948 



means of series of 5 or 6 successive eclipses, ') and he assnmes that 

 when, in the Assyrian period, predictions are made, this rule is 

 made use of. No proof can be given of this, as ihere is no text in 

 which mention is made of the means by whicli the results were 

 obtained. And a prediction is possible, as shown above, with even 

 more primitive knowledge. Thai these series of 5 or 6 were known 

 in the succeeding centuries is certain, because a knowledge of them 

 is a station on the line that leads to the discovery of the saros. 



To trace this road we must first consider what phenomena and 

 what regularity an attentive observer would further be able to notice 

 in the lunar eclipses. If a certain series of full moons (by a series 

 we here mean always full moons following each other with intervals 

 of six lunar months) recedes ^more and more from the nodal line, 

 the series of full moons preceding them constantly approaches it. 

 The longitude of the former full moon is 29°, 11 smaller, a month 

 earlier the node lay upon 1°,564 greater longitude, therefore for 

 this previous full moon L — il^ P— 29°,11 — 1,564 = P— 30°67. 

 We will now tabulate P for both series beside each other beginning 

 with the 4-'' of the above ro\v, and beside it a few of the preceding 

 full moons. 



1) G, ScHiAPARELLi, I primordi dell' astronomia presso i babilonesi. (Scientia IV p. 36). 



