334 Scientific Proceedings, Royal Dublin Society. 



numerals in fig. 3. The symbol at each end of the second row 

 corresponds to the word "month." In the third row the two 

 symbols *f and if occur alternately. These symbols denote the 

 first half and second half of the month, respectively. 



The comparison of these periods of time with our ordinary 

 months is very much complicated by the intricacies of the Chinese 

 calendar ; for although the Chinese, if their records are authentic, 

 had determined the length of the year with very great accuracy 

 ages before any precision was attained in this quantity by 

 Europeans — reckoning it at 365 \ days as early as 2000 B.C. — still 

 the years which are commonly employed and which are used for 

 chronological purposes are lunisolar years, and an elaborate system 

 of intercalary months has to be maintained in order to keep the 

 seasons in their proper places. 



As far as I have been able to ascertain, the Chinese civil year 

 commences at the first new moon after the Sun enters the sign of 

 Pisces. The Chinese new year must therefore lie between 22nd 

 January and 20th February. The months also are lunar months, 

 and the year consists of twelve of them. The civil year, therefore, 

 consists of 354 or 355 days only, and consequently it becomes 

 necessary every third or fourth year to introduce an intercalary 

 month. Now it is obvious that according to this style of reckoning, 

 no two years, unless they happened to be separated by some 

 multiple of the Metonic cycle of 19 years, would be similar. 



I find, however, in Hoffman's Japanese Grammar a reference 

 to a solar year, which is divided into twelve equal parts, each of 

 which is again divided into two. In Williams' Chinese Obser- 

 vations of Comets, there is given a Table of the 24 divisions of the 

 Chinese year in which the 24 periods contain in all 365 or 366 

 days, but, unfortunately, there is no reference to or explanation of 

 it in the text. 



I conclude, therefore, that in addition to the lunisolar year of 

 varying length, which was used for official and chronological pur- 

 poses, there was the uniform solar year of 365^ days employed for 

 other purposes for which it was found convenient. 



I find, according to Williams, that the first of these 12 double 

 divisions of the solar year corresponds to the 5th of February, 

 according to our style of reckoning, and that the second half of the 

 second period and the second half of the eighth period correspond 



