TERJE LØMO’S TIME AS A PHYSICIAN IN THE NORWEGIAN NAVY WAS COMING TO AN END, and he was looking for his next job. Walking from one Oslo hospital to another, Lømo ran into neuroscientist Per Andersen, who was looking for a few good scientists to launch a lab at the University of Oslo. The year was 1964, and their project would be to explore the neurology of the hippocampus, a brain region thought to play a role in learning and memory.

Memory had always been a great physiological mystery. Aristotle thought of memory as impressions made in wax, but if this were the case, what is the “wax” in the brain that holds them? In the early 20th century, researchers began to search for signs of an engram, the theoretical manifestation of a memory in the body. In 1949, Canadian psychologist Donald Hebb suggested that such a trace might be found in the way that neurons connect themselves. When an experience activates a group of neurons, he theorized, they might form stronger connections with one another. That bond between neurons might itself act as the storage of a memory.  In other words, “neurons that fire together might wire together.”

In 1953, an American epilepsy patient named Henry Molaison had several sections of his brain removed, including most of the hippocampus. Although the procedure curbed his seizures, Molaison could no longer form new memories. This was possibly a clue to where the wired neurons of memory might be formed.

As a Ph.D. student in Andersen’s lab, Lømo began working through the layers of tissue in the hippocampi of anesthetized rabbits, stimulating one cell after the next. Andersen had already found that, along certain pathways, a researcher could stimulate a neuron and watch that signal grow stronger after a series of stimulations. But while this seemed like evidence of “wiring together,” the effect lasted for only seconds and wasn’t an example of the kind of long-term storage a memory would require.

In 1966, Lømo sent several high-frequency stimulations, spaced seven minutes apart, through a single neuron. With each successive stimulus, Lømo found, the response in the next neuron became stronger—an effect that lasted for hours. Sitting by himself in the lab, late at night, he stared at the screen of the oscilloscope he used to track the neurons’ electrical fluctuations, becoming the first human to witness memories being physically written on the brain.

That process, confirmed in Lømo’s experiments, came to be called “long-term potentiation,” or LTP. In the 1980s, other researchers would discover the neurochemical components of LTP. When a neuron receives repeated doses of a neurotransmitter for another cell, it triggers the development of additional receptors. Those additional receptors become available to pick up the message of the neurotransmitter, and so a stronger connection forms between the two neurons.

Lømo’s discovery paved the way for the discoveries about the mechanisms of memory that are being made today. Cumulative lifetime marijuana use, for example, has been associated with poorer verbal memory in middle age. A study from January found that certain synthetic formulations of marijuana significantly interfered with LTP in the mouse hippocampus, because they inhibit the release of neurotransmitters. Researchers believe this process contributes to impaired memory formation.

Understanding more about the physical processes of memory may transform therapy for patients with Alzheimer’s disease and other forms of memory loss. And while much more remains to be understood about the mechanics of memory—how memories are stored and retrieved by the conscious mind, for example—Lømo’s contribution gave the first answer to what had been a constant riddle since ancient times.