The appearance of the sky on Saturday, December 23d, will be historic. At least those who saw the heavens then from eight o'clock till two, will long relate the scene as one of the things most wonderful.
— The Boulder County News, 29 December 1876
On Saturday, 23 December 1876, a halo complex occurred over the Denver-Boulder area that challenges the Saskatoon display of 3 December 1970 as the holder of the first reliable observation of zenith arc extension and 44° parhelia. Two accounts, written for separate publications by Sergeant James A. Barwick of the U.S. Signal Service at the Denver station, provide a strong case for these halos having been in the sky, not just artefacts of the mind that had crept in after the fact.
An after-storm display in a hard arctic outbreak
The display was the product of a massive arctic outbreak, and on the preceding night the area was in the grip of a blizzard: "Denverites need not be told about Friday night's storm, but for the benefit of non-residents THE NEWS will state that for several hours it was one of the worst known in this city," writes the Rocky Mountain News in 27 December issue.
Come morning, the show was visible right from sunrise. After describing the storm, the RMN article continues: "Yesterday morning the sun rose clear but the air was filled with particles of frost, the refraction from which caused the appearance of 'mock suns', or 'sun dogs'." Another article, in the Colorado Banner 28 December issue: "The air is filled with frost, reminding one of a snow storm...". So this was, as expected, diamond dust.
Yet from Barwick's newspaper account in the Boulder County News (29 December issue – said to be copied from an original in the Denver Tribune, but the issue is not available online) one might infer instead a cirrostratus sheet. Barwick writes that on "Saturday morning the sky showed a dull milky appearance," and then puts the display ending at 2:45 in the afternoon because that was "when the clouds became too thick to permit the rays of the sun being reflected or refracted."
Alas, Barwick seems to have been influenced here, at least on the morning part. Elias Loomis's Treatise on Meteorology (1873) – which Barwick in the same article names as the "Signal Service text book on meteorology" and which he says "has been very largely drawn upon for the explanation, causes, etc., of the occurrence of the optical phenomena above described" — carries, in its section on 22° halo, a strikingly similar phrase: "When the sky is hazy, and presents a dull, milky appearance…"
As to the afternoon thickening, I could find nothing in Loomis that Barwick might have drawn upon. Interestingly, the snippet in the epigraph, which is about the action in the 40 km distant Boulder, contains vague parallels with Barwick's afternoon remarks: "At least those who saw the heavens then from eight o'clock till two..." Then again, the full text from which this comes from, continues to describe a blazing parhelion after the sun had already gone behind the hills:
As the sun fell behind the hills, the lower end of the upper and western quarter of this brightest ring was all that remained; but that point seemed in all its colors even more brilliant than before, as it stood on the face of the high front hill at Gregory Cañon like a prismatic sun doing its best in honor of the coming Christmas.
Even if the timing of sun falling behind the hills is dependent on the observers position relative to the hills, this feels like an observation further down in the afternoon, because around 2 pm the sun was still quite high.
But I would rather not get lost in the fog of conflicting accounts here, because the big picture is in any case clear: in
this kind of outbreak the high-cloud-bearing systems have already swept
through ahead of the front, and what is left behind is the cold dome
itself, with a clear upper atmosphere And this outbreak was one of the hardest the area has known, for the very next morning, Sunday
the 24th, Denver fell to its second-coldest temperature on record, −25
°F / −33 °C (https://www.weather.gov/bou/lowtempextremes).
Solar elevation change in Denver on 23 December 1876.
Now, while Barwick gives an impression of high cloud origin for the display, he also has language that is suggestive of diamond dust: "The air on Saturday being calm and still, these snow crystals descended slowly to the Earth." But this is said in the context of explanative talk on crystal orientations, so doesn't seem to be an observation of the actual conditions. And again, it looks like it's not really Barwick talking here, because in Loomis the section "Parhelic Circle" has similar and in part identical wording: "When the air is tranquil, the flakes of snow which are present in the atmosphere descend slowly to the earth…" More of Professor Loomis will be seen creeping into Barwick's account below.
The zenith arc extension
On to the actual business. The zenith arc extension already comes out compelling in Barwick's Monthly Weather Review account (Vol. 4, Issue 12, p. 8, December 1876): "That portion of the zenithal circle (SS) which lay nearest the sun, was a strikingly brilliant rain-bow, the red being nearest the sun, and all its glowing colors were very clearly defined." This tells us the zenith arc was not merely extended to a uniform full circle afterwards – the circle consisted of two distinct parts. The reading is further bolstered by a note at the end about a second display, seen on 10 December at Pembina, in Dakota Territory (now North Dakota), which is credited with only the rainbow segment of the circle.
But the most striking evidence is in Barwick's newspaper article:
When the sun had reached an altitude of 12° above the horizon, a brilliant arch in the form of an inverted rainbow, represented by the letters S S, was seen to form a tangent with the halo of 46° radius C C, at its highest point above the sun. As the sun rose higher in the heavens this arc became more curved, and between 9:40 and 11:50 a.m. it had formed itself into a complete circle of about 15° or 18° radius. One-quarter of the circle was of the most brilliant and glowing colors, showing the different colors of the rainbow, so very distinct that they could be traced one after the other, from red, nearest the sun, to violet, which was farthest from the sun, the centre of this arch was the centre of the zenith, observed from where the observer stood. The remainder of the circle was distinct enough, but did not show any of the prismatic colors, and was not so wide as the portion where the colors were so distinct.
I have had goosebumps many times over my halo career, but reading this moved me for the first time to tears. And yet – is Professor Loomis's shadow looming over Barwick here too? For it is not only the falling-crystals passage and the "dull milky appearance" that Barwick took from him. The mangled tangent arc in Barwick's drawing (NN, OO) echoes Loomis's own diagram, in which two elevation forms of the tangent arc float slightly detached from the 22° halo. And Barwick has the zenith arc appearing only once the sun reached 12° elevation – precisely the figure Loomis gives (erroneously) as the lowest possible elevation for the halo. Moreover, Loomis writes of reported parhelia at 50° and 98°, of which Barwick duly notes the latter as visible in Denver (HH, alongside the 120° parhelia LL). I could cite more similarities, but enough. So was I an old sentimental fool for those tears? Actually, I am starting to get paranoid. What if Barwick didn't even see the display and his two reports are simply pure Loomis channeled through him?
Not so fast. Loomis gives not a word nor a drawing of any zenith arc
extension. The only pre-Denver observations of the extension I know of
are those of Dufay's in Paris on 6 March 1735 and J. W. Lambert's in
Wetzlar, Germany, on 24 January 1838. To be in the know, Barwick would
in practice have had to be reading Bravais, yet he cites only Loomis as
his source.
And even if, for argument's sake, we grant that Sergeant Barwick had read the French of Bravais, or perhaps had parleyed with someone in the know (J. D. most likely, who wrote in RMN and BCN), we still run into the problem that neither of the two earlier observations resemble what he describes. Neither Dufay nor Lambert reports the dual feature – a bright, colourful zenith arc with a fainter white extension. Dufay's is a uniformly colored circle (though gapped opposite the sun); Lambert's is, oddly, a uniformly white one. Bravais does also carry a theoretical note on the possibility of a multiple-scattering extension, but I cannot see that transforming easily into Barwick's description either.
Turn instead to modern displays in which the extension has been reported visually, and we find two brethren to Barwick. Ripley and Saugier, in the Weather article "Photometeors at Saskatoon on 3 December 1970", describe a white extension in the display. In the second modern visual I know of – the display of 29 February 2016 at Klaukkala, Finland – Marko Pekkola writes of having seen Kern arc as a "faint, fragmented, suggestive glow," but gives no indication of colour perception (in the photos the arc is quite distinctly coloured, https://www.taivaanvahti.fi/observations/show/49209). And in the third case – the Saskatoon-class Siziwang Qi display of 14 February 2020 – of the two persons who reported seeing the extension, Tian Xiangyang and Zheng Dan, the former called it coloured while the latter recalled a "mostly whitish arc all around" (from personal communication with Jia Hao).
So while absolute certainty is probably beyond reach, I judge it unlikely that Barwick knew anything of the zenith arc extension. His describing the dual nature of SS was where Loomis momentarily relaxed the possession of him, letting him set down the extension, true as it stood in the sky.
44° parhelia
The parhelia at the crossing of the 46° halo and the parhelic circle are pictured and discussed in Loomis's book – most insightfully, as will be shown below – so they, too, could have found way into Barwick's account without having actually been in the sky. But the photographed displays tell us that in a supermassive diamond-dust show of this kind carrying the zenith arc extension, the 44° parhelia (which at higher sun elevations would rather appear as 46° parhelia) are as a rule part of the package. So long as we accept the zenith arc extension real, we can count in also Barwick's XX parhelia, despite him not giving any helpful details on the feature.
Until now, I have regarded Saskatoon display as the first reliable documentation of 44° parhelia. The halo was photographed, so there is no question whatsoever on the veracity. However, my skepticism towards the earlier would-bes is not for the lack photos – a non-photographed observation is fully acceptable as long as it meets the standards. And the pre-Saskatoon reports of parhelia on 46° halo just don't meet the standards for 44° parhelia. Most observations are in high clouds, which automatically renders them highly suspect for the reason that no high cloud 44° parhelia have been photographed. And the diamond dust ones neither survive a closer look for whatever reason. Below I have provided two historical diamond dust examples that fail to convince with their parhelia on 46°.
The drawing by J. R. Blake of one of the diamond dust displays in his book "Solar Halos in Antarctica" looks the part for 44° parhelia. But reading the account pancakes it. Blake describes both parhelia on concentric halos as very bright – an impossibility regarding the outer ones if one supposes they were 44° parhelia. As for the arc at the zenith, Blake was in the know of the Kern arc possibility, but is not saying a word about it in the display description. Clearly, in his mind this was something different. The location ("passing through or close to zenith") of course doesn't match, but neither do the other aspects ("fairly bright coloured arc... the colours being very distinct"). The date for the display is 29 November 1958.
How Denver zenith arc extension fits into the historical picture?
No camera was turned at the zenith arc extension in the Saskatoon display, was hinging only in the drawing and description by Ripley and Saugier, fuelling endless speculation among enthusiasts. At least in Finland, no two halo people ever met without asking themselves, in hushed voices: "Did Saskatoon have Kern?"
But after the first photographed extension materialized from the stack Marko Mikkilä made of the 17 November 2007 display at Sotkamo, Finland, the possibility of Saskatoon really having contained the extension began to gain foothold. Fresh observations piled on more weight and finally, a visual sighting backed by photographs in the Saskatoon clone at the Siziwang Qi area of northern China (https://thehalovault.blogspot.com/2020/02/intense-kern-arc-from-china.html) settled for good the question of whether a zenith arc extension can be intense enough for visual.
The Denver report by Barwick moves the first reliable detection of the extension almost a century back from Saskatoon. The jump actually pushes 19 years past H. F. A. Kern's "original" observation of 8 October 1895 at Loenen aan de Vecht in the Netherlands, making this highly doubtful (and possibly even faked) report lose even more its standing, and rendering the case for using a name other than "Kern arc" stronger.
The 44° parhelia ride along with the extension. This possibly pushes Denver past all diamond dust displays that have parhelia depicted on 46° halo.
The drawing of Saskatoon display from the Ripley and Saugier Weather article "Photometeors at Saskatoon on 3 December 1970". The similarities with Barwick's drawing go even as far as using a thinner line for the zenith arc extension. My blogpost "The Saskatoon display" has some photos: https://submoon.wordpress.com/2011/01/24/the-saskatoon-halo-display/
Miscellaneous notes
With the significance of the Barwick's Denver sighting laid out, a word on some other issues in and around the display.
Denver-Boulder and Saskatoon similarities
The parallels between the two displays are clear for all to see. Both were unusually long-lived, starting at sunrise and going well over midday. According to Ripley and Saugier, in Saskatoon the show kept going for five hours until 14:30. The Denver-Boulder display was possibly lasting all the way till the sunset.
The compositions match too. Set aside the misplaced tangent arc in the Denver drawing – possibly a non-existent Loomis import – and the two are identical but for the sun elevation in the drawings. Both even carry the dubious extra parhelia between 44° and 120° parhelia.
On the naming
In the world of halos, extensions (one halo continuing another) are everywhere, and I see "zenith arc extension" — or "circumzenith arc extension" if you prefer the longer form — always a legitimate neutral choice to run alongside the observer name (with optional specifications of single-scattering plate, single-scattering Parry and multiple-scattering domains). My nomenclature article has a section on the zenith arc extension naming, which I have for convenience prepared into a post here in The Halo Vault: https://thehalovault.blogspot.com/2026/07/on-zenith-arc-extension-naming.html The text sets out the Dufay's extended zenith arc, which is one and half century earlier than Kern's, as the more convincing case of the two, thus raising the possibility of calling the halo rather "Dufay arc" than "Kern arc". In my X posts I have experimentally done exactly that, hard as it has been, having basically grown up on "Kern." But that's just when you have to steel your jaw and push harder; comfort zone is no place to be when halo naming problems need fixing.
And now Barwick has arrived on the block, and one thing he certainly does is to make the "Kern arc" even more unearned. Where there was one more convincing observation of a date earlier to Kern's, now there are two, the "Kern arc" growing ever more uncomfortable choice by it. But the high confidence level of Barwick's extension makes the ground unsteady also under "Dufay arc".
About daytime cold in Denver-Boulder
Denver and Boulder residential areas sit between 1500 and 1700 meters at a relatively low latitude of 40th parallel, so sun tends to warm the daytime air appreciably. Even so, in winter the temps stay below freezing on number of days. In Denver this averages at 21 days, and the coldest daily maximum on record – 3 February 1883 – was as low as −10 °F / −23 °C. So when such a deep freeze sets in, it is not difficult to imagine diamond dust persisting until afternoon and even to sunset.
For comparison, the Siziwang Qi banner display on the Mongolian Plateau seems to have lasted all day. I don't remember where exactly within that area it was photographed, but the whole administrative region, lying between the 41st and 43rd parallels, ranges from 1,000 to 2,100 m in elevation, comparable to Denver-Boulder. Enthusiasts have for some while understood the high plains as the home for the most supermassive diamond dusts, and these days photographs of several displays a winter seem to come out of the Mongolian Plateau.
Two picks from Loomis
Not only was the Denver-Boulder display revelation to me, but also Loomis's book was a fresh acquaintance that delivered the goodies. It's with those that I finish this text.
In the section for the 46° parhelion, Loomis argues that crystals with vertically oriented prism faces (the alternate Parry orientation) would be unstable, and concludes that such a halo is better explained as parhelion of parhelion:
These parhelia are probably produced by rays which have experienced the minimum deviation in the same direction in two vertical prisms, in which case the total deviation of the rays would be double of that produced by a single prism. Upon this hypothesis the parhelia should not exactly coincide with the 46° halo, but for elevations not exceeding 30° the difference might easily escape observation. The observations are not sufficiently precise to decide whether this explanation is admissible or not.
Is Loomis the first to recognise the mechanism in plate-oriented crystals? Bravais conceived of multi-scattering through 60° prism in two crystals of the same orientation, but strangely considered only column and random orientations (far as what I have gathered from the book).
The second point of interest is Loomis's attempt to explain the Hevel's halo. He invokes the triple-alternate raypath 3-5-7 in triangular crystals, saying it produces "an illuminated surface of 92° radius, and having a tinge of violet on the side next to the sun." Today the feature we mainly associate this raypath with is the relatively common blue edge or blue arc at 87° from column oriented crystals, the equivalent and similar looking random-orientation feature being a much rarer catch. Is Loomis in fact touching here upon the blue-edge mechanism?
Loomis's text on the triple-alt raypath in triangular crystals to explain the Hevel's halo. He seem to imply in the first sentence, however, that the idea is not his. Did Bravais consider this?
Marko Riikonen
No comments:
Post a Comment