Many thanks to those who have answered my questions. Geoffrey Thurston’s summary is clear and accords with the rudimentary, but effective, instrumental approach advocated by Lancelot Hogben in his epic book, Science for the Citizen.
Mike Shaw has added succinct advice, and the reason for choosing 1000 years was to divorce myself from sophisticated instruments and assume a certain naivety. However, Frank’s note make it clear that some prior knowledge would be useful. Indeed, makes the point indirectly, that we are always linked to the past. The comment about using the stars in the 1850s fits with my own understanding. The handy publication On the Line: The Story of the Greenwich Meridian, by Graham Dolan, National Maritime Museum (2005), shows a diagram containing 67 clock stars for a list dated 1850. Say three an hour! Pendulum clocks had reached a plateau in performance, but some horologists are fantasising by asking ‘what if’ questions. Specifically if John Harrison’s ideas ‘for 1 second in 100 days’ had been implemented there could have been a big improvement in longer term accuracy and thus of benefit to astronomers. Others have countered, given the complexity of making a clock as per Harrison, by saying that there was no demand for significant improvements. In fact at Greenwich, G B Airy concentrated on improving the transit telescope and observation techniques. Pendulum clocks only improved when the air pressure around the pendulum was held constant in a chamber (Riefler, 1900s) or lowering the air pressure (Hamilton Shortt, 1930s). Given my relatively limited knowledge of sundials, clocks and astronomy, and what is still to learn, I will demote myself to the making the tea in the Observatory… Regards, Doug > On 21 Feb 2017, at 09:19, Frank King <f...@cl.cam.ac.uk> wrote: > > Dear Doug, > > The answers to your first question about > noting seasonal drift 1000 years ago have > all missed something crucial... > > The astronomical comments have been very > sound but what is missing is any account > of just how you would maintain your > records? > > These days you could buy a desk-top diary > and write your observations in it on a > daily basis. I would like to know how > people did that 1000 years ago, never > mind in Babylonian times. > > You need to have invented a calendar and > some form of writing implement and you > need to be able to count. > > The Babylonians had this infra-structure > but their calendar had intercalary months > which were added ad hoc. I would have > found this hard to live with! > > Even if think about the relatively recent > times of just 1000 years ago you would, > today, have headed your entry: > > a d IX kal Mart MMXVII > > and you would need to be able to work out > what the date was CCCLXV days ago. > > In doing your calculations you have the > added difficulty that zero hasn't yet > been invented. Fractions are understood > but expressing results to a certain > number of decimal places is way into > the future. > > I have often pondered an even more primitive > question: I am dumped on a desert island and > I want to count the days since I arrived. > What discipline should I follow? > > I could, of course, cut a notch in a stick > every morning when I first wake up but > what happens when one day, around noon, I > think "Oh, er, did I cut today's notch?" > > I am by no means convinced that I could > reliably count up to CCCLXV. > > Can someone please come up with an > error-detecting and error-correcting > approach. > > I am not sure I understand your second > question: > > Assuming that in 1850s I had access > to a good transit telescope, and a > reasonable clock (daily errors > about 1 second a day)... > > Like you, I don't see why the clock > needs to be all that precise in a fixed > observatory. You can put it right on > a daily basis using your transit > telescope. > > Of course you wouldn't actually adjust > the clock, you would just log the error. > Indeed, by the 1850s, star tables were > sufficiently good that you could have > logged the error several times a night! > > Frank >
--------------------------------------------------- https://lists.uni-koeln.de/mailman/listinfo/sundial
