________________
COMPUTATION OF DATES.
423
eclipses mentioned in inscriptions are not always actually observed eclipses, but cal. culated ones. My reasons for this opinion are the following :-Firstly, eclipses are auspicious moments, when donations, such as are usually recorded in inscriptions, are particularly meritorious. They were therefore probably selected for such occasions, and must accordingly have been calculated beforehand. No doubt they were entered in the pañohdngas or almanacs in former times as they are now. Secondly, even larger eclipses of the sun, up to seven digits, pass unobserved by common people, and smaller ones are only visible under favourable circumstances. Thirdly, the Hindus place implicit trust in their Sastras, and would not think it necessary to test their calculations by actual observation. The writers of inscriptions would therefore mention an eclipse if they found one predicted in their almanacs.
For determining the occurrence of eclipses the columns showing the sun's distance from the moon's nodes in Tables VI, VII, VIII, serve. The quantities are given in thousandth parts of the semicircle. In Table VI this quantity is given from modern European tables and also according to the Arya, Surya, and Brahma Siddhantas, and the Siddhanta Siromani. In the remaining tables the difference between the various authorities is so small that it is neglected. According to Hindu science
certain, if from Dode is between 0 and 90, or 910 and 1000 At new-moon a solar eclipse is 3 doubtful > >
91 105 909 895 Cimpossible,
106,894
certain, if from node is between 0 and 58, or 942 and 1000 At full-moon a lunar eclipse is doubtful
69 75, 911 928 Cimpossible, » » 76 , 922 41. A solar eclipse can only happen at the time of new-moon, i.e. when tithi is 0 or 30, and a lunar eclipse only at the time of full-moon, i.e. when the tithi is 15.00. It is also obvious that an eclipse of the moon is visible only when the moon is above the horizon during the eclipse, i.e. after sunset; and a solar eclipse is invisible after sunset. Therefore, in computing lunar eclipses, we calculate the moment of mean sunset, i.e. 30gh. For this we must add 0-51 to the tithi, 18 to anomaly, 3 to pode as shown below:
Ex.- Saka 861, 4030 K.Y. Magha-sudi 15, Sunday, a lunar eclipse. According to Tables I-III, and (node) Tables VI-VIII:
Per. Tithi. ('An. Node. 4000 K, Y. . . . (1) 8.98 523 62 Ind. 18.83 30 years ,
2:19 684 228 Ind. T'ithi 3.88 4080 K. Y..
11.17 27 Mágba, 30 ghatikdo.
• (1) 14-95 C's an. 40, eq.
Tithi 15.01
207
290)
2.81
52
» Anvelipeo which was not visible in Indin is recorded in Professor Kielhorn's paper, "Examination of questions con. nected with the Vikrama era."--Ind. Ant. vol. XIX, p. 116, eclipse No. 83.
* The limits of a solar eclipse are approximate only. They determine eclipses that might be visible at some point of the whole earth. The Hindu method of calculating solar eclipses is cumbrous, and the results cannot be given in a convenient tabular form. It is different with lunar eclipses. In the middle of solar Asbadha a lunar eclipse occurs, as calculated by the Sdrya Siddhants, when at fali-moon the Anomaly is 500 and 0 and distance from node 76 or 925, or anomaly 0 and distance of node 62 or 988; in the middle of solar Pausha, when at full-moon the anomaly is 600 and distance from node 74 or 926, or Anomaly O and distance from node 68 or 942. It will be seen that the limit is influenced more by the value of the anomaly tban by the time of the year. Details need not be entered upon here; these remarks will serve for most cases.
