Aristotle wrote about eels in Historia Animalium around 350 BCE. He observed that they had no testes, no ovaries, no obvious reproductive organs of any kind. He concluded — reasonably, given what he could see — that eels must arise spontaneously from mud. Specifically, from the entrails of the earth.
This was wrong. Twenty-three hundred years later, we know it’s wrong. We also still cannot reliably explain, in much detail, how eels reproduce.
What we know now
The European eel (Anguilla anguilla) and the American eel (Anguilla rostrata) are catadromous: they spend most of their lives in fresh water and migrate to the ocean to spawn. The spawning ground is the Sargasso Sea, a region of the western Atlantic delimited not by land but by the gyre of currents that surround it. Johannes Schmidt, a Danish biologist, spent eighteen years on twenty-three expeditions narrowing this down between 1904 and 1922. He never witnessed mating. He inferred the spawning location from the geographic distribution of the smallest larvae — leptocephali — that he could find. Where the smallest larvae cluster, the eggs were laid. The clustering pointed to the Sargasso. Nobody has ever, before or since, observed European eels mating in the wild.
The leptocephalus larvae drift across the Atlantic on the Gulf Stream for one to three years. They look nothing like adult eels — they are leaf-shaped, transparent, and were classified as a separate species (Leptocephalus brevirostris) until 1896, when an Italian zoologist watched one metamorphose into a glass eel in a tank. The glass eels enter European estuaries, become elvers, become yellow eels, and remain in fresh water for somewhere between five and twenty years. Then, on a trigger that has been described but not in any way mechanistically explained, they undergo a final transformation. Their eyes enlarge. Their gut atrophies. They stop eating. Their skin silvers. They migrate, in autumn, downstream to the sea, swim three to four thousand miles back to the Sargasso, spawn, and die.
The reproductive organs on which Aristotle stalled — and on which Sigmund Freud, in his first published paper in 1877, also stalled while searching dissection trays in Trieste for male eel testes — develop only during this final migration. A yellow eel pulled from a lake has nothing to find. The organs are produced after the trigger fires, and the trigger fires once.
What we don’t know
The list is uncomfortably long.
We do not know what initiates the transformation from yellow eel to silver eel. We have correlations with body length, condition factor, age, and water temperature, but no individual eel can be reliably predicted to silver in any given year. Some eels remain in fresh water for fifty years without silvering, then die without reproducing. Others silver at five years. The mechanism is not understood.
We do not know how silver eels navigate. The journey from a Norwegian lake to the Sargasso involves three to four thousand miles of open ocean, in deep water, in the dark, by an animal that has never made the trip. The current best guess involves geomagnetic orientation combined with olfactory cues, but the only experimental evidence we have is from short-distance manipulation experiments that do not scale to the actual journey. The animal does it. We cannot say in detail how.
We do not know what eel mating looks like. We do not know whether it occurs at depth or near the surface, individually or in aggregations, with what courtship behavior, on what timing relative to the lunar cycle. We have models. We have inferences from genetics. We have direct observations of zero events.
We have not, except in research conditions that are nowhere close to industrial, succeeded in breeding eels in captivity. The first complete life cycle was closed in captivity in Japan in 2010, with hormone injections, careful temperature control, and a feed slurry made from spiny dogfish eggs that took thirty years of trial and error to formulate. The yields are vanishingly small. The world’s eel aquaculture industry — a multi-billion-dollar business serving Japanese, Chinese, and European markets — is still entirely dependent on glass eels caught in the wild, then grown out in tanks. The wild glass eel populations have collapsed by approximately 95 percent since the 1980s. Nobody can quite produce the substitute.
Why the gap doesn’t close
There is a category of scientific question whose persistence is not a function of effort but of empirical access. The eel sits at the center of it. The reproductive biology of the eel is unobservable in the wild because the wild reproductive site is approximately two kilometers underwater in a region of ocean that has neither landmarks nor a navigable bottom. It is unreproducible in captivity because the trigger conditions for sexual maturation are themselves part of the unknown. We cannot provoke the animal to do the thing we are trying to study, and the wild does it in a place we cannot watch.
This is structurally different from problems that yield to better instruments. We have better instruments than Aristotle, better instruments than Schmidt, and we are not meaningfully closer to filming an eel mate. The blocker is not resolution. It is the geometry of the system. The eel reproduces, once, in a location it took the field eighteen years to narrow down, and the act of being captured for observation appears to interrupt the very physiological cascade that produces the gametes we want to see.
There is a further, awkward implication. Most of the canonical facts about eel reproduction in textbooks are inferences from auxiliary data — larval distributions, hormone profiles in pre-spawning silver eels caught en route, post-mortem dissection of females intercepted on migration. We have stitched together a plausible picture from the things we can see. Whether the picture is correct in its specifics is not testable, and probably will not be testable in our lifetimes. The eel question is not unsolved because it is unimportant. It is unsolved because the system is, in a strong sense, hiding.
A category, not a curiosity
There is a small but real class of questions that are this kind of unsolved. The atmospheric chemistry of the upper Venusian cloud deck, where temperature and pressure are nearly Earth-like but every probe melts within hours. The neurology of slow-wave sleep in cetaceans, where unihemispheric sleep is observed in captivity and the same protocols cannot ethically be run on free-swimming whales. The genetic dispersal of the Polynesian sweet potato, which appears to have crossed the Pacific from South America before European contact in a way that the canonical archaeological models do not allow.
These questions persist because the data we would need to answer them lives in a place that resists being observed. The eel’s spawning ground is the cleanest case because the inaccessibility is geometric and the animal is otherwise extraordinarily well-studied. We have catalogs of eel fisheries going back five centuries. We know the toxicology of the flesh. We know the parasites. We know the chromosomal karyotype. We know everything except the thing the entire life cycle is organized around.
The temptation, in a culture used to questions yielding to compute, is to treat persistent ignorance as a temporary failure of measurement. Sometimes it is. Sometimes it is not. The eel has been beating us since Aristotle, and there is no instrument under development that I’m aware of which is going to change that in the next decade. We will most likely close the gap, when we close it, by someone getting a fully gravid female and a fully gravid male in the same volume of water at the right temperature in the right season — and even then, what we will see is an act of reproduction in a tank, which is not the same act that occurs in the Sargasso.
What you have, after twenty-four hundred years, is a fish that is on European supermarket shelves, that we eat by the metric ton, whose reproductive biology has resisted Aristotle, the Roman naturalists, Linnaeus, Cuvier, Freud, Schmidt, the entire postwar Japanese aquaculture industry, and modern molecular ecology. The eel is not a hard problem. It is a problem with a different shape than the shape our instruments have.
I find that genuinely interesting. I also think more questions are like this than we admit, and I think we mostly stop noticing the ones our generation cannot solve, on the grounds that the next generation might — even when there is no clear path by which they would. Aristotle assumed the next generation would solve the eel. We are the next generation, twenty-three centuries deep. We have not.