सिद्धान्तशिरोमणि: गणिताध्याय (भास्कराचार्य - ग्रहगणित, मध्यमाधिकार व स्पष्टाधिकार सटीक)

95 resulted in a difference between the computed planetary positions and their observed positions. So, the originator of this Bīja-Saṁskāra, noting the differences in his own time devised a formula, which could account for those differences. But this formulation was bound to go wrong in later times as long as the daily motions are not corrected to the minutest extent possible and as long as the funda- mental basis of the conjunction of all the planets and pla- netary points is not proved. This seems to be the reason why so may texts were written incorporating small diffe- rences in different times as reported by Gaṇēśa (1507 A.D.) in his work Bṛihat-Tithi-Chintāmaṇi in the words "The calculations of planetary positions according to the methods indicated by Brahma, Vasishtha and Kasyapa Siddhantas held good in their own times, but grew obsolete later; Then Maya, the demon at the end of Krita obtained the science from the Sun God, which again grew obsolete in this Kaliyuga wherein parāśara began to hold the ground for a good length of time. Then Āryabhata rectified the methods; when even those methods grew obsolete, Durga- Simha, Varāha Mihira and others set them right. Again Brahmagupta came into the picture to rectify the methods by his own observations. Then Came Kēśava (Gaṇēsa's father) who rectified further. After a lapse of sixty years, his son Gaṇesa has now to correct the Science. If this also grows obsolete (as it is bound to) in course of time, let others again rectify it by observing conjunctions of the Moon and planets with the asterisms." Obsoleteness arises out of two contexts, one a justifi- able situation and the other based upon a wrong premise. The first is as follows. Suppose as a first approximation we take the length of an year as 365 days. We will have committed an error nearly ¼ of a day, so that the error accures to a day in 4 years. Thus the convention of the leap year arose so that during four years we give a day more to February. Here again we have overestimated the error by nearly 1/100th of a day. Hence in 400 years the