Clock of Strasbourg Cathedral |
The
clock is inside the building, rather than at a tower, like those of many other
cathedrals. It has a long history, as it dates back since the 14th century,
although it was completely rebuilt in the 16th. By the end of the 18th century
it stopped working. Legend says that, at the beginning of the 19th century, an
orderly who was showing the cathedral to a group of visitors mentioned that the
clock had not been working for a long time. Then a boy who was part of the group
exclaimed: I will fix it! Forty years
later, he did. That child would have been Jean-Baptiste Schwilgué, who
remodeled the clock around 1840.
Charles Babbage |
The cathedral chapter did not allow Schwilgué to modify the external appearance of the clock, but let him completely rebuild the mechanisms. Using gear wheels, Schwilgué built a device that simultaneously provides five different types of time: mean solar time at Strasbourg; official time; local solar time; sidereal time; and lunar time. The first is an ordinary clock moved by a pendulum and adjusted by hand; the others are derived from it. The official time is easy, because in Strasbourg and in winter time it differs by half an hour from the average solar. The local solar time is more complex, as it takes into account the variations in the length of the day throughout the year, and uses elliptical gears to calculate it. To get sidereal time from solar time, it would have been necessary to use wheels with millions of teeth, which is not technically feasible. Schwilgué came up with an approximation of the relationship between the two, with an error less than one second per century: 1+450/(611×269). Combining gears with 450, 611 and 269 teeth, it is easy to multiply by that fraction. Adding 1 to the result is equivalent to advancing one extra tooth.
In addition to the time, the clock provides information about the day of the year, taking into account the Gregorian calendar. A figure of the god Apollo marks the date on a sliding circular band, which each day advances one position and is divided into 365 positions, marked with the days of the year. In leap years, part of the band slides one position, covers an empty space located between December 31 and January 1 (which bears the inscription beginning of the ordinary year) and discovers another space, usually covered by the 28 of February, marked with the 29th of the same month. This mechanism is driven by a gear with 100 positions, of which 24 have teeth, separated from each other by three spaces without teeth, except for two teeth that are separated by seven spaces. The toothless spaces correspond to the ordinary years of 365 days. The teeth move the mechanism that causes the appearance of February 29. The seven spaces in a row correspond to the final years of the century, which according to the Gregorian calendar should not be leap years, although they would have been with the Julian calendar (see this post). This gear moves forward one position at the beginning of the year, making one revolution per century. Finally, there is another gear that rotates once every four hundred years and causes the appearance of an additional tooth on the previous gear, in the center of the area with seven spaces in a row, so that the years multiples of 400 become leap years, thus completing the correct treatment of all the years in our calendar.
Finally,
the clock contains complex mechanisms to calculate the date of Easter, whose algorithm
is quite complicated, as I explained in this
post. This part of the mechanism is not well understood, but the fact is
that it works correctly and causes the slippage of a section associated with
the band of the days of the year, marked with the movable feasts of the Church.
The clock is built to work for many centuries. After doing so for over 150 years, it successfully passed the year 2000. Unless the Gregorian calendar is reformed to correct its current error (about three days every 10,000 years), all the calculations performed by the clock will remain valid until the year 9999, a limit imposed because the count of years has just four figures. Schwilgué proposed that the life of his clock could be extended beyond the year 10,000 by writing a 1 in front of the figures of the year, but then the calculation of the date of Easter would no longer be correct, as this date does not follow a 10,000 year cycle.
No comments:
Post a Comment