Sky & Telescope, Dec. 1984

The problem of solar neutrinos has been with us for 16 years now, and it apparently ir not going to leave without a fight. Neutrinos are the only particles created by nuclear fusion in the Sun's core that leave it unaltered. Traveling at the speed of light, they arrive at Earth in eight minutes and provide the only direct test we have for nuclear fusion in the closest star. The problem is that we're detecting only a third as many neutrinos as theory predicts.
In 1968 Raymond Davis of Brookhaven National Laboratory set up a 100,000 gallon tank of cleaning fluid (665 tons of it!) in the Homestake Gold Mine in South Dakota. This expcriment to detect solar neutrinos still runs. The latest counts are 2.1. give or take 0.3. solar neutrino units (SNU's). Current predictions, according to ]ohn Bahcall at Prinston's Institute for Advanced Study, are 6.8 plus or minus 2.8 SNU's.
In a recent talk to the American Astronmical Society, Bahcall said he counted around 40 proposed solutions to the problem, at least 39 of which must be wrong! In general, the solutions reduce to three alternatives. Maybe our theoretical model of the Sun is wrong; in some curious way the Sun could be opaque to neutrinos, or it could have a slightly different composition than we believe. Maybe ideas about the neutrino itself are inconrrect: this particle might change form between the Sun and Earth, and Davis detects only one fom (S&T: August. 1980. page 115). Or maybe the theory of fusion is wrong; the exact process that produces neutrinos detectable in Davis' apparatus (a rare side process of the fusion chain, the decay of boron-8) is exquisitely senritive to temperature. A small difference in temperature from that calculated would make a huge difference in the number of boron-8 neutrinos.
An experiment substituting 45 tons of gallium for the cleaning solution would decide which alternative in correct - the third seems to be the favorite. A gallium detecfor would be sensitive to neutrinos coming from the main fusion process, the proton-proton chain. But with a price tag of $22.5 million, the Department of Energy and the National Science Foundstion have both declined funding.
Bahcall, backed by a computer simulation, now says that an experiment with only 15 tons of gallium. running for three years, could count the neutrinos from the proron-proton reaction with a 15 percent error, all for $5.25 million. Will the Department of Energy buy that? According to Davis. "It's really hard to say." In any case the Soviets have 60 tons of gallium available, plus the resources of what they call Neutrino Village. They intend to run a gallium experiment soon and have offered to collaborate.
Meanwhile, Davis and colleagues have discovered what they think is an interensting pattem in the data. The neutrino counts seem to vary inveresly with the sunspot cycle. The counts dropped off from 1977 to 1979, a period when the Sun was getting more spots, then rose again from 1979 to the present, when the spots were declining.
Two problems are apparent. In the first place, any correspondence between neutrino rates and sunspot numbers in difficult to reconcile with what we know about the Sun. The spots occur only on the Sun's surface. while neutrinos come from its core. Since the core is supposed to be stable, the neutrino counts should remain steady. Secondly, any inverse relation between neutrino counts and the sunspot cycle needs, and lacks, explanation. Davis points out a possible relation between sunspots and the solar diameter: if the Sun gets larger, it would have fewer spots.
"All measurements involved are kind of ragged." said Davis. "but the fact that they sort of fit together is interesting." Bahrall agreed that the counts could conceivably vary with the sunspot cycle, but he notes that "the statistics don't demand that they do." Says Davis. "The situation's a little vague right now, but if it's true if would change our thinking about the Sun."

By the time we were in mid-cycle, various adherents to the standard model were making predictons...