... a tower of three storeys each over ten feet in height, within
which was concealed all the machinery. It was round at the top to symbolize
the heavens and square at the bottom to symbolize the earth. Below there
was set up the lower wheel, lower shaft, and the framework base. There were
also horizontal wheels, vertical wheels fixed sideways, and slanting wheels;
bearings for fixing them in place; a central stopping device and a smaller
stopping device [i.e. the escapement] with a main transmission shaft. Seven
jacks rang bells on the left, struck a large bell on the right, and beat
a drum in the n-dddle to indicate clearly the passing of the quarter-hours.
Each day and night [i.e. each 24 hours] the machinery made one complete
revolution, and the seven luminaries moved their positions around the ecliptic.
Twelve other wooden jacks were also made to come out at each of the double-hours,
one after the other, bearing tablets indicating the time. The lengths of
the days and nights were determined by the varying numbers of the quarters
passing in light or darkness. At the upper part of the machinery there were
the top piece, upper gear-wheels, upper stopping device [escapement], upper
anti-recoil ratchet pin, celestial ladder gear case [possibly the first chain
drive in history, or otherwise this was invented shortly afterwards by Su
Sung for his clock], upper beam of the framework, and the upper connecting-rod.
There were also on a celestial globe the 365 degrees to show the movements
of the sun, rnoon, and five planets; as well as the Purple Palace [north
polar region], the lunar mansions in their ranks and the Great Bear; together
with the equator and ecliptic which indicated how the changes of the advance
and regression of heat and cold depend upon the measured motions of the sun.
The motive power of the clock was water, according to the method which had
come down from Chang Heng in the Han Dynasty through I-Hsing ... But ...
as during the winter the water partly froze and its flow was greatly reduced,
the machinery lost its exactness, and there was no constancy between the
hot and the cold weather. Now, therefore, mercury was employed as a substitute,
and there were no more errors . . .