The photosphere, the Sun's visible surface, radiates like a blackbody at about 6,000 K. Move outward and the temperature ought to fall. Across a transition region only 100 kilometers thick, above the chromosphere, it leaps instead, to between 1 and 3 million K — the Sun's outermost and most tenuous layer running hundreds of times hotter than the surface it sits on. Heat does not flow spontaneously from a cooler body to a hotter one, so nothing conducting upward from the surface can account for it. Something must be depositing energy up there continuously, and it has been traced to the Sun's magnetic field.
That much is not in dispute. The corona's measurement history is the oldest of the three, and it opens the way the first chapter did — with an instrument recording an answer to a question nobody had asked.
At the total solar eclipse of August 1869, early spectrographs trained on the exposed corona returned a bright green line at 5303 Å matching no known element. For seventy years it was assigned to a hypothetical element, coronium. In 1939 Walter Grotrian proposed instead that the line came from familiar metals stripped of many electrons, a state that takes extreme heat. Bengt Edlén confirmed it in the laboratory in 1940–41: the green line was iron stripped of thirteen electrons, and the corona therefore exceeded a million degrees. Sounding rockets confirmed intense solar X-ray emission in 1949, and a puzzle in chemistry became a problem in plasma physics.
What is disputed now is the mechanism, and two frameworks have held the field for decades with real observational support on each side. Wave heating has convective churning at the photosphere launching magnetohydrodynamic waves up the field lines into the corona; the waves are observed to damp low there, and what fraction of the heating budget they carry is unsettled. Nanoflare theory, formalized by Eugene Parker in 1988, has convection braiding magnetic field lines until they reconnect in innumerable small, impulsive bursts, each releasing stored magnetic energy as heat.
Nanoflare theory has the advantage of a clean test, which few questions here have. For nanoflares to dominate the heating rather than merely contribute, the power law relating flare frequency to flare energy must have a spectral index above 2. Measured indices across different studies range from about 1.4 to 3.3 — not a cluster near the threshold, which would be a hard measurement of a borderline case, but a spread across it. The test is well posed, and the data answers it both ways at once.
The reason a century and a half has not closed it is by now familiar. The events at the heart of both theories happen at scales of meters to kilometers over fractions of a second, far below anything a telescope can now resolve from 93 million miles away. They blur into one smooth average glow. Solar Orbiter's Extreme Ultraviolet Imager has watched some large coronal loops hold a steady, nearly un-bursty state where nanoflares would predict flickering. But no current mission can measure the magnetic field directly in the chromosphere and transition region — the altitudes where the disputed energy is stored and released. Instruments reliably measure only the photospheric field below. The magnetic free-energy budget above it is inferred.
Set the three side by side and one shape repeats. Two rulers whose gap widened as both grew more precise. An alignment that cannot be tested further in the temperature, because the universe holds only five independent numbers at that scale. A furnace whose mechanism runs below what any instrument can now resolve. In each case the missing thing is not effort or care but access to the place where the assumption lives.
They are not inaccessible in the same way, though. A regime before recombination is closed to light in principle; a sample of five is fixed by there being one universe; a scale beneath resolution is only beneath current resolution. Which makes the corona the one case here where patience is a strategy, not a posture.
What the corona adds is duration. The question has been open since 1869, and holding it has not been idle. The green line was identified, coronium retired, the discipline moved from chemistry into plasma physics. And the question changed shape, from why is it hot to which mechanism, and in what proportion — a narrower question, asked with better instruments, that did not exist in 1869.
Seventy of those years were spent confidently describing an element that does not exist. It is tempting to call that time lost. It was not. Coronium was the best available account of a real signal, and it was the kind of account that could be killed. A named element is a claim about what a laboratory ought to find. Not knowing what makes a line is a claim about nothing, and could have been held indefinitely at no cost. The wrong answer is what gave Grotrian and Edlén something to overturn.
The corona also shows what partial resolution looks like from inside. The temperature question closed. The mechanism question did not. An open question is rarely one thing; it has parts, they close at different times, and the closing of one is what makes the next askable. An unresolved measurement is not a gap in the record. It is a part of the record that is still being read.
Argued from