ABSTRACT:
     The purpose of this paper is the demonstrate by photos and calculations that it is impossible by observations of the rising/setting sun or a gnomon to determine the two annual solstice days.
     On the solstice day and the day before and after the sun appear to follow the same path, which in Indian mythology is described as that Surya, the sun, rests for 3 days before continuing his annual circle from solstice to solstice.
     Determination of solstice in relation to ancient megalithic structures in NE-Europe (Stonehenge, Newgrange, etc.) has been described in the literature on astro-archaeology by Alexander Thom in the 1970es and discussed in detail in later considerations on archaeoastronomy (see literature below)..

Solar observations at winter solstice 2004:
				In 2004 it was winter solstice the 21st of December at 19:44 (local time = UT + 7 h.). Photos of the setting sun was taken daily from the 12th of December to the 1st of January.      The 20th, 21st and 22nd the sun seemed to set the same place.      Two days before and after equinox a slight movement was observable: The setting sun was moving a little towards north (right).
18 December 2004, sunset		19 December 2004, sunset
20 December 2004, sunset		21 December 2004, sunset		22 December 2004, sunset
		23 December 2004, sunset		24 December 2004, sunset

Solar observations at summer solstice 2007:

     In 2007 it was winter solstice the 21st of June at 06:59 (local time = UT + 7 h.).
     One major obstacle doing solar observations around summer solstice is that it falls in the rainy season. In 2007 the monsoon came later than usual and for the first time in 10 years it was possible to mark the southernmost path of the shadow of the sun on the sundial, but it was not possible to take comparative photos of the setting sun over a two-week period or not even for the five days around solstice.

21 June 2007, sunset		29 June 2007, sunset

      Because sunset observations were (and mostly are) not possible during the monsoon, we have to rely on observations of the gnomon shadow:

The gnomon below consists of a vertical iron rod, which casts its shadow on a water-levelled terrace floor. The 3 pictures are taken at the major solar events: summer solstice, equinox and winter solstice.
Summer solstice: 21 June 2007 at 10:21		Equinox: 21 March 2007 at 12:21		Winter solstice: 22 December 2007 at 10:03

9 June 2007 at 8:05 (the stippled lines are added)		23 June 2007 at 08:24
Above: Left: The paths of the shadow of the sun differ from the 8th (red line) to the 9th (blue line). Right: Two weeks later, around summer solstice the 21st, the daily paths of the shadow appear to follow the same track. The 24th the path changed slightly to the north, and the observer knows that equinox was some days before.   Calculations: Above: Daily change of path of the rising sun 4 weeks before summer solstice 2007. The closer to the solstice day the smaller the daily difference. Below: Enlargement of the above paths of the rising sun from the 17th to the 22nd of June, summer solstice. The difference between the 21st and the 22nd is so small that the naked eye cannot discern between the 2 days, so plus/minus 1 days from the equinox day the naked human cannot tell the difference. Probably one more day should be added making the 'resting period' of the sun at least 3 days, probably 5 days.          The two above pictures were taken at winter solstice 2007 and together with the two graphics above gives an impression of the difficulties in determining solstice day by naked eye observation.

CONCLUSION:      After 10 years of experiments and observations of sunsets/sunrises the author regards it as impossible to determine the solstice days by naked eye observations no matter how long sight-lines are used.      Observations of the daily path of the gnomon shadow gives similar result: Not until 2 days after solstice the observer knows that the solstice was 2 days before.

A method to determine the solstice day in advance

     If for some reason it is/was required to know the solstice day in advance of time, then it is possible to predict the day based on observations of the rising/setting sun or a gnomon. The latter got the advantage of not being dependent of a cloudless sky at the horizon and is also the only option on locations where the horizon is obscured by mountains, vegetation or constructions.

     Background:
     The 12th of December the daily path of the gnomon shadow followed a previously marked line made the 1st of January - 20 days ahead. Dividing 20 by 2 gives 10, which added to 11 gives 22 = the solstice date.
     Calculations confirm the observation: At solar noon the apparent altitudes of the sun on the two above mentioned dates were respectively 50°17'11'' and 50°17'24''. The azimuth of the sunrises were 113°56'00'' and 113°56'40''. The azimuth of the sunsets were 246°01'44'' and 246°03'35''.
     The 12 December / 1 January line was marked on the sundial, so next year the gnomon shadow will announce the solstice day 10 days in advance.

2008-01-01, sunset		2008-01-02, sunset
The pictures above show the sunsets 10 and 11 days after solstice. When the sun sets at the same location in December, the observer can easily predict the date of the approaching solstice by adding 10 or 11 days.