Until late May, 1999, I thought that James Prescott Joule's 1869 letter to the Manchester Literary and Philosophical Society was the earliest scientific publication on green flashes. Then the Inter-Library Loan people came up with a paper that had been written by one Pietro Giuseppe Maggi in 1852, which I'd found in the Catalogue of the Royal Society.
Actually, it wasn't a paper, and it wasn't written by Maggi; it was an account of a talk he had given to the Venetian Institute of Science, Letters, and Arts. Really, it's little more than an abstract of his talk. But it's enough to show that Maggi had discovered the blue flash at sunset over the mountains, had observed the red and blue rims telescopically and correctly attributed them to atmospheric dispersion, and had noticed that they are enlarged enough by distortions of the solar disk to become visible to the naked eye — a formidable set of discoveries. Furthermore, he connected these blue flashes with the pieces of the Sun that appear separated from the disk and then melt away, which I would call mock mirages.
As you can imagine, I was excited to have discovered such a thorough piece of work on sunset flashes, done some 17 years before Joule's letter was written. But I was nagged by the feeling that I'd seen Maggi's name somewhere before. Had someone else scooped me on this discovery?
I thought about this occasionally for a week or two without coming up with anything definite. Then, while cleaning old, unfilled ILL requests off my desk (in preparation for adding the unfound items to my wanted list), I noticed that I had already asked for this item back in April and May of 1996 — almost exactly three years earlier. The old request even gave the Roy. Soc. Catalogue as the source of the reference. So why hadn't I gotten Maggi's abstract three years sooner?
Then it came back to me. That was an item I'd asked for a couple of times, without success. The reply always came back: Not found as cited. That usually means the reference is screwed up; often, the volume number doesn't match the year given for the journal named.
So, how had they found it this time? On thinking about it, I remembered that again I'd had to ask twice. The first attempt failed, but the second one worked. Why?
Well, it seems that the first time I tried (in 1999) I simply copied the reference from the citation in the Physics Index to the Roy. Soc. Cat. But, as it was excessively cryptic (“Ven. At.” or something equally elliptical), I consulted the list of journal abbreviations in the front of the index, which gave me the full title of the journal. Yet that still hadn't worked.
What did work was to look up Maggi's paper in Volume IV of the Catalogue, and then check the journal abbreviation given there to the full list at the front of Volume I. There, I finally found that the journal had started a second series in 1850; and it was the detailed specification of “second series” in the request that allowed the ILL people to find the paper at last.
Obviously, I hadn't gone back to Vol. I in 1996, and hadn't learned that an 1852 paper must be in the second series. So of course the volume and year didn't appear to match, and my request had bounced.
So, what have I learned from all this? I'd already learned (in the intervening 3 years) not to rely on the brief journal abbreviations used in the Roy. Soc. Catalogue; so the moral of the story must be something else.
Obviously, if I had been careful enough, I could have come up with this gem three years ago. So one lesson is certainly to be as careful as possible, all the time. Sloppiness costs you lost time.
Maybe the moral is that persistence pays off. Or that letting a problem lie there until you have forgotten about it, and can make a fresh start, is sometimes the fastest way to get things done.
If intelligence is being able to learn from the mistakes of others, let me know what you've learned from my mistakes here.
© 1999 – 2007 Andrew T. Young
If you have comments, please let me know.