In astronomy, we sometimes make discoveries that really make us scratch our heads. Observations that appear to contradict what we believe we know about the Universe. These kind of discoveries are always exciting, as it gives us an opportunity to learn new things about the Universe, and gain a deeper understanding of the world around us. It is certainly something that most astronomers hope for, at some point in their career. That whatever you are researching into, does something unexpected. I mean, who wouldn’t want to have your name associated with a new astrophysical phenomenon?

Thus, one can imagine the excitement, when, in 1962, the French astronomers Daniel Barbier and Nina Morguleff looked at a spectrum they had taken of the G5 dwarf star HD117034 (yes, we’re good with names), and observed two bright emission lines of neutral potassium. A spectra taken the following night, showed no sign of these emission lines. They quickly wrote a summary of their observations and announced the discovery of this new class of “potassium flare stars” in the Astrophysical Journal.

It is not an unknown phenomenon that low-mass stars exhibit flares. These intense, short, outbursts of energy we know even from our own Sun, the solar flares, which are typically followed by coronal mass ejections. However, flaring stars often show emission lines from multiple elements, like silicon, iron and oxygen. Thus, it was a surprise to discover this new class of stars, that seemed to show emission only in the same two potassium lines. Over the next couple of years, another two potassium flaring stars were discovered, both observed from Observatorie de Haute-Provence, as was the case for the initial discovery. Again, the potassium emission lines were only visible in a single spectrum. Adding to the mystery was the fact that the three flaring stars were of wildly different spectral types, namely a G5 dwarf, a K7 dwarf and a B9 dwarf, where particularly the latter stands out, as B-type stars pumps out a lot of energy, meaning that potassium should be fully ionized and no emission lines from neutral potassium would be expected.

Nevertheless, since this had now been observed in different stars at different times, it would appear that a real, transient, astrophysical phenomenon had been discovered, although no one was able to explain it. This spawned a lot of interest, and an extensive search was carried out by Robert F. Wing, Manuel Peimbert and Hyron Spinrad at Lick Observatory, to discover more of these mysterious objects. Although they surveyed 162 stars, not a single potassium flare was observed. Thus, the head-scratching began. Had the astronomers who discovered the three flare stars just been incredibly lucky and made a serendipitous discovery? Or might the explanation be something a bit more down to Earth?

It had previously been suggested that the emission lines could be originating in the Earth’s atmosphere, but in this case the emission lines should have been very narrow, which was not what was observed. The astronomers at Lick Observatory then began looking for other sources of potassium emission. After some experimentation, it was found that the observed emission spectra matched quite closely the spectra of ordinary matches, if the light from the matches were able to reach the spectrograph. Several different types of matches were used, and all produced similar results. The same thing was also tested at the Observatorie de Haute-Provence, and it was found that, in certain positions, the light from the matches did indeed reach the spectrograph. The full paper can be read here.

Thus it would appear that the potassium flaring stars was in reality a potassium flaring member of the observatory staff, who had lit up briefly in the science literature. Following this, I bet this observatory may have been the first to see a smoking ban on the premises. Or at least anywhere near the spectrograph.

This is also a perfect example of how astronomy, and science in general, works when something weird and unexpected it found (remember the super luminal neutrinos?). Such findings need to be independently confirmed and checked rigorously, to rule out other potential explanations. In many cases these things can be explained by already known phenomenon, or an overlooked mistake somewhere, but sometimes genuinely new things are discovered, which is amazing!

As a side-note, the extensive testing of the spectra of matches also showed that French and American matches were very similar products, in terms of composition, so there’s no blaming foreign matches when you can’t light your fire.