336 Scientific Proceedings, Royal Dublin Society. 



each of which is again divided into two. These 12 portions are 

 called after the signs of the Zodiac, and are, reckoning from 

 midnight, as follows : — Mouse-time, Bull-, Tiger-, Hare-, Dragon-, 

 Serpent-, Horse-, Groat-, Ape-, Cock-, Dog-, and Swine-time. The 

 middle of Mouse-time corresponds to midnight, so that this period 

 lasts from 11 p.m. till 1 a.m. The figures on the left-hand side 

 of the dial are Chinese, representing these different divisions of 

 the day. They- commence at the top with Cock, and proceed in 

 the order just given. 



The Japanese, on the other hand, reckoned time by hours of 

 variable length, depending on the season of the year. But at what 

 moment they commenced the day, and when their night began, is 

 not at all clear. Hoffman is most inconsistent with regard to this 

 point. He says that the Japanese measurement of time fixed 

 according to our hours is alone of value for an equinoctial day, and 

 then immediately afterwards states that a midsummer's day, 

 including the morning and evening twilight, being about 17 h. 58 m. 

 long, one of the divisions of a Japanese day would contain 

 2 h. 58 m. 



Now, it is evident that if the curves are hour-lines at all, they 

 cannot be intended to divide the day from sunrise till sunset, and 

 the night from sunset till sunrise each into equal portions, as 

 was in ancient times common enough (cf. the Hemisphere of 

 Berosus), and is still done in some of the out-of-the-way parts of 

 Europe. For on this hypothesis all the divisions would be of the 

 same length at the equinoxes only. Hence the first line would 

 have to represent the equinoxes. That this is not the case is clear 

 from the curves themselves, to say nothing of the evidence afforded 

 by the symbols contained at the top of the dial. For if it were, we 

 should expect the day to increase continually till the solstice 

 (which would be represented by a rise in the curve), from which 

 point the day would again diminish, and the curve would fall in 

 a similar manner till the equinox was again reached. The last line 

 therefore would represent the solstice. There would, however, on 

 this hypothesis be a provision for only one-half of the year. But at 

 the equinox the change in the length of the day is most rapid, so 

 that we should expect the curves to rise most abruptly close to the 

 first line. Now this is not at all the nature of the curves which 

 rise most steeply at the middle line, and this fact alone would lead 



