About the Annotated Green-Flash and Mirage Bibliography


For many years, I have been collecting references on atmospheric refraction and sunset phenomena, starting with papers I needed to calculate airmass, back when I was reducing astronomical photometry. Then when the sunset work began, I started a file on distorted sunsets; interesting things that didn't belong there got put in file folders marked “Green Flash” or “Dip” or “Wegener” or “Mirage File”. The more useful references were collected in a computer-readable index called “refraction.”

As I learned more, I realized that these topics were inseparable: astronomical refraction, airmass, the refractivity of air, dip of the horizon, mirages, green flashes, etc., etc. — so all are combined in one monster “refraction” file (now about 2.5 MB, and slowly growing.)

If you want a quick peek at all this, click away. Just remember, 2.5 MB takes a while to work its way over the wires. And it's too big to print (over 700 pages), so don't try.

You may be baffled by the way I've arranged these thousands of references, so you might want to read a little more about what it is before deciding to hit that mouse button. I strongly recommend that you read this introduction thoroughly before trying to make use of the bibliography.

Paper files vs. computer files

Almost everything here is a reference I have in hard-copy form, usually a photocopy of the original article. I currently have over 1000 references on green flashes, and hundreds more on the related topics.

Even so, this is not a complete list. I have tried to be as complete as possible for the green-flash references, but have only bothered to type in the most interesting papers on mirages, etc. There are more reprints in my paper files than entries in the computer file, except for green flashes, where I have tried to keep things one-to-one.

Some idea of the completeness of the green-flash list can be gleaned from the list of 17 GF references Shaun Hardy sent me in August, 2000, from the files of the Department of Terrestrial Magnetism of the Carnegie Institution of Washington. Of these 17, I already had 13 in my file, and 4 were completely new to me. That suggests that the literature is at least 17/13 or 1.3 times as large as the thousand-odd items in my list at that date, which means there are still 300 or so yet to be discovered. Probably this is an underestimate, if one allows for items in newspapers. Still, there is the hint that I have appreciably more than half of the whole GF literature. Of course, in other areas of refraction, the present bibliography is only a small fraction of what exists.

In particular, I have only a superficial knowledge of meteorology and atmospheric physics, so those topics are largely missing from the bibliography. Likewise, I know I've missed lots of papers in the field of navigation, where lots of mirage and other refraction phenomena have been reported. And the literature of surveying is full of reports of anomalous refraction measurements, most of which I have not seen.

I have read every word of nearly everything here; there are a few long papers and books that turned out to have only a short section of interest, so I skipped the dull parts. I've also read a couple of hundred other papers (mostly routine mirage observations), but didn't find them interesting enough to include here (i.e., I figured I'd never cite them anywhere.)

The annotations

Obviously, you can't keep track of all that in your head, which is why I started the computer file in the first place. For my own benefit, I made some notes to go with each reference. Most of these are simply a brief excerpt or salient quotation; not all of them should be taken at face value, for I have excerpted some outrageous remarks from some of the singularly ill-informed writers.


This is the place to warn prospective users: NEVER copy items off a list like this and think it takes the place of reading the original papers. Your take on any given paper will probably be different from mine; what I found dull, you might consider crucial — and vice versa. ALWAYS read the paper yourself before you make use of it; don't take my word, or anyone else's word, for anything.

You'll also find some editorial comments of mine from time to time, indicating what I think of a paper or its author, or how different papers in the list are related. These are my own personal views, and should not be attributed to any sponsoring agency or organization.

You're looking over my shoulder as I work, so to speak. The annotations are notes for my own use. Another person would probably select different excerpts as the most significant in most papers. As you read along, you'll notice ideas for papers I intend to write in the future, and occasional unresolved questions. This is the raw material of research, not the finished product.

Why should I make this sort of half-baked stuff available to the world at large? Well, William Beasley urged me to publish an annotated bibliography on “the green flash,” and asked that it be complete . As he was doling out money to support the work at the time, I figured it was a Good Idea to make him happy, and tried  to make that part complete. And it was  a good idea to make the references available to other people; we shouldn't all have to dredge through the unreliable indices and reviews again and again. Besides, there are lots of papers here that can't be found in any of the standard indexes. You can regard the green-flash part of this as a further step in the work started by Mulder, Fisher, and O'Connell.

As hard-cover publication seems unlikely for something so protean, the Web seemed the best place for it. I'd also been sending versions of the bibliography to George Kattawar from time to time, and he too thought it was worth posting to the Web pages. So, here it is.

What's here and what's not

What's here is much of what I've read: everything that is relevant to green flashes, and a lot of interesting and useful papers on mirages, dip, atmospheric refraction, and so on. I've deliberately omitted a lot of routine mirage reports, and a few papers of marginal interest on other subjects. There are also several paper files of marginal relevance to the task at hand that have been left out. But everything I have found on green flashes is here.

But of course there are still things I've missed, and the search for items is continuing. If you're actively interested in this sort of thing, stop by in a couple of months and you'll find some newly added material. Several errors in the formatting software have been corrected; the references you now see are in many cases more complete than before. If you'd like to help, I have a list of wanted items I haven't been able to find.

Format and presentation

In my own file, I have the comments before  each reference. In the HTML version, I thought it was more readable (and more consistent with the style of the reading page) to put the references first, followed by the annotations. This may cause some confusion; when I say “the above” I mean the previous  citation, not the paper cited directly above my comments on it. Citation/comment pairs are separated from neighboring pairs by a horizontal rule, so you can be sure which comments go with which reference.

The annotations are comment lines in my file, marked by asterisks in col. 1. It seemed easiest to let the asterisks stand in the HTML version; this is not intended to be a work of art, but only a ready reference for people really interested in these phenomena.

I've also indicated the beginning of each of my paper files by enclosing a title for it between horizontal rules. Each file-title line also contains the word FILE in caps, which sometimes is useful in scrolling back and forth, to find out which paper file a given photocopy is in.

How it's arranged, and why

The computer file is intended to be a faithful mirror of my paper files (see above). I have the habit of arranging files in chronological order; so within each section [whose comments start with a line containing XXX FILE or some such label] the items are arranged with the oldest first. The very beginning is a group of very old items, starting in antiquity. When the density of papers got to be enough for topical subdivisions, I split up the material into categories like mirages, green flashes, dip, and so on.

In a few cases, some people have been so influential that I have grouped their papers together separately. So I have separate files on Alfred Wegener, Colton and Chappell and Fisher (along with related papers on distorted sunsets), Lehn (and other Novaya Zemlya work), Forel and the Dufours, etc. I should probably also have done this for J. B. Biot, William Groff and a few others.

The chronological arrangement has some obvious drawbacks, and a few great advantages. One advantage is that you can easily follow the arguments that raged about green flashes in the pages of English Mechanic, or Symons's Met. Mag. Furthermore, when the correspondents in these fights refer to other publications (such as the passing mention of T.P.'s Weekly in the English Mechanic series of 1904), you have the relevant articles nearby. Often, papers in one journal were quoted or excerpted soon afterward in another; or authors referred to their earlier works. The chronology of such cross-references is (I hope) correctly preserved here; while it's fairly easy to get the years in order, it's more difficult to order papers within each year.

It has become apparent, after filling in a lot of these bits and pieces, that the green-flash literature is considerably more tightly connected than (say) O'Connell's account would lead you to believe. Though there are many papers that have never been cited, there are also many that have had considerable influence on later writers. In many cases, I have notations to indicate later citations of earlier papers. For example, “O'C #45” would denote reference no. 45 in O'Connell's bibliography.

It will become obvious, as you read through the bibliography, that a lot of papers could have been put in any of several different files. The particular pigeonhole I actually chose depended on what I was thinking about when I filed the photocopy; don't expect anything like consistency here. Some mirage papers mentioned green flashes, and got put in the green-flash list (but only if the GF reference was substantial); others deal with dip and other refraction phenomena. The phase of the Moon and what I had for lunch are probably also significant independent variables. So, be careful; don't expect to find all the references to mirages in the mirage file, or all the stuff about dip in the dip file. If you're trying to find all my references on some such sub-topic, use the search function of your browser; don't rely on my file-folder headings.

Bear in mind that this is a reflection of my own paper files, arranged according to the way the subject is organized in my own head. It isn't arranged in a way that's convenient for anyone else to use; to do so would require a major amount of work, and would also require that I “freeze” the bibliography at some point instead of keeping it up to date. This is a working document, not a polished finished product. Proceed at your own risk.

If you want to use the bibliography, I suggest that you download a local copy for yourself, and then employ whatever search tools you happen to be used to. (I'm a *nix person, so I use grep and the pattern-search features of less and vi.) It certainly doesn't make sense to print it all out, and try to page through it manually: the last person who tried to print it found it was over 700 pages!

I add an item or two a month to this list, so you might want to get a fresh copy a few times a year.


One of the nuisances of interdisciplinary work is that every field has its own journal abbreviations, evidently as a means of keeping the riff-raff at arm's length. Astronomers are no exception. So, out of habit, I use short abbreviations for some of the commonest journals:

A.N. = Astronomische Nachrichten

BSAF = Bulletin de la Société Astronomique de France = “l'Astronomie” after these two journals merged, about 1890

JBAA = Journal of the British Astronomical Association

M.N. = Monthly Notices of the Royal Astronomical Society

AJB = Astronomischer Jahresbericht

Occasionally I refer to J. C. Poggendorffs biographisch-literarisches Handwörterbuch für Mathematik, Astronomie, Physik mit Geophysik, Chemie, Krystallographie und verwandte Wissengebiete  (now partially available at Google Books; Vol. 2 is at http://books.google.com/books?vid=OCLC05073007&id=zXcJAAAAIAAJ) as just plain “Pogg.” Likewise, the Dictionary of Scientific Biography  is just the “DSB”.

Notice that both the Royal Astronomical Society and the Royal Meteorological Society have had a “Quarterly Journal;” so when an astronomer refers to “Q.J.” it means Q.J.Roy. Astr. Soc.; but if a meteorologist says “Q.J.” it is Q.J.Roy. Met. Soc. There was also once upon a time a “Quarterly Journal of Science.” So I have tried to avoid using the “Q.J.” contraction.

I hope the remaining abbreviations are clear enough for the general reader to understand. Let me know if you have problems figuring anything out.

Sources of copies

Most of the references cited are available only in large libraries. Some are fairly rare, and exist in only a handful of libraries. Occasionally, I indicate an on-line source for a particular item. Some recent technical articles are available online, but many of the older ones are hard to get — with a few important exceptions, such as Gallica.


The French national library has made available many old journals, and a few books whose copyrights have expired. These include C.R.  and the Mémoires  of the old Royal (i.e., pre-revolutionary) French Academy of Sciences; Proc. Roy. Soc.  and Phil. Trans . A few other journals of occasional interest here, such as Annalen der Physik , and Acta Eruditorum , are also available. Another valuable resource is the Catalogue of the Royal Society, which covers most of the 19th Century. Likewise, the cumulative index of the Mém . and Hist . is available. Unfortunately, little else of interest to me is available on-line as of Feb. 18, 2005. (A decade later, in 2016, the situation was only a little better.)

All these are available at the Gallica website. The catch is that most of the material is present only as scanned images of the pages; very little of it has been OCR'd. (Unfortunately, I had some difficulties in retrieving material from their site, until I solved a subtle configuration problem with my browser; see the Cookie Problem page for details.)


Another source of old volumes is the ADS site at Harvard. A particularly useful feature is the ability to search the full text of some works that have been OCR'd.

ADS is one of the most useful and user-friendly on-line sources. Unfortunately, they have concentrated on English-language documents; but many important works are in French or Italian or Russian and have not been scanned.


Google Books has scanned many old volumes; but the quality of the OCR work is poor, so you can easily miss items in a full-text search. One of their biggest drawbacks is the failure to identify the volumes of bound journals correctly. A trick I've found useful is to search for the names of a few months in each volume, hoping to find a title or contents page that would give the volume and year. A related problem is their mis-classification of many, if not most, scientific works.

One very serious problem in Google Books is their failure to open the foldout plates that were a standard feature in the 19th Century. The original volumes were designed to have the plates folded out, so you could read the text while looking at the engravings. A nice feature of this policy was that the same plate could be referenced repeatedly from several different pages of the text, so you could keep a plate opened out and read a long section of the text without having to page back and forth — an irritating problem with more modern books. A perceptive and Web-using reader will notice that this feature of 19th-Century printing made those old volumes a convenient sort of hypertext that is not matched by modern mass-produced books with figures embedded in the text.

The failure of Google to provide even marginally adequate scans of figures and plates in old books largely destroys the value of their scans, and has been harshly and justifiably criticized by Denis Roegel on the LORIA website. As of early 2020, Google Books is still not a reliable source. But they seem to be learning that this weakness needs to be remedied; a few books on their site now have the formerly missing figures.

Biodiversity Heritage Library

Some rare old journals devoted to the vague topic of “natural history” contain mostly papers devoted to botany, zoology, fossils, and geography; but they also contain some papers on navigation, astronomy, and meteorology, which can include matters of interest ranging from mirage reports to measurements of the dip of the horizon. The BHL often has high-quality scans of those journals, such as the Schriften der naturforschenden Gesellschaft zu Danzig , which is at the BHL at https://www.biodiversitylibrary.org/bibliography/14678#/summary.

The BHL has pretty good OCR text; but when you ask to download it along with the page images, what you get is a PDF of the bare page images, with the virtual pages of text added at the end. So this isn't a “searchable” PDF file in the usual sense. It's probably better to ask for the plain page images, and then add your own OCR, if you have a decent OCR engine available. (In 2020, the current stable version of Debian contains a pretty decent version of the well-established tesseract OCR engine, and the ocrmypdf script; so this task is possible if the page images are not too bad.) So the BHL's own OCR is useful mainly for finding interesting documents.

However, they seem to be improving. As of 2023, I find they have some important resources downloadable as PDF files with good OCR'd text. One fine example is all the volumes of the Royal Society Catalogue at https://www.biodiversitylibrary.org/bibliography/1931 . That reference is arranged alphabetically by author's last name, and chronologically within each author's list, which is convenient for tracing the works of a particular individual. The original set of volumes covered the period from 1800 to 1869; later supplements carried the indexing forward in time, and became the International Catalogue of Scientific Literature in 1902.


Lots of Google scans of old journals and books are kept at the HathiTrust digital library. Some of the scans are non-Google, and useful for general research; but because of Google's discrimination against pages of plates, and images generally, you can expect to find only badly-OCR'd text of most items.

The HathiTrust website serves as an inter-library loan pool for a group of participating libraries — mostly in North America — so if you don't have an affiliation with one of them, you won't be able to download many of their holdings.

Berlin Academy

The Berlin Academy has made some of its old publications available on-line as well. Fortunately, the Monatsberichte , which contain Kummer's wonderful paper, are finally available on-line.

Furthermore, as of 2021, I see that even more of the Berlin Academy's publications are available; see https://bibliothek.bbaw.de/digitalisierte-sammlungen/akademieschriften for the whole list.

Bavarian National Library

Another important on-line archive of German publications is at the MDZ at the Bayerische Staatsbibliothek. They have rare items like Bessel's record books of his observations at Königsberg, including the important observations of refraction near the horizon made by Argelander. To my great delight, these have recently been re-analyzed by Marcel Tschudin.


Finally, if you have access to JSTOR, you will have access to the Royal Society journals (i.e., Proc. Roy. Soc . and Phil. Trans .) in text-searchable format, as well as better-quality page images than Gallica provides. JSTOR has tried to provide high-resolution scans of figures; but, as the resolution of computer screens is much coarser than that of copper-plate engravings, these are often broken into sections that are impossible to reassemble — another sign that the printed page is far from obsolete. (JSTOR also has Science ; but that's of little use here.)


This site contains information about digitized mathematical and astronomical tables.

There is also an extended discussion of the requirements for productive use of digitization. Too bad the Google people haven't read this.

Other sources

The American Journal of Physics is available on-line, but your institution has to have a subscription to give you access.

A number of other societies that publish journals have similar policies (e.g., the Optical Society). These are useful only if you are an academic researcher with suitable library privileges.

Some commercial publishers of technical journals have similar policies. This makes access to some important old journals impossible, except through inter-library loan. An outstanding example is “Phil. Mag.”, which is difficult to find online; however, some volumes of it are available. There is a guide to them at the Online Books Page at the University of Pennsylvania. Its current publisher is Taylor & Francis, which has a restricted archive of past issues.

I'm happy to see that the U.K. Met. Office has scanned the volumes of Symons's Meteorological Magazine and made them available at


Historical indices

If you are interested in carrying this kind of work on, here are a few hints about how to proceed. Periodically, there have been attempts to summarize the history of astronomy up to the time of writing. A classic in this is the index of the journals published by learned societies before about 1800:

J. D. Reuss
Repertorium Commentationum. Tom. V. Astronomia
(Henricum Dieterich, Gottingae, 1804)

That's useful for articles in journals published by scientific societies, but useless for independent journals like A.N. and the likes of Crelle's Journal or Nicholson's Journal.

Another series of historical summaries is Delambre's volumes on the history of astronomy. He lived long enough to produce the history of ancient astronomy, the history of astronomy in the Middle Ages, and the history of "modern" astronomy (i.e., before about 1820); but he died before he could complete the history of astronomy in the 18th Century, which was edited and published by Claude-Louis Mathieu.

Already in the early 19th Century, astronomy was becoming too big a field for such synoptic reviews, and specialized works like Carl Christian Bruhns's 1861 history of astronomical refraction, which was later updated by Bemporad's 1907 encyclopedia article, became necessary. However, a final attempt to index all scientific literature was made by the Royal Society during the 19th Century; see the description of their Catalogue above, under the BHL.

But by the 20th Century, annual indexes like the Astronomischer Jahresbericht  (and its successor, Astronomy and Astrophysics Abstracts ), and Meteorological and Geoastrophysical Abstracts , became important sources of citations for the beginner in historical searches. Later, citation indexes appeared, and are currently a convenient way to trace the evolution of a topic forward in time from classic reviews of the past.


“To err is human . . . ,” so I'm sure there are many errors here. Some are errors of transcription, others are errors of translation. (I wrote out translations of the languages I don't read easily, and simply copied the German and French quotations, following Mulder's policy.) Sometimes I have only a very muddy photocopy to work from; if you have the original publication available, please check it when reporting apparent errors.

I've also found a host of copying errors, at first by chance while modifying the reference file, and later by using my multilingual spell-checking script. Occasionally I find and fix a page-number error, too. Even with these improvements, I'm sure there are mistakes I've missed.

There are also a few errors caused by the script that converts the original file into HTML. Most of these should be obvious; they are fairly few by now.

A fresh pair of eyes will detect mistakes faster than I can; so pleaselet me know when you find some. I hope I'll get them corrected quickly.

However, be careful. Many journals used to change volume numbers in the middle of a calendar year; so it isn't necessarily a mistake when two citations from the same volume have different years. Also, some older articles used spellings that are now archaic; don't panic if you see thun  in an old German quotation where you'd expect to see tun , or seyn  for sein , for example. Please try to check the original source when you report errors.

OK, ready to download the whole monster bibliography? Here it is.

If you find this material useful, I'd appreciate an acknowledgement in any publication that makes use of it.


© 1999 – 2001, 2003 – 2008, 2010, 2014, 2016, 2019 – 2023 Andrew T. Young

Back to the GF home page, or the Alphabetical Index, or the Tables of Contents, or the Overview page