204 Shinzo Siiinjo 



The 28 lidu system lL'-]r AJ^^ikj widely iu use in the ancient 

 Orient, is nothing but a mode of dividing the whole ecliptic into 28 

 parts, roughly corresponding to the number of days in a sidereal 

 month (27.3 days), thereby providing 28 reference points for the deter- 

 mination of the position of the conjunction points. Although the 

 original meaning has long been lost in oblivion, we see that the 

 system is still in use ia China, India, and Persia, with variations 

 necessarily resulting from long ranges of transference to different 

 countries. 



The birth-place of the 28 lisiu system has been and is now also in 

 dispute among scholars. Permit me to call attention here, however, 

 to some important points regarding this question. If we notice that 

 by prolonging the direction of the handle of the Pei-to2i ;jb^|'; we 

 arrive at the bright star Arcturus and still further at Spica, then the 

 facts, first that the Chinese system begins with Spica, secondly that 

 Arcturus is called Tai-cMao ::^^, while Spica is called GJiiao -f^, 

 and thirdly that the Indian system contains Arcturus among its 

 members, notwithstanding the distance of this star from the ecliptic, 

 — all these facts seem to indicate that the 28 hsiu system must have 

 come into use as a sequence to the Pei-tou system. Further, the fact 

 that the Indian system contains Vega and Altair among its members, 

 both distant from the ecliptic, seems to suggest the origin of the 

 system in a country like China, where the love-story of Chien niu ^^ 

 and Chih nii ^;^, i.e., Vega and Altair, has been very familiar to 

 the public since remote ages. 



From this and other evidence, it seems to me almost without 

 doubt that the 28 hsiu system was first invented and adopted in 

 China about the 12th century B.C. in sequence to the Pei-tou system, 

 that it was transferred into India, probably about the Ghan-kuo age 

 in China (480 B.C.- 250 B.C.), and that the Chinese system has since 

 undergone several readjustments, of which at least one in the Han 

 dynasty is recorded in history, while the Indian system still represents 

 the primitive form. 



A further important step was made by the introduction of 

 gnomon, a little before the middle of the Ch'un-ch'iu period ^liltB^i^ 

 (722-481 B.C.). Although the use of the gnomon ±^^ is described 

 iu the Ghou-li ^|^, the date of compilation of which is commonly 

 taken to bo at the beginning of the Chou dynasty, we shall not con- 

 sider this point, the Ghou-U itself being surely the production of a 



