twenty seconds, a click-clot stops it in five. The 1944 sponge was a leftover product of a national wartime plant Cohn ran out of a Harvard basement. The 2026 click-clot is a designer object built by mechanical engineers in a Montreal lab. The story is the same. The answer keeps getting better.

Every generation of surgeons believes it has finished the war on bleeding. Each generation is wrong. The next generation will go faster.

Bailey and Ingraham would, one suspects, have approved.