Prehistoric Solar Calendar

Seasons of the year may be measured by the rising and setting positions of the sun on the horizon. With apologies to those living in the southern hemisphere, the two extreme positions are: To the South, Winter Solstice (shortest day / longest night) and, to the North, Summer Solstice (longest day / shortest night). Mid-way between them are Spring Equinox and Autumn Equinox when day and night are equal.

These basic divisions split the year into four parts. Halving each of these parts gives us the four days known as Cross-Quarter days. Imbolc, Bealtaine, Lughnasadh & Samhain / Halloween (Candlemas, May Day, Lammas & Martinmas). These are commonly held to have come to us from the iron-age celtic calendar but they really are very much older.

A prehistoric sixteen part division of the year was discovered by Alexander Thom in the 1960's. Whole Horizon Analysis confirms it. Pre-iron age peoples of the Atlantic Coasts of North-West Europe really did split the year this way:

Solar rising and setting positions move along the horizon from south to north and back again over the course of a year. Like a pendulum, rate of change is greatest in the centre and smallest at the ends. Average daily shift of about one solar diameter. Slightly more around the equinox. A little less by the cross-quarters, much less around the solstices. Prehistoric peoples watched it all very carefully and used the horizon profile as a measuring tool.

This diagram of the pattern with equinox, cross-quarters and solstices labeled shows how just nine points on the horizon split the year into sixteen equal time periods. Moving towards the equator would make the pattern more compact and upright. Going the other way would expand and flatten it still more.

Both horizon height and latitude have a significant effect on the calendrical meaning of any particular bearing. The uncertainties of time have another one, with the overall azimuth spread now less than it was. Further back in time than 1800BC it was a bit more extreme but only a bit. Obliquity and Declination values peaked during the European Neolithic so that things were stable for quite a long time then. As a rule of thumb: For Summer Solstice sunrise, the sun first becomes visible now at roughly the same place that it lifted off the horizon in prehistoric times. Winter Solstice sunrise is the opposite. [Diagram]. Sets mirror the rises.

Beyond this repeated halving into sixteen parts of about 23 days each, Professor Thom thought he might have found some evidence of further subdivisions of the solar year. This would make sense. If you intend to have a socio-religeous event on a particular day, then simply marking that exact day is of little practical use. Some early warning of its approach would be required.

Thom wondered about 32ndths. I claim to have discovered markers for days that are about one and two weeks away from the major solar dates. Looking at these, I realised that they could have been splitting the nominal 23 day periods into thirds. As an attempt to rationalise and generalise these divisions without excessively complicated terminology, they were labeled as 8 & 15 day brackets to the major solar events: solstices, cross-quarters and equinox. Together with the original sixteen part division, these 8 & 15 day brackets make a year of forty-eight (variable-length, 7.6 day mean) "Tweeks" that are better regarded as quarter-months.

The next diagram is a bit crude because we are really dealing with a continuous sequence, not an alternating one but it demonstrates the over­lapping months (that are really eclipse seaons):

The 15 day brackets to any major solar event effectively define the period within which the closest full moon to that event must occur. The 8 day brackets do the same for the sixteenths.

Evidence for these calendrical divisions is found primarily in their coincidence with prominent horizon features and reinforced by examples of monument axes that indicate them. Such examples, for each class of division if not each single division, may be found here and statistical analysis of many prehistoric ritual horizons may be found here.

The values that have been used in this work are:

• Declination is a measure of the position of an object in the sky, relative to the celestial equator.
• Given Declinations are for the sun's centre, those for the upper or lower limb would be further north or south by approx. 0.26 degree.
• Spring Equinox, Summer Solstice, Autumn Equinox and Winter Solstice are the four quarter days or "natural solar festivals".
• Cross-Quarter days are halfway in time between the Solstices and the Equinoxes, and with them make eight "major solar festivals".
• Sixteenths are the midpoints between adjacent "major festivals" (nominally 23 days from each, 22.8 days mean).
• The "8 day" and "15 day" periods are nominal compromises, not necessarily of the stated length which split a nominal 23 day periods into thirds.
• + & - refer to more northerly or southerly declination than the adjacent "major festival".

This theoretical model was constructed to rationalise and explain the repeated occurrance of certain particular declination values in early survey data. It seems to do this quite well but I do not claim certainty, just a first stab at an explanation. Variations of a day or two here and there would not be unexpected in practice, because the values given are means. The basic sixteen fold division was originally discovered by Alexander Thom (1967) and some of his declination values were slightly different to these. My own version was derived from an 1800BC ephemeris by day counting and calculation of mean values. Prehistoric peoples would perhaps have used rule-of-thumb or intuitive methodologies.

• With all divisions taken into account, the tropical year may be seen as being divided into twelve months, approximations to synodic months, which are effectively a mapping of the zodiac onto the horizon. Note that this does NOT mean they used the zodiac as it is known to us, nor that they actually counted a twelve month year
  
  
• The half-month markers either side of the major solar festivals make eight "months" with a festival at the centre of each.
  
  
• The quarter-month markers either side of the major solar festivals mark the outer limits of eight "months" having a sixteenth division at the centre of each.
  
  
• These "months" overlap each other by a week at each end - marked grey on the diagram.
• Thom's sixteenths may therefore be seen as the solar centres of sixteen overlapping lunar periods. Prehistoric peoples also seem to have had a similar Prehistoric Lunar Calendar. Using rule-of-thumb or intuitive measurements they could have combined it with this solar calendar to make a fairly accurate Prehistoric Eclipse Prediction System.