By Duncan Steel 25/06/2019

[avatar user=”duncansteel” size=”thumbnail” align=”right” /]

We have just passed the solstice, the shortest day of the year in the southern hemisphere. From now on, the hours of daylight will get longer through until the December solstice. Here I discuss why the June solstice occurs a few days before the Feast Day of St John the Baptist, the traditional time of ‘midsummer’ in the northern hemisphere (and the reason for the festivities depicted by Shakespeare in the comedy whose title includes that word). There is also a serious side to this: few people (including many climatologists) seem to realise that the durations of the seasons are varying, and this might be a significant factor to encompass in climate change studies; quite apart from that, analysis of historical climate records should take into account the fact that the seasonal cycle was, until about 70/80 years ago, apparently the anomalistic year rather than the mean tropical year as now seems to be the case. 

Some might argue that I should have written this post last week, but I have rather deliberately left it until today (June 25th) to commit these thoughts to the internet.

As most will be aware, the shortest day of the year (in terms of the time between sunrise and sunset) in New Zealand occurred last Saturday, June 22nd. On Universal Time (UTC) – essentially the time on the Greenwich meridian – the event was on Friday, because the time that the Sun reached its northernmost declination was 15:54 UTC on June 21st.

I am by no means sure why, but the June solstice nowadays appears to be the time at which many Matariki (essentially, Māori New Year) celebrations are held, despite the fact that Matariki (the te reo Māori word for the star cluster known variously as the Pleiades or Messier 45 [astronomically], Subaru [in Japanese], or the Seven Sisters in common parlance) undergoes its heliacal rising in late May or early June, though different groups have chosen to mark the event at later dates.

The Pleiades or Matariki, courtesy Brian Goff of Petaluma, California.

The English word ‘solstice’ is derived from the concatenation of two Latin words, though which two words is unclear (to me, at least, who was never higher than 30th in a Latin class of 33 pupils). The beginning is obvious: ‘Sol’ means ‘Sun’, as in solar. Sources I have consulted over the years, however, have variously ascribed the rest of ‘solstice’ as coming from one of the Latin verbs stare (to stand still, rest, remain) or sistere (to cause to stand, stop, or check [something’s progress]).

Those who have studied Latin will know that the first verb one learns is generally amo, amare, amavi, amatus, driven into memory by frequent repetition, the whole class chanting it. Amare (i.e. the infinitive) means ‘to love’, from which we get amorous, amateur (someone who does things for love, not money, though its accepted meaning has regrettably altered to become a pejorative description of someone), or the French mon amour and the Italian amore. Amare is the archetypal first conjunction verb, and its bedfellow is sto, stare, steti, status, from which we have received stand, status, station, statue and stature.

On the other hand sistere is a 3rd conjunction verb: sisto, sistere, stiti, status. The grounds for confusion is apparent. English words stemming from this one include resist, persist, desist, subsist, insist, consist, and even exist (the ‘s’ having been dropped from ‘ex-sist’). I would assume that interstitial also stems from that source, but maybe it comes from stare (which, I think, is pronounced like star-ray rather than the English word for looking intently at something or someone).

Where have we gotten to (I write, with poor English construction)? My preference is to think of ‘solstice’ as deriving from sol + stetit, literally ‘the Sun has stood still’ as stetit is the third-person singular perfect of stare.

Anyhow, the implied meaning of ‘solstice’ is that the Sun appears to stand still. That is, its rising point in the east sweeps along the horizon moving north between the December solstice (see what I did there, to avoid worrying about which hemisphere one is in?) and at the June solstice, when it reaches its furthest northerly rising point; it then starts moving southward again, reaching its most-southerly rising point at the December solstice. At the solstices the Sun reaches its extreme declinations (furthest angular separations from the celestial equator), those declination angles being equal to the tilt of Earth’s spin axis (technically known as the obliquity of the ecliptic). Those extreme declinations also equal the latitudes of the tropics, these being the furthest parallels at which the Sun can ever appear directly overhead, this occurring at the solstices.

So much for the etymology of the word ‘solstice’. Surely ‘equinox’ is simpler? Well, it is simpler, but it is misleading. Obviously enough, the noun ‘equinox’ means ‘equal night’ (Latin: aequinoctium), implying that the durations of daytime and nighttime are equal. Problem is, those are not equal at the equinoxes.

Astronomically-speaking, the equinoxes occur when the Sun in the sky crosses the celestial equator, moving north at the March equinox (the start of spring in the northern hemisphere, but autumn here down south), and swinging in the opposite direction at the September equinox. If the Earth had no atmosphere, there would be equal lengths of daytime and nighttime. But we do have an atmosphere.

It happens that the atmosphere deviates (refracts) the image of the Sun as it is rising or setting by almost exactly the same angle as its apparent size in the sky: half a degree. In consequence, as you watch the Sun setting and its lower limb appears to touch the horizon, in fact on an atmosphere-less planet it would already have disappeared, the upper limb just dipping below the horizon. The same applies at sunrise. This means that the Sun appears to be above the horizon for several minutes more than 12 hours at the equinoxes, so ‘equinox’ is a misnomer. And I haven’t even mentioned whether we should consider also the effect of atmospheric scattering of sunlight and so the lengths of dawn and dusk (or twilight).

Properly, the seasons start and end with the solstices and equinoxes, despite many  people imagining that March 1st, June 1st, September 1st and December 1st mark the beginnings of the seasons. In fact, the seasonal change occurs about three weeks later: in the northern hemisphere spring begins with the Spring Equinox around March 20th, summer with the Summer Solstice near June 21st, autumn/fall with the Autumnal Equinox on about September 23rd, and winter with the Winter Solstice around December 22nd. In the southern hemisphere, of course, the seasons are reversed.

In a previous post I described why the solstices and equinoxes are displaced from their traditional dates (e.g. Christmas Day is on December 25th, and the solstice near the 22nd, about three days earlier: this is due to the Catholic Church, in the Gregorian reform of the calendar in AD 1582, choosing to use the dates of the solstices/equinoxes in AD 325, the year of the Council of Nicaea, rather in the era of Jesus Christ three centuries earlier). Regardless, the instants of the solstices/equinoxes range over about 53 hours across the 400-year leap-year cycle.

Whilst we are in the southern hemisphere in New Zealand, so that the recent solstice marks the beginning of winter, please might I request you to think of yourself instead being in the northern hemisphere, so that summer has just begun with the June solstice.

The way that people thought of the seasons in the past is a bit different to the present, when society was largely agrarian. The summer was often considered to begin with May Day, at the start of that month, and the solstice seven weeks or so later was thought of as being midsummer. 

In a religion-dominated society, this was an important event. Christmas at midwinter (taken to be December 25th) was one important festival, marking the assumed birthdate of Jesus Christ, and midsummer (traditionally on June 24th, though the solstice has been shifted a few days earlier by the Gregorian reform) was celebrated as the nativity of St John. Take a gander at Shakespeare’s A Midsummer Night’s Dream and you will find it is set on St John’s Eve (i.e. the evening of the 23rd).

But which Saint John? Well, June 24th is the feast day of St John the Baptist, but there has been a profusion of saints named John. The biblical Gospel According to John is ascribed to John the Evangelist, otherwise John the Apostle, who lived decades after the Crucifixion (and his feast date is December 27th). After a quick search I have found at least eighteen saints named John (one of whom shares June 24th as a dedication day with the Baptizer) and that number does not include the Jeans and Juans and so on.

John the Baptist was a contemporary of Jesus, and has appeared in several movies due in part to his unfortunate end in about AD 28/29: he was beheaded at the request of Salome, the step-daughter of Herod Antipas, or so the story goes.

Caravaggio’s 1610 depiction of the head of John the Baptist being presented to Salome.

I knew already that it was John the Baptist who is commemorated on June 24th, despite having no religious beliefs myself (all I am recounting here is history, and astronomy). The reason I knew is that I was born in a small town in Somerset which is named Midsomer Norton, though some centuries ago the spelling was ‘Midsummer’. The small river running through the town centre is called the Somer (sew-mur) but I think that too is a later alteration in spelling.

The reason for the original spelling being ‘Midsummer’ is surely by association with the local church (of England), which is dedicated to St John the Baptist. As a matter of fact, my primary school was named for this St John, and I remember well being dragged off to the church down the road to sing in the choir and so on. (Coincidentally, my elder son Harry is just completing his DPhil at the Oxford college also named for John the Baptist.)

There is, it happens, another Midsomer Norton in the solar system. You can find it here, or by using a suitably-large telescope.

The title I gave to this rather discursive essay was The day the Sun stood still. That echoes two quite different matters.

The first is that there is a biblical account of the Sun standing still. The following text comes from Joshua 10:13.

          And the sun stood still, and the moon stopped,
          until the nation took vengeance on their enemies.

If you search on the internet you will find many discussions of what this means, and believers arguing that this statement is literally true. As a scientist, of course I prefer explanations that conform to the laws of physics. A couple of years ago Colin Humphreys of the University of Cambridge and my old friend Graeme Waddington of the University of Oxford published an analysis in which they made a strong case for the celestial event described by Joshua as being a total solar eclipse in 1207 BCE. This report is also of significance in that it enables a better dating of the pharoanic reigns in that era.

My title also resonates with a movie from long ago, which should be known to aficionados of science fiction: The Day the Earth Stood Still. As an old-timer I prefer the 1951 version, although younger readers will likely be more familiar with the remake that appeared in 2008.













Anyone reading this far might be thinking: where’s the science? In my discussions above I may mostly have been flitting around, discussing this and that point I find of interest. But I wanted to finish off with a scientific matter that actually informed my original decision to write about the solstice, and its date.

So, nowadays the soltices and equinoxes occur on calendar dates that are generally about three days earlier than their traditional dates. Christmas Day is traditionally December 25th, but the solstice occurs around the 22nd. The Nativity of St John the Baptist is traditionally June 24th, but the solstice in that month occurs on about June 21st. One can take a look at the actual instants of the solstices and equinoxes and see how they shift around; I have previously given in some detail the range of March equinox dates in a graphical representation stretching over more than five centuries.

But here is a question to ponder. Do the solstices and equinoxes stay in step? For example, if ‘spring’ in the southern hemisphere lasts from the September equinox until the December solstice (as it does), is it of constant duration, or does it vary across decades or centuries?

The answer is that it varies in length. In the plot below I show the durations of the northern (sorry) seasons across six millennia, from 3001 BCE until 3000 CE (i.e. almost a thousand years into the future). As is obvious, the seasons vary in length.

Durations of the seasons for the northern hemisphere between -3000 and +3000 (3001 BCE to 3000 CE, though this dating span is imprecise because the ‘year’ used here is the mean Julian year of 365.25 days rather than the mean Gregorian year).

From the perspective here of how the actual solstice/equinox occurrences relate to the traditional dates, the plot above might be regarded as being informative. For a subset of the four solstices/equinoxes the trend is to get closer to the traditional dates; for others the trend is to get further away. I will leave it as an exercise to the reader to work out which is which.

But why are the seasons varying in length? The answer is that perihelion is precessing, slowly moving through the calendar year from its present occurrence close to January 4th (and it steps forward by one day every 57/58 years). This puts it within (northern) winter, which is therefore the shortest season; and correspondingly summer is the briefest season in the southern hemisphere, as it has been since about 1250 CE. (To understand why, consider that the Earth’s speed varies with its distance from the Sun, as per Kepler’s second law.)

The sum of the four seasonal durations in the above plot renders the mean tropical year. This is very slowly reducing (as measured in mean solar days) in the present epoch, but its rate of change is small compared to the ways in which the distinct tropical years are varying. That is, one can define a type of year for any desired start-and-end point on the ecliptic (i.e. the tropical zodiac), and the four obvious points to use are the solstices and the equinoxes. This leads to a graph as below:

Durations of the different tropical years as counted between the specific solstices or equinoxes. Currently the mean tropical year – the average of the four others – is slowly decreasing. One should remember, though, that whilst the above calculations give year lengths in terms of the mean solar day at present, in fact the mean solar day length itself is increasing in time, due to tidal drag imposed largely by the Moon (hence the imposition of occasional leap-seconds).

Quite distinct from all of the above is the fact that, over the history of temperature records since the invention of the thermometer, the dominant annual cycle of temperature (which one might regard as being the seasonal cycle) has generally been the anomalistic year (the interval between perihelion passages, currently near 365.2596 mean solar days) rather than the mean tropical year, as seems to be assumed in most climate change publications.

This I find puzzling. I would welcome anyone being able to direct me towards climate analysis publications in which:
(a) The anomalistic year rather than the mean tropical year is used as the ‘year’ for phasing of climate records; or
(b) The varying durations of the seasons are explicitly encompassed by the analysis.



The photograph at the head of this blog post shows Stonehenge soon after sunrise. The fact that this megalithic structure appears to be oriented such that sunrise at the June solstice appears close to its axis of symmetry is well-known. I am pleased to acknowledge that the source of this photo is a lady named Nik who has made many of her wonderful images freely available on the Unsplash website


Addendum (June 29th): Midsummer? Midsommar in Swedish, the title of a new movie.