When one considers how widely researched and revered the applied and decorative arts of Japan generally are in the West, it is surprising that more is not commonly known of the art and craft of the Japanese clockmaker. The aim of this article is to offer an introduction to the way in which the concept of time itself was viewed in Japan and the various mechanical instruments that were created to measure it. Although rare, some of these clocks and timepieces can still be found on the open market and this article will focus on those most likely to be encountered.

Figure 3. An early 19th century shitan Shaku dokei. The movement frontplate is cast and engraved in the form of a vase of chrysanthemums. The silvered dragonfly at the top of the movement moves as the movement runs. The time is indicated by the pierced brass pointer. 16in high.

Japanese clocks made prior to 1873 do not have the traditional dials with fixed numerals that are used today. Until 1st January of that year, the Japanese day was divided into six equal parts, and the night into another six equal parts. It is immediately obvious that this system means that the six equal parts changed in length as the duration of hours of daylight and darkness changed with the seasons. It is also evident that, even at their shortest duration, one Japanese 'hour' would last much longer than a European hour. The former might range from about 90 minutes in winter daytime to about 150 minutes in summer daytime, while the latter always lasts 60 minutes, regardless of the time or season. Japanese clocks, therefore, had to have some way of taking into account the complex and elastic nature of their hours and several methods were developed as mechanical timekeepers evolved.
The first mechanical clocks were brought into Japan in the mid-16th century by Jesuit missionaries. Foreign governments and traders also took them to Japan in the hope of gaining political advantage or opening up trade routes. The earliest surviving clock in Japan is dated 1581 and was a gift from the Governor of Mexico to the founder of the Tokugawa shogunate, lyeyasu, in 1612. It is interesting that the Governor did not think it unreasonable to offer a 40-year-old clock as a gift; it implies that clocks were still sufficiently rare for this to be regarded as a fit offering.
Japan has a long tradition of skilled blacksmiths, swordsmiths and metal workers, and the missionaries trained some of these craftsmen in the upkeep and restoration of their timepieces. In doing so, a Japanese clockmaking tradition was established and while some European lineage can be seen in the movements and cases, a Japanese clock cannot be mistaken for as product of any other civilisation. Much to our chagrin, researching individual makers is extremely difficult because, unlike European and American clocks, Japanese clocks are very rarely signed.
Both spring-driven and weight-driven movements were made, the earlier iron ones gradually replaced by brass. While many Western clockmakers regularly made movements of 8-day duration or longer, Japanese movements rarely run for more than a single day on one winding. Bearing in mind that the movement or dial was adjusted every day to take into account the variable hours, it was no additional bother to wind the clock at the same time. At all times, clocks were regarded as precious objects and in the earlier stages of development were found only in temples or the houses of the nobility where 'clock doctors' were employed to adjust and maintain them regularly. Figure 1 shows a small clock in a domestic setting with other treasures.
The simplest solution to overcome the problem of movable hours came in the form of an easily regulated movement controller. The foliot is a horizontal bar set above the movement that regulates the speed at which a clock movement runs down. Japanese clockmakers used this device far longer than their Western counterparts because it suited their demands so well. The foliot bars were cut with notches along their length to carry pairs of weights on opposite sides of the bar. Moving the weights away from the centre caused the foliot, and therefore the clock, to beat more slowly, while moving them nearer to the centre meant that the clock would run more quickly. This change had to be effected twice every day - once at sunrise and again at sunset - so that the clock would run at the correct rate for each 6-hour period.
An improvement to this very labour-intensive system was the double foliot. In these clocks, there are two foliot bars set above the case, each with the timing weights set at pre-determined levels responding to the current daylight or night-time going rate. A mechanism was tripped inside the movement at sunrise and sunset which automatically dis-engaged one foliot bar and activated the other. Figure 2 shows a clock with traditional dial and double foliot.
The third manner in which the flexible hours were accommodated was to have individual numerals that could be moved relative to one another. The most commonly found example of this type of dial is used on 'stick' or pillar clocks, known locally as Shaku dokei.