On zenith arc extension naming

The text below is a copy of a section from my halo nomenclature article published earlier this year (https://thehalovault.blogspot.com/2026/01/on-halo-naming.html) that deals with historical zenith arc extension observations and naming issues. I have it here as a post because I thought it might be useful companion for my upcoming post about a quite remarkable old display. I have quietly corrected a handful of typos. Two bigger changes are marked with asterisks and explained at the end.

*    *    *    *    *    *    *    *

And so on to the naming of the arc observed by H. F. A Kern in Loenen aan de Vecht, Netherlands, on the morning of 8 October 1895. In the original report in the Dutch weather amateur's Onweders Optische Verschijnselen circular (1896), it is opined that "Earlier observations in which this arc is mentioned are doubted, because in most cases the imagination of the observer clearly emerges from the report" (the original text shown in the illustration below). I know only two older observations, and the other one, dating a century and half prior to Kern's, doesn't, in the light of the current knowledge, quite fall under that remark, at least in comparison to Kern's. This is the display by a Frenchman of the name Dufay in Paris, on 6 March 1735.

                                                             H. F. A. Kern's observation in Onweders.

All photographed high cloud zenith arc extensions – to date 6 displays – have been in big displays, and by that standard, H. F. A. Kern's display comes out oddly light weight. Ignore his arc and what's left is just the zenith arc and 22° halo. Elsewhere in the country nothing of consequence was seen either. The best was at Gravenhage, where additionally parhelia and 22° tangent arc was reported.

In contrast, Dufay's Paris display is already a multihalo. It's five halos against Kern's two. The little details – the slight asymmetry of zenith arc and its extension combo, the zenith arc not being in minimum deviation position, and the one-sidedness of the 46° halo * – add to the observation's street cred. It also contained parhelia, which is expected in a zenith arc extension display. It is also interesting to note that in none of the high-cloud-six (Jutland, Kissimmee and Reading x 4) the extension appears as a separate segment on the opposite side. The extension is faint in many of these displays, but the general picture is rather a one of continuations of the zenith arc than a separate arc – just like is the case with Dufay. (This in itself is admittedly strange as it suggests highly triangular plates and a persuasive line of evidence exists for not supporting high-triangulars taking stable plate orientations).

Two displays containing an extended circumzenith arc predating H. F. A. Kern. On the left by Dufay, on the right by J. W. Lambert on 24 January 1838 in Wetzlar, Germany. In line with his drawing, Dufay writes the zenith arc encompassed more than half a circle. Curiously, Lambert reported the whole circle AB as having been white. A particular detail in Dufay's drawing is the zenith arc grazing the 46° halo: "Le rouge de l'arc tangent touchait le bleu du halo extraordinaire". Bravais regarded this with significance as it backed up his theory of circumzenith arc against the other theory, the top 46° contact arc, which would always be fully embedded in 46° halo. In his book, Bravais comments on this observation on page 103. H. Ekema mentions Lambert's display in his Maandblad voor Natuurwetenschappen article, writing: "In the rich collection of observations included in his treatise [Bravais' book] there is only one case that somewhat resembles the phenomenon at Loenen, namely on 24 January 1838 at Wetzlar Lambert observed a white circle around the zenith at a solar elevation of 20° 15′. Whether both observations truly relate to the same phenomenon is, however, doubtful."

In Onweders, Kern's display is given an explanation by a raypath in plate oriented crystals that we notate as 136, which, as the text goes, makes a fainter copy of the circumzenith arc opposite to the zenith (in Optische Verschjinselen aan de Hemel (1957) Visser refers to the H. F. A. Kern's observation, and has the arc alternatively called as "de gespiegelde circumzenithale boog"), covering the same azimuthal span as zenith arc, about 120 degrees at the display's sun elevation of 8.75 degrees. The theory seems to originate from one H. Ekema, who presented his calculation in a parallel article in Maandblad voor Natuurwetenschappen, 1896, No. 5 (it has recently come available on Google Books). Aside from the calculation, it repeats much of what is said in Onweders.

Today we have the benefit of simulations to know that while 136 makes an arc weighted at 180 degree azimuth from the sun, it is also exceedingly faint. The raypath that we are concerned with when looking at any single scattering display photograph of zenith arc extension, is 135. It is by far the dominating raypath in any realistically shaped plate oriented crystals and makes an arc that leaves a gap opposite to the zenith.

To be true, the raypath 136 for zenith arc extension is strengthened, to a varying degree depending on the crystal shape and light elevation, by similar raypaths 13457 and 13537. But the combo of these three raypaths still comes weak next to 135. Unlike 136, the two longer raypaths may not be perfectly opposite-weighted because they can have, depending on the crystal shape, an intensity drop at the opposite position and even an outright gap.

Plate zenith arc extension in three different crystal scenarios at +8.75° of the Loenen aan de Vecht display. All crystals h/d 0.4 dev 0.1. Regular hexagons and full triangles are without basal face shape variation. The intermediate crystal shape of the middle row has variation, shown is the average shape. 13537 in the middle row is not empty, just a meagre few dots. Zenith arc raypaths and raypaths similar to it (such as 13573) are not included.

But having the raypath wrong is not the issue here. It is simply that Dufay has an earlier observation that is more competent than Kern's.

And I can't shake off the feeling of Kern's observation being purposely faked. The observation is just too convenient. As if the theory of the gespiegelde circumzenithale boog came first and then an observation was concocted around it. But if it was fabricated, surely they would have taken care to include the parhelia, right? In his article, H. Ekema understands that the crystals required for the zenith arc and its extension also make parhelia, but manages to put a positive spin on the lacking of parhelia in Kern's observation by brushing the issue under the carpet and instead noting that parhelia were indeed reported in other localities that day. So maybe the Kern's arc was a real feature in the sky – but not a halo but a funny cloud.

I guess there is a difference on how to regard, in a historical drawing, an extension that is continuous from the zenith arc and an extension that is separate. The former may have come about because of the typical phenomenon of drawing things longer than they actually were. Supposing the continuation was not real (which must be true in majority, if not all cases; in none of the six visual was gotten) such completionism may have been unintentional – it's harder to remember things when you jot down your impressions hours, days, weeks or months later. The latter is harder to reconcile with this psychology. Here, if it wasn't something that the observer actually saw (a halo or funny cloud), the other option is intent fabrication.

Examples of completionism. On the right a display by Schult on 27 March 1826 in Oslo. Wegener arc reaches the sun and circumscribed halo is interpreted as two full circles. On the left a display by F. Lebland ** on 1 February 1882 in Fort Conger, Canada. There are a number of cases of parhelia drawn at the junction of 46° halo and parhelic circle. Many are likely due to completionism, though other explanations may also apply, one being 46° infra- and supralateral arcs giving an impression of weak parhelia at their crossing with the parhelic circle. Here also 44° parhelia is a prospect because this was most likely an ice fog display.

Of course it is possible that at some point a name has been so long around that it's now indifferent to attempts of correcting wrongs. So the Kern arc may be here to stay. The man goes by the full name of Hendrik Frederik Anton Kern, and he was 30 years old at the time of the observation. He trained as a teacher and settled in Loenen aan de Vecht where he became the headmaster of the primary school. He doesn't seem to have been that much of a halo man. I have copies of all the Onweders Optische Verschijnselen halo sections from 1896 to 1952 observations, but except for his 1895 observation, I don't recall having seen H. F. A. Kern's name in them. But he must have been specialized in other kinds of observations because he is listed as one of the observers in Onweders Optische Verschijnselen 1938 issue, which is the only digitized OOV currently available (not on a Dutch but Indonesian institution website, mind you). Possibly he was an onweders chaser. It would be worth perusing the full OVVs to see what kind of stuff he reported.

But should someone prefer to start using the Dufay arc, I see it as a legitimate name. Possibly his is the first zenith arc extension ever reported. Which is of course an additional merit, because, as I have already said above, naming arcs after silver or bronze medalists is bad practice (I am brushing here of course under the carpet the issue of common practice of naming halos after persons who have never observed them).

Dufay probably went by the full name of Charles-Francois de Cisternai du Fay – a scientist whose main bread was electricity. Or would it be necessary to consider a double name because Dufay starts his account by telling that he observed the display with Mr. Condamine (most likely Charles Marie de La Condamine)?

Actually, there is the option for keeping both Kern and Dufay. The Parry orientation born zenith arc extension is different from plate born extension in that it is not a full circle but an arc, with more than 180 degree circumference. And the Dutch theory, 136 in plate oriented crystals, draws an arc identical to the Parry orientation born arc – both are from raypaths in which the exit face is opposite to the reflection face. So if the single scattering plate born extension would be Dufay, Kern could be reserved for the special case of Parry born extension on the basis of the theory, even if the orientation is not the same (no other raypaths contribute to zenith arc extension in Parry crystals). But shouldn't the arc in that case be named after its theoretician – the Ekema arc?

For the multiple scattering version I see Ripley-Saugier and Bravais as person name candidates, the former for their observation in the Saskatoon display, the latter for his theoretical genius to envision the possibility of MS extension. This is found on page 99 in Bravais' book Memoire sur les halos. He discusses the azimuthal span of the zenith arc and writes: "However, it may happen that the theoretical amplitude is exceeded, and even that a complete circle is formed: 1) because each point of the arc gives rise to the formation of a parhelic circle, and because all these secondary parhelic circles, by superimposing themselves, can produce an appreciable light" ("Cependant il peut arriver que l'amplitude théorique soit dépassée, et même qu'il se forme un cercle complet: 1) parce que chaque point de l'arc donne lieu à la formation d'un cercle parhélique, et que tous ces cercles parhéliques secondaires, en se superposant, peuvent produire une lumière appréciable"). Bravais also mused on the secondary parhelic circle at 22° to explain observations like the one by J. W. Lambert shown in the illustration. It's a certain mammoth in the room that Bravais does not have a halo named after him. True, the zenith arc has been called 'Bravais arc' in some writings, but I don't hear it used today. Which I think is good because by tradition person names are for more special halos. There is also a third MS halo that could be named after Bravais, the 44° tangent arc, which he also wrote about in his book.

Marko Riikonen

* in the original this reads unintelligibly "one-sidedness of the extension"
** the original credits this to F. Lebland, which is a double mistake. First, it is F. Leblanc. Second, he was not on the expedition; he was a wood engraver who prepared the printing block based on a sketch that someone had made of the display. The display appears in "Three years of Arctic service : an account of the Lady Franklin Bay Expedition of 1881-84, and the attainment of the farthest north, Vol I", written by the expedition leader Adolphus Greely.