Ninety-one, ninety-two percent it’s going to be at peak, so it’s pretty significant. There’s a debate on whether we’ll see any dimming of the daylight, but perhaps we will, perhaps we won’t.

Eric Walker of the Highlands Astronomical Society, interviewed on BBC Reporting Scotland Evening News, 10 August 2026

As this post goes live, on 12 August 2026, here in Scotland we’re just a few hours away from the event depicted and described above—a fairly dramatic partial solar eclipse.

Across most of Europe, West Africa, Arctic Russia, Canada and New England the moon will fall into an imperfect alignment with the sun, partially obscuring its disc. In a narrow band at the core of that broad area, the lunar-solar alignment will be perfect, and a total solar eclipse will occur. That’ll be visible only in limited parts of Siberia, the Arctic Ocean, East Greenland, west Iceland, the North Atlantic and Spain. (For more on the geometry of solar eclipses, see my post about Annular Solar Eclipses.)

But I wanted to pick up on something you may have found quite striking in Eric Walker’s short interview above—despite more than 90% of the solar disc being obscured, he said it’s debatable whether we’ll notice any diminution of daylight. How can that be? Your solar panels will certainly notice a 90% fall in illumination.

But the human eye is extremely adaptable to varying lighting conditions—there’s about a hundred-fold difference between bright sunlight and good indoor illumination, for instance, but we only ever notice the difference when we make an abrupt transition from one to the other. And the change in illumination during a partial eclipse is usually anything but abrupt—it can take an hour or more for the moon to crawl in from the edge of the solar disc to a central position.

The illuminance we receive from direct sunlight is about 100,000 lux. The illuminance under an overcast sky can be anything from 30,000 lux (a light cloud layer) down to 1,000 lux (a heavy overcast). A 90% partial eclipse will reduce the illuminance until it bottoms out at a pretty bright 10,000 lux, well within the range our eyes and brains have evolved to compensate for. But for such extreme partial eclipses, there is also an issue of rate of change. While the moon must move a long way across the solar disc to reduce its visible area by half, it doesn’t need to move so far to reduce that visible area by half again, and even less far to halve the area of the now quite slim visible crescent. And so on. So when a partial eclipse gets close to being total, there’s a period of relatively rapid dimming (and then brightening) around the time of maximum eclipse. And that might be noticeable to observers who are looking out for it, when an eclipse covers more than 90% of the sun’s disc. So, yes, there’s room for debate about what we’ll actually be able to see.

Which brings me to the Battle of Isandlwana*, which took place during the Anglo-Zulu War, on 22 January 1879. It was a massive defeat and humiliation for the British, during which 20,000 Zulu warriors surprised, overwhelmed, and almost entirely wiped out 1,800 British soldiers.

A solar eclipse occurred during the battle and, even by the standards of internet discussion, a truly colossal amount of garbage has been written about the eclipse’s supposed effect on the battlefield. A brief search of the internet will turn up stories of the battle being fought under “surreal light” or “apocalyptic gloom”, or even in “eerie darkness”. The onset of this darkness is often portrayed as being sudden. And I’ve seen it suggested that this “gloom” made it difficult for the British to select targets for their Martini-Henry rifles as the Zulus rushed towards them. There are long social media arguments about this, during which nobody seems to realize that it’s perfectly possible to find out the exact circumstances of the eclipse—you just ask NASA. Which is what I did.

NASA’s eclipse data for the date show that there was an annular solar eclipse, with a track much farther north than Isandlwana—so only a moderate partial eclipse occurred over the battlefield. For more information, I zoomed in on their map until I could identify the car park at the Isandlwana battlefield site, and requested data for that location. Here’s what I got:

Only about half the sun was obscured. Here’s a simulation of what it looked like on the day:

With such a fat crescent, there’s no issue here of an episode of rapid dimming and brightening around the time of maximum eclipse. Given that the change in lighting happened slowly over a period of more than an hour, and that the fall in illumination could be completely compensated for by a millimetric dilation of everyone’s pupils, it’s difficult to see how it would even have been noticeable, let alone “surreal”, “apocalyptic” or “eerie”. And, indeed, the eclipse seems to have gone largely unremarked in the region at the time.

NASA gives its precise eclipse timings in Universal Time, which corresponds to Greenwich Mean Time. In 1879 there was no standard time zone operating in South Africa, so Local Mean Time (based on the position of the sun) was in use. Isandlwana’s longitude of 30.652°E gives us a Local Mean Time two hours and a couple of minutes ahead of Greenwich, so it’s easy enough to convert NASA’s timings. Here’s the progress of the eclipse according to Isandlwana LMT:

But no-one’s watches seem to have been synchronized on the day, and there are some fairly gross disagreements about the timing of some events during and after the battle. And the British forces might well have carried the local time with them from Transvaal when they moved eastwards to invade Zululand. So if we’re trying to compare eclipse events to the imprecise local timings we have from the battlefield, the best we can say is that the eclipse started around 13:00 hours, reached its modest maximum at about 14:30, and ended around 16:00.

Timings for the battle are sparse and disputed. For reference I’m using Keith Smith’s essay “Isandlwana: A Timetable” from the Journal of the Anglo Zulu War Historical Society.

If you care to look at Smith’s essay, you’ll see that a lot of madly complicated things were going on as the Zulus moved towards the British encampment. But at about 13:00 the Zulus of the central group became pinned down by “withering fire” from an infantry line deployed to protect the camp. This lasted about half an hour, while two Zulu “horns” swept around the infantry line and invaded the camp at 13:30, by which time the infantry had begun retreating towards the camp. There was then chaotic fighting among the tents, and by 14:00 British survivors were fleeing towards the Buffalo River.

So the volley fire directed from the infantry line towards the approaching Zulus happened just as the first tiny rim of the sun was disappearing, and the chaotic fighting in the camp ensued when the eclipse was not much more advanced. The maximum eclipse was only reached during various last-stand actions among the British survivors.

---

And yet there are some battlefield reports that describe an episode of darkness. In his Companion to the Anglo-Zulu War (2008), Ian Knight quotes one Zulu warrior as saying:

The sun turned black in the middle of the battle; we could still see it over us, or we should have thought we had been fighting till evening. Then we got into the camp and there was a great deal of smoke and firing. Afterwards the sun came out bright again.

And another says:

The tumult and the firing was wonderful, every warrior shouted “Usutu!” as he killed anyone, and the sun got very dark like night with the smoke.

But notice that the first report comes from a time before the warrior arrived in the camp—so very early in the eclipse. And both accounts emphasize the presence of smoke. The Martini-Henry rifle used black powder and produced a lot of smoke, particularly during volley firing. There are reported occasions during the Zulu War when British rifle smoke grew so thick that firing needed to be halted to allow visibility to improve. And the camp itself is another potential source of smoke—when a relief force arrived (too late) they found many of the tents burned.

In her History of the Zulu War and its Origin (1880) Frances Colenso provided another witness statement that’s sometimes invoked as evidence of “eclipse darkening”. An officer in the relief force approaching the camp is quoted as saying:

There certainly were some tents standing then, but seemed very few, and away to the left front of the camp there was some smoke, though not much, and it was high up, just as if there had been musketry fire and the smoke had floated away; but there was certainly no musketry fire going on then. A few seconds afterwards a sergeant . . . . said: ‘There go the guns, sir.’ I could see the smoke, but we could hear nothing. In a few seconds we distinctly saw the guns fired again, one after the other, sharp. This was done several times—a pause, and then a flash—flash! The sun was shining on the camp at the time, and then the camp looked dark, just as if a shadow was passing over it. The guns did not fire after that, and in a few minutes all the tents had disappeared.

But eclipses don’t produce localized, transient areas of darkness the size of a military camp. Smoke from gunfire and burning tents might.

In an article entitled “The Sun Turned Black”, for the Journal of the Anglo Zulu War Historical Society, Ian Knight cites the recollection of a trooper some distance from Isandlwana as supporting evidence for a noticeable eclipse darkening:

Trooper Fred Symons of the Natal Carbineers, out at Mangeni [Falls] with Lord Chelmsford, was almost alone in recalling a strange, still, oppressive gloom—all characteristics of a partial degree of black-out—and felt a ‘presentiment that something was going to happen either to us or those at the tents.’

One can’t help but feel that this speaks more of Symons’ state of mind than anything to do with an actual eclipse.

And we also have this, from Harry Struben’s memoir, Recollections of Adventures (1920):

A strange thing happened at this time. I was walking with Sir Theophilus Shepstone^† outside Utrecht^‡, discussing the situation, when suddenly it became dark, and for a moment neither of us realised that it was a total eclipse of the sun. When we did he said “Struben this may have a strange effect on the Zulus, who are superstitious”. Next day we heard of the disaster at Isandhlwana and he said to some old Kaffirs who came to hear news “Umtwan ami u fele, George ugwazile.” “My child is dead, George is stabbed,” and he was quite overcome.

There was, of course, no total eclipse of the sun anywhere in the world that day; nor, as I’ve described, is “sudden darkness” a feature of partial eclipses. (The eclipse took an hour and twenty-five minutes to reach a maximum of 58% in Utrecht.) Struben was writing thirty years after the event, and long after Shepstone’s death, and it has the feel of a tale that grew in the telling.

Do you need more reassurance that the eclipse had no effect on visibility during the battle? Here’s Pat Rundgren’s personal report, from the battlefield itself:

Much has been made of the eclipse of the sun that took place on the day of the battle, and the “surreal light” that it allegedly cast on the proceedings. It has become part of the superstitious folk-lore associated with this most inauspicious day for the British.
Yet I personally have witnessed three x 60% eclipses at Isandlwana in my 20 years on the battlefields, and I have to say that I wouldn’t have known that one was taking place without (a) being told so and (b) without dark glasses. On those three occasions it made no noticeable difference to the light.

---

The battlefield is now a poignant place, dotted with white cairns marking the places where British soldiers fell.

(It wasn’t until 1999 that a memorial was built to honour the Zulu dead.)

---

* In older texts, the name is given as Isandhlwana, but that central “h” is now deprecated. In Zulu, it’s properly spelled iSandlwana, “i” being a locative prefix, indicating “place of”. Sandlwana is the Zulu name for the reticulum, the second stomach of a cow. The reticulum has a complex honeycomb pattern to its interior, reminiscent of the eroded surface of the outcropping that towers over the battlefield in the picture above. (For much more on all that, see another of Ian Knight’s articles for the Journal of the Anglo Zulu War Historical Society.)
^† Shepstone had been responsible for the British annexation of the Transvaal in 1877. His son, George, died at Isandlwana .
^‡ Utrecht, in South Africa, was at that time the seat of the British adminstration of the Transvaal.