Pramāṇa · Search · About · API

The Indian Calendar, with Tables for the Conversion of Hindu and Muhammadan into A.D. Dates, and vice versa c7: Part III. Description and Explanation of the Tables

The Indian Calendar, with Tables for the Conversion of Hindu and Muhammadan into A.D. Dates, and vice versâ c7

Language
English

Part III. Description and Explanation of the Tables

c7.2

73. Table I.—Table I. is our principal and general Table, and it forms the basis for all calculations. It will be found divided into three sections. (1) Table of concurrent years; (2) intercalated and suppressed months; (3) moments of commencement of the solar and luni-solar years. All the figures refer to mean solar time at the meridian of Ujjain. The calculations are based on the Sûrya-Siddhânta, without the bîja up to 1500 A.D. and with it afterwards, with the exception of cols. 13 to 17 inclusive for which the Ârya-Siddhânta has been used. Throughout the table the solar year is taken to commence at the moment of the apparent Mêsha saṅkrânti or first point of Aries, and the luni-solar year with amânta Chaitra śukla pratipadâ. The months are taken as amânta.

c7.3

74. Cols. 1 to 5.—In these columns the concurrent years of the six principal eras are given. (As to current and expired years see Art. 70 above.) A short description of eras is given in Art. 71. The years in the first three columns are used alike as solar and luni-solar, commencing respectively with Mesha or Chaitra. (For the beginning point of the year see Art. 52 above.) The Vikrama year given in col. 3 is the Chaitrâdi Vikrama year, or, when treated as a solar year which is very rarely the case, the Meshâdi year. The Âshâdhâdi and Kârttikâdi Vikrama years are not given, as they can be regularly calculated from the Chaitrâdi year, remembering that the number of the former year is one less than that of the Chaitrâdi year from Chaitra to Jyeshtha or Âśvina (both inclusive), as the case may be, and the same as the Chaitrâdi year from Âshâdha or Kârttika to the end of Phâlguna.

c7.4

¹ Prinsep's Indian Antiquities, II., Useful Tables, p. 171.

c7.5

Cols. 4 and 5. The eras in cols. 4 and 5 are described above (Art. 71.) The double number is entered in col. 4 so that it may not be forgotten that the Kollam year is non-Chaitrâdi or non-Meshâdi, since it commences with either Kanni (Kanyâ) or Chîngam (Simha). In the case of the Christian era of course the first year entered corresponds to the Kali, Śaka or Chaitrâdi Vikrama year for about three-quarters of the latter's course, and for about the last quarter the second Christian year entered must be taken. The corresponding parts of the years of all these eras as well as of several others will be found in Table II., Parts ii. and iii.

c7.6

75. Cols. 6 and 7.—These columns give the number and name of the current samvatsara of the sixty-year cycle. There is reason to believe that the sixty-year luni-solar cycle (in use mostly in Southern India) came into existence only from about A. D. 909; and that before that the cycle of Jupiter was in use all over India. That is to say, before A. D. 909 the samvatsaras in Southern India were the same as those of the Jupiter cycle in the North. If, however, it is found in any case that in a year previous to A.D. 908 the samvatsara given does not agree with our Tables, the rule in Art. 62 should be applied, in order to ascertain whether it was a luni-solar samvatsara.

c7.7

The samvatsara given in col. 7 is that which was current at the time of the Mesha saṅkrânti of the year mentioned in cols. 1 to 3. To find the samvatsara current on any particular day of the year the rules given in Art. 59 should be applied. For other facts regarding the samvatsaras, see Arts. 53 to 63 above.

c7.8

76. Cols. 8 to 12, and 8a to 12a. These concern the adhika (intercalated) and kshaya (suppressed) months. For full particulars see Arts. 45 to 51. By the mean system of intercalations there can be no suppressed months, and by the true system only a few. We have given the suppressed months in italics with the suffix "Ksh" for "kshaya." As mean added months were only in use up to A.D. 1100 (Art. 47) we have not given them after that year.

c7.9

77. The name of the month entered in col. 8 or 8a is fixed according to the first rule for naming a lunar month (Art. 46), which is in use at the present day. Thus, the name Âshâdha, in cols. 8 or 8a, shows that there was an intercalated month between natural Jyeshtha and natural Âshâdha, and by the first rule its name is "Adhika Âshâdha", natural Âshâdha being "Nija Âshâdha." By the second rule it might have been called Jyeshtha, but the intercalated period is the same in either case. In the case of expunged months the word "Pausha", for instance, in col. 8 shows that in the lunar month between natural Kârttika and natural Mâgha there were two saṅkrântis; and according to the rule adopted by us that lunar month is called Mârgaśirsha, Pausha being expunged.

c7.10

⁷⁸ Lists of intercalated and expunged months are given by the late Prof. K. L. Chhatre in a list published in Vol. I., No. 12 (March 1851) of a Mahrâthi monthly magazine called Jñânaprasâraka, formerly published in Bombay, but now discontinued; as well as in Cowasjee

c7.11

Patell's "Chronology", and in the late Gen. Sir A. Cunningham's "Indian Eras,"¹ But in none of these three works is a single word said as to how, or following what authority, the calculations were made, so that we have no guide to aid us in checking the correctness of their results.

c7.12

¹ Gen. Cunningham admittedly (p. 91) follows Cowasjee Patell's "Chronology" in this respect, and on examination I find that the added and suppressed months in these two works (setting aside some few mistakes of their own) agree throughout with Prof. Chhatre's list, even so far as to include certain instances where the latter was incorrect. Patell's "Chronology" was published fifteen years after the publication of Prof. Chhatre's list, and it is not improbable that the former was a copy of the latter. It is odd that not a single word is said in Cowasjee Patell's work to shew how his calculations were made, though in those days he would have required months or even years of intricate calculation before he could arrive at his results. [S B. D.]

c7.13

79. An added lunar month being one in which no saṅkrānti of the sun occurs, it is evident that a saṅkrānti must fall shortly before the beginning, and another one shortly after the end, of such a month, or in other words, a solar month must begin shortly before and must end shortly after the added lunar month. It is further evident that, since such is the case, calculation made by some other Siddhānta may yield a different result, even though the difference in the astronomical data which form the basis of calculation is but slight. Hence we have deemed it essential, not only to make our own calculations afresh throughout, but to publish the actual resulting figures which fix the months to be added and suppressed, so that the reader may judge in each case how far it is likely that the use of a different authority would cause a difference in the months affected. Our columns fix the moment of the saṅkrānti before and the saṅkrānti after the added month, as well as the saṅkrānti after the beginning, and the saṅkrānti before the end, of the suppressed month; or in other words, determine the limits of the adhika and kshaya māsas. The accuracy of our calculation can be easily tested by the plan shewn in Art. 90 below. (See also Art. 88 below.) The moments of time are expressed in two ways, viz., in lunation-parts and tithis, the former following Prof. Jacobi's system as given in Ind. Ant., Vol. XVII.

c7.14

80. Lunation-parts or, as we elsewhere call them, "tithi-indices" (or "t") are extensively used throughout this work and require full explanation. Shortly stated a lunation-part is 1/10000th of an apparent synodic revolution of the moon (see Note 2, Art. 12 above). It will be well to put this more clearly. When the difference between the longitude of the sun and moon, or in other words, the eastward distance between them, is nil, the sun and moon are said to be in conjunction; and at that moment of time occurs (the end of) amāvāsyā, or new moon. (Arts. 7.29 above.) Since the moon travels faster than the sun, the difference between their longitudes, or their distance from one another, daily increases during one half and decreases during the other half of the month till another conjunction takes place. The time between two conjunctions is a synodic lunar month or a lunation, during which the moon goes through all its phases. The lunation may thus be taken to represent not only time but space. We could of course have expressed parts of a lunation by time-measure, such as by hours and minutes, or ghaṭikās and palas, or by space-measure, such as degrees, minutes, or seconds, but we prefer to express it in lunation-parts, because then the same number does for either time or space (see Art. 89 below). A lunation consists of 30 tithis. 1/30th of a lunation consequently represents the time-duration of a tithi or the space-measurement of 12 degrees. Our lunation is divided into 10,000 parts, and about 333 lunation-parts (1/10000ths) go to one tithi, 667 to two tithis, 1000 to three and so on. Lunation-parts are therefore styled "tithi-indices", and by abbreviation simply "t". Further, a lunation or its parts may be taken as apparent or mean. Our tithi-, nakshatra-, and yoga-indices are apparent and not mean, except in the case of mean added months, where the index, like the whole lunation, is mean.

c7.15

Our tithi-index, or "t", therefore shows in the case of true added months as well as elsewhere, the space-difference between the apparent, and in the case of mean intercalations between the mean, longitudes of the sun and moon, or the time required for the motions of the sun and moon to create that difference, expressed in 10,000ths of a unit, which is a circle in the case of space, and a lunation or synodic revolution of the moon in the case of time. Briefly the tithi-index "t" shews the position of the moon in her orbit with respect to the sun, or the time necessary for her to gain that position., e.g., "0" is new moon, "5000" full moon, "10,000" or "0" new moon; "50" shews that the moon has recently (i.e., by 50/10000ths, or 3 hours 33 minutes—Table X., col. 3) passed the point or moment of conjunction (new moon); 9950 shews that she is approaching new-moon phase, which will occur in another 3 hours and 33 minutes.

c7.16

81. A lunation being equal to 30 tithis, the tithi-index, which expresses the 10,000th part of a lunation, can easily be converted into tithi-notation, for the index multiplied by 30 (practically by 3), gives, with the decimal figures marked off, the required figure in tithis and decimals. Thus if the tithi-index is 9950, which is really 0.9950, it is equal to (0.9950 × 30 =) 29.850 tithis, and the meaning is that 9950/10000ths of the lunation, or 29.850 tithis have expired. Conversely a figure given in tithis and decimals divided by 30 expresses the same in 10,000ths parts of a lunation.

c7.17

82. The tithi-index or tithi is often required to be converted into a measure of solar time, such as hours or ghaṭikās. Now the length of an apparent lunation, or of an apparent tithi, perpetually varies, indeed it is varying at every moment, and consequently it is practically impossible to ascertain it except by elaborate and special calculations; but the length of a mean lunation, or of a mean tithi, remains permanently unchanged. Ignoring, therefore, the difference between apparent and mean lunations, the tithi-index or tithi can be readily converted into time by our Table X., which shews the time-value of the mean lunation-part (1/10000th of the mean lunation), and of the mean tithi-part (1/1000th of the mean tithi). Thus, if t = 50, Table X. gives the duration as 3 hours 33 minutes; and if the tithi-part 1 is given as 0.150 we have by Table X. (2 h. 22 m. + 1 h. 11 min. =) 3 h. 33 m.

c7.18

It must be understood of course that the time thus given is not very accurate, because the tithi-index (t) is an apparent index, while the values in Table X. are for the mean index. The same remark applies to the nakshatra (n) or yoga (y) indices, and if accuracy is desired the process of calculation must be somewhat lengthened. This is fully explained in example 1 in Art. 148 below. In the case of mean added months the value of (t) the tithi-index is at once absolutely accurate.

c7.19

83. The saṅkrāntis preceding and succeeding an added month, as given in our Table I., of course take place respectively in the lunar month preceding and succeeding that added month.

c7.20

84. To make the general remarks in Arts. 80, 81, 82 quite clear for the intercalation of months we will take an actual example. Thus, for the Kali year 3403 the entries in cols. 9 and 11 are 9950 and 287, against the true added month Āśvina in col. 8. This shews us that the saṅkrānti preceding the true added, or Adhika, Āśvina took place when 9950 lunation-parts of the natural month Bhādrapada (preceding Adhika Āśvina) had elapsed, or when (10,000 - 9950 =) 50 parts had to elapse before the end of Bhādrapada, or again when 50 parts had to elapse before the beginning of the added month; and that the saṅkrānti succeeding true Adhika Āśvina took place when 287 parts of the natural month Nija Āśvina had elapsed, or when 287 parts had elapsed after the end of the added month Adhika Āśvina.

c7.21

¹ A thousandth part of a tithi is equal to 1.42 minutes, which is sufficiently minute for our purposes, but a thousandth of a lunation is equivalent to 7 hours 5 minutes, and this is too large; so that we have to take the 10000th of a lunation as our unit, which is equal to 4.25 minutes, and this suffices for all practical purposes. In this work therefore a lunation is treated of as having 10,000 parts, and a tithi 1000 parts.

c7.22

85. The moments of the saṅkrāntis are further given in tithis and decimals in cols. 10, 12, 10a and 12a. Thus, in the above example we find that the preceding saṅkrānti took place when 29.850 tithis of the preceding month Bhādrapada had elapsed, i.e., when (30—29.850 =) 0.150 tithis had still to elapse before the end of Bhādrapada; and that the succeeding saṅkrānti took place when 0.861 of a tithi of the succeeding month, Āśvina, had passed.

c7.23

To turn these figures into time is rendered easy by Table X. We learn from it that the preceding saṅkrānti took place (50 lunation parts or 0.150 tithi parts) about 3 h. 33 m. before the beginning of Adhika Āśvina; and that the succeeding saṅkrānti took place (287 lunation parts, or .861 tithi parts) about 20 h. 20 m. after the end of Adhika Āśvina. This time is approximate. For exact time see Arts. 82 and 90.

c7.24

The tithi-indices here shew (see Art. 88) that there is no probability of a different month being intercalated if the calculation be made according to a different authority.

c7.25

86. To constitute an expunged month we have shewn that two saṅkrāntis must occur in one lunar month, one shortly after the beginning and the other shortly before the end of the month; and in cols. 9 and 10 the moment of the first saṅkrānti, and in cols. 11 and 12 that of the second saṅkrānti, is given. For example see the entries against Kali 3506 in Table I. As already stated, there can never be an expunged month by the mean system

c7.26

87. In the case of an added month the moon must be waning at the time of the preceding, and waxing at the time of the succeeding saṅkrānti, and therefore the figure of the tithi-index must be approaching 10,000 at the preceding, and over 10,000, or beginning a new term of 10,000, at the succeeding, saṅkrānti. In the case of expunged months the case is reversed, and the moon must be waxing at the first, and waning at the second saṅkrānti; and therefore the tithi-index must be near the beginning of a period of 10,000 at the first, and approaching 10,000 at the second, saṅkrānti.

c7.27

88. When by the Sūrya-Siddhānta a new moon (the end of the amāvāsyā) takes place within about 6 ghaṭikās, or 33 lunation-parts, of the saṅkrānti, or beginning and end of a solar month, there may be a difference in the added or suppressed month if the calculation be made according to another Siddhānta. Hence when, in the case of an added month, the figure in col. 9 or 9a is more than (10,000—33 =) 9967, or when that in col. 11 or 11a is less than 33; and in the case of an expunged month when the figure in col. 9 is less than 33, or when that in col. 11 is more than 9967, it is possible that calculation by another Siddhānta will yield a different month as intercalated or expunged; or possibly there will be no expunction of a month at all. In such cases fresh calculations should be made by Prof. Jacobi's Special Tables (Epig. Ind., Vol. II.) or direct from the Siddhānta in question. In all other cases it may be regarded as certain that our months are correct for all Siddhāntas. The limit of 33 lunation-parts here given is generally sufficient, but it must not be forgotten that where Siddhāntas are used with a bīja correction the difference may amount to as much as 20 ghaṭikās, or 113 lunation-parts (See above, note to Art. 49).

c7.28

In the case of the Sūrya-Siddhānta it may be noted that the added and suppressed months are the same in almost all cases, whether the bīja is applied or not.

c7.29

89. We have spared no pains to secure accuracy in the calculation of the figures entered in cols. 9 to 12 and 9a to 12a, and we believe that they may be accepted as finally correct, but it should be remembered that their time-equivalent as obtained from Table X. is only approximate for the reason given above (Art. 82.) Since Indian readers are more familiar with tithis than with lunation-parts, and since the expression of time in tithis may be considered desirable by some European workers, we have given the times of all the required saṅkrāntis in tithis and decimals in our columns, as well as in lunation-parts; but for turning our figures into time-figures it is easier to work with lunation-parts than with tithi-parts. It may be thought by some readers that instead of recording the phenomena in lunation-parts and tithis it would have been better to have given at once the solar time corresponding to the moments of the saṅkrāntis in hours and minutes. But there are several reasons which induced us, after careful consideration, to select the plan we have finally adopted. First, great labour is saved in calculation; for to fix the exact moments in solar time at least five processes must be gone through in each case, as shewn in our Example I. below (Art. 148). It is true that, by the single process used by us, the time-equivalents of the given lunation-parts are only approximate, but the lunation-parts and tithis are in themselves exact. Secondly, the time shewn by our figures in the case of the mean added months is the same by the Original Sūrya, the Present Sūrya, and the Ārya-Siddhānta, as well as by the Present Sūrya-Siddhānta with the bīja, whereas, if converted into solar time, all of these would vary and require separate columns. Thirdly, the notation used by us serves one important purpose. It shews in one simple figure the distance in time of the saṅkrāntis from the beginning and end of the added or suppressed month, and points at a glance to the probability or otherwise of there being a difference in the added or suppressed month in the case of the use of another authority. Fourthly, there is a special convenience in our method for working such problems as are noticed in the following articles.

c7.30

90. Supposing it is desired to prove the correctness of our added and suppressed months, or to work them out independently, this can easily be done by the following method: The moment of the Mesha saṅkrānti according to the Sūrya-Siddhānta is given in cols. 13, 14 and 15a to 17a for all years from A.D. 1100 to 1900, and for other years it can be calculated by the aid of Table D. in Art. 96 below. Now we wish to ascertain the moment of two consecutive new moons connected with the month in question, and we proceed thus. The interval of time between the beginning of the solar year and the beginning or end of any solar month according to the Sūrya-Siddhānta, is given in Table III., cols. 8 or 9; and by it we can obtain by the rules in Art. 151 below, the tithi-index for the moment of beginning and end of the required solar month, i.e., the moments of the solar saṅkrāntis, whose position with reference to the new moon determines the addition or suppression of the luni-solar month. The exact interval also in solar time between those respective saṅkrāntis and the new moons (remembering that at new moon "t" = 10,000) can be calculated by the same rules. This process will at once shew whether the moon was waning or waxing at the preceding and succeeding saṅkrāntis, and this of course determines the addition or suppression of the month. The above, however, applies only to the apparent or true intercalations and suppressions. For mean added months the Śodhya (2 d. 8 gh. 51 p. 15 vi.) must be added (see Art. 26) to the Mesha-saṅkrānti time according to the Ārya-Siddhānta (Table I., col. 15), and the result will be the time of the mean Mesha saṅkrānti. For the required subsequent saṅkrāntis all that is necessary is to add the proper figures of duration as given in Art. 24, which shews the mean length of solar months, and to find the "a" for the results so obtained by Art. 151. Then add 200 to the totals and the result will be the required tithi-indices.

c7.31

91. It will of course be asked how our figures in Table I. were obtained, and what guarantee we can give for their accuracy. It is therefore desirable to explain these points. Our calculations for true intercalated and suppressed months were first made according to the method and Tables published by Prof. Jacobi (in the *Ind. Ant.*, Vol. XVII., *pp.* 145 to 181) as corrected by the errata list printed in the same volume. We based our calculations on his Tables I to 10, and the method given in his example 4 on pp. 152–53, ¹ but with certain differences, the necessity of which must now be explained. Prof. Jacobi's Tables I to 4, which give the dates of the commencement of the solar months, and the hour and minute, were based on the *Ārya-Siddhānta*, while Tables 5 to 10 followed the *Sûrya-Siddhānta*, and these two *Siddhāntas* differ. In consequence several points had to be attended to. First, in Prof. Jacobi's Tables I to 4 the solar months are supposed to begin exactly at Ujjain mean sunset, while in fact they begin (as explained by himself at p. 147) at or shortly after mean sunset. This state of things is harmless as regards calculations made for the purpose for which the Professor designed and chiefly uses these Tables, but such is not the case when the task is to determine an intercalary month, where a mere fraction may make all the difference, and where the exact moment of a saṅkrānti must positively be ascertained. Secondly, the beginning of the solar year, *i.e.*, the moment of the Mesha-saṅkranti, differs when calculated according to those two *Siddhāntas*, as will be seen by comparing cols. 15 to 17 with cols. 15*a* to 17*a* of our Table I., the difference being *nil* in A.D. 496 and 6 gh 23 pa. 41.4 pra. vi. in 1900 A.D. Thirdly, even if we suppose the year to begin simultaneously by both *Siddhāntas*, still the collective duration of the months from the beginning of the year to the end of the required solar month is not the same, ² as will be seen by comparing cols. 6 or 7 with cols. 8 or 9 of our Table III. We have applied all the corrections necessitated by these three differences to the figures obtained from Prof. Jacobi's Tables and have given the final results in cols. 9 and 11. We know of no independent test which can be applied to determine the accuracy of the results of our calculations for true added and suppressed months; but the first calculations were made exceedingly carefully and were checked and rechecked. They were made quite independently of any previously existing lists of added and suppressed months, and the results were afterwards compared with Prof. Chhatre's list; and whenever a difference appeared the calculations were completely re-examined. In some cases of expunged months the difference between the two lists is only nominal, but in other cases of difference it can be said with certainty that Prof. Chhatre's list is wrong. (*See note to Art. 46.*) Moreover, since the greatest possible error in the value of the tithi-index that can result by use of Prof. Jacobi's Table is 7 (*see his Table p. 164*), whenever the tithi-index for added and suppressed months obtained by our computation fell within 7 of 10,000, *i.e.*, whenever the resulting index was below 7 or over 9993, the results were again tested direct by the *Sûrya-Siddhānta*. ³

c7.32

¹ For finding the initial date of the luni-solar years Prof. Jacobi's Tables I. to XI. were used, and in the course of the calculations it was necessary to introduce a few alterations, and to correct some misprints which had crept in in addition to those noted in the already published errata-list. Thus, the earliest date noted in Tables I. to IV., being A.D. 354, these Tables had to be extended backwards by adding two lines more of figures above those already given. In Table VI., as corrected by the errata, the bīja is taken into account only from A.D. 1601, whereas we consider that it should be introduced from A.D. 1501 (see Art. 21). In Table VI. the century correction is given for the New (Gregorian) Style from A.D. 1600 according to the practice in the most part of Europe. I have preferred, however, to introduce the New Style into our Tables from Sept. A.D. 1752 to suit English readers, and this necessitated an alteration in the century data for two centuries. [R. S.]

c7.33

² It is the same according to Warren, but in this respect he is in error. (*See note to Art. 24.*)

c7.34

³ 42 calculations were thus made direct by the *Sûrya-Siddhānta* with and without the bīja, with the satisfactory result that the error in the final figure of the tithi-index originally arrived at was generally only of 1 or 2 units, while in some cases it was *nil* It was rarely 3, and only once 4. It never exceeded 4. It may therefore be fairly assumed that our results are accurate. [S.B D.]

c7.35

As regards mean intercalations every figure in our cols. 9*a* to 12*a* was found correct by independent test. The months and the times of the saṅkrāntis expressed in tithi-indices and tithis were calculated by the present *Sûrya-Siddhānta*, and the results are the same whether worked by that or by the Original Sûrya-Siddhânta, the First Ârya-Siddhânta, or the Present Sûrya-Siddhânta with the bîja.

c7.36

We think, therefore, that the list of true added and suppressed months and that of the mean added months as given by us is finally reliable.

c7.37

92. *Cols. 13 to 17 or to 17a.* The solar year begins from the moment of the Mesha saṅkrânti and this is taken as apparent and not mean. We give the exact moment for all years from A.D. 300 to 1900 by the Ârya-Siddhânta, and in addition for years between A.D. 1100 and 1900 by the Sûrya-Siddhântas as well. (*See also Art. 96*). Every figure has been independently tested, and found correct. The week-day and day of the month A.D. as given in cols. 13 and 14 are applicable to both the Siddhântas, but particular attention must be paid to the footnote in Table I., annexed to A.D. 1117-18 and some other subsequent years. The entries in cols. 15 and 15a for Indian reckoning in ghaṭikâs and palas, and in cols. 17 and 17a for hours and minutes, imply that at the instant of the saṅkrânti so much time has elapsed since mean sunrise at Ujjain on the day in question. Ujjain mean sunrise is generally assumed to be 6.0 a.m.

c7.38

93. The alteration of week-day and day of the month alluded to in the footnote mentioned in the last paragraph (Table I., A.D. 1117-18) is due to the difference resulting from calculations made by the two Siddhântas, the day fixed by the Sûrya-Siddhânta being sometimes one later than that found by the Ârya-Siddhânta. It must be remembered, however, that the day in question runs from sunrise to sunrise, and therefore a moment of time fixed as falling between midnight and sunrise belongs to the preceding day in Indian reckoning, though to the succeeding day by European nomenclature. For example, the Mesha saṅkrânti in Śaka 1039 expired (A.D. 1117) took place, according to the Ârya-Siddhânta on Friday 23rd March at 58 gh. 1p. after Ujjain mean sunrise (23 h. 12 m. after sunrise on Friday, or 5.12 a.m. on Saturday morning, 24th); while by the Sûrya-Siddhânta it fell on Saturday 24th at 0 gh. 51 pa. (=0 h. 20 m. after sunrise or 6.20 a.m.). This only happens of course when the saṅkrânti according to the Ârya-Siddhânta falls nearly at the end of a day, or near mean sunrise.

c7.39

94. In calculating the instant of the apparent Mesha-saṅkrântis, we have taken the śodhya at 2 d. 8 gh. 51 pa. 15 vipa. according to the Ârya-Siddhânta, and 2 d. 10 gh. 14 pa. 30 vipa. according to the Sûrya-Siddhânta. (*See Art. 26.*)

c7.40

95. The figure given in brackets after the day and month in cols. 13 and 19 is the number of that day in the English common year, reckoning from January 1st. For instance, 75 against 16th March shows that 16th March is the 75th day from January 1st inclusive. This figure is called the "date indicator", or shortly (d), in the methods of computation "B" and "C" given below (*Part IV.*), and is intended as a guide with reference to Table IX., in which the collective duration of days is given in the English common year.

c7.41

96. The fixture of the moments of the 1600 Mesha-saṅkrântis noted in this volume will be found advantageous for many purposes, but we have designed it chiefly to facilitate the conversion of solar dates as they are used in Bengal and Southern India.¹ We have not given the moments of Mesha-saṅkrântis according to the Sûrya-Siddhânta prior to A.D. 1100, so that the Ârya-Siddhânta computation must be used for dates earlier than that, even those occurring in Bengal. There is little danger in so doing, since the difference between the times of the Mesha-saṅkrântis according to the two Siddhântas during that period is very slight, being nil in A.D. 496, and only increasing to 1 h. 6 m. at the most in 1100 A.D. It is, however, advisable to give a correction Table so as to ensure accuracy, and consequently we append the Table which follows, by which the difference for any year lying between A.D. 496 and 1100 A.D. can be found. It is used in the following manner. First find the interval in years between the given year and A.D. 496. Then take the difference given for that number of years in the Table, and subtract or add it to the moment of the Mesha-saṅkrānti fixed by us in Table I. by the Ārya-Siddhānta, according as the given year is prior or subsequent to A.D. 496. The quotient gives the moment of the Mesha-saṅkrānti by the Sūrya-Siddhānta.

c7.42

¹ See Art. 21, and the first footnote appended to it.

c7.43

TABLE

c7.44

Shewing the difference between the moments of the Mesha-saṅkrānti as calculated by the Present Sūrya and the first Ārya-Siddhāntas; the difference in A.D. 496 (Śaka 496 current) being 0.

c7.45

| No. of years. | Difference Expressed in | No. of years. | Difference Expressed in | No. of years. | Difference Expressed in |
| | gh. | pa. | minutes. | | gh. | pa. | minutes. | | gh. | pa. | minutes. |
| 1 | 0 | 0.3 | 0.1 | 10 | 0 | 2.7 | 1.1 | 100 | 0 | 27.3 | 10.9 |
| 2 | 0 | 0.5 | 0.2 | 20 | 0 | 5.5 | 2.2 | 200 | 0 | 54.6 | 21.9 |
| 3 | 0 | 0.8 | 0.3 | 30 | 0 | 8.2 | 3.3 | 300 | 1 | 22.0 | 32.8 |
| 4 | 0 | 1.1 | 0.4 | 40 | 0 | 10.9 | 4.4 | 400 | 1 | 49.3 | 43.7 |
| 5 | 0 | 1.4 | 0.5 | 50 | 0 | 13.7 | 5.5 | 500 | 2 | 16.6 | 54.7 |
| 6 | 0 | 1.6 | 0.7 | 60 | 0 | 16.4 | 6.6 | 600 | 2 | 44.0 | 65.6 |
| 7 | 0 | 1.9 | 0.8 | 70 | 0 | 19.1 | 7.7 | 700 | 3 | 11.3 | 76.5 |
| 8 | 0 | 2.2 | 0.9 | 80 | 0 | 21.9 | 8.7 | 800 | 3 | 38.6 | 87.5 |
| 9 | 0 | 2.5 | 1.0 | 90 | 0 | 24.6 | 9.8 | 900 | 4 | 6.0 | 98.4 |

c7.46

Example. Find the time of the Mesha saṅkrānti by the Sūrya-Siddhānta in A.D. 1000. The difference for (1000—496 =) 504 years is (2 gh. 16.6 pa. + 1.¹ pa. =) 2 gh. 17.7 pa. Adding this to Friday, 22nd March, 42gh. 5pa., *i.e.*, the time fixed by the Ārya-Siddhānta (*Table I.*, cols. 14, 15), we have 44 gh. 22.7 pa. from sunrise on that Friday as the actual time by the *Sūrya-Siddhānta*.

c7.47

97. *Cols. 19 to 25.* The entries in these columns enable us to convert and verify Indian luni-solar dates. They were first calculated, as already stated, according to the Tables published by Prof. Jacobi in the *Indian Antiquary* ¹ (Vol. XVII.). The calculations were not only most carefully made, but every figure was found to be correct by independent test. As now finally issued, however, the figures are those obtained from calculations direct from the *Sūrya-Siddhānta*, specially made by Mr. S. Bālkṛishṇa Dīkshit. The articles *a, b, c*, in cols. 23 to 25 are very important as they form the basis for all calculations of dates demanding an exact result. Their meaning is fully described below (*Art. 102.*).

c7.48

¹ *See note 1 to Art. 91.*

c7.49

The meaning of the phrase "moon's age" (*heading of cols. 21, 22*) in the Nautical Almanack is the mean time in days elapsed since the moon's conjunction with the sun (amāvāsyā, new moon). For our purposes the moon's age is its age in lunation-parts and tithis, and these have been fully explained above.

c7.50

98. The week-day and day of the month A.D. given in cols. 19 and 20 shew the civil day on which Chaitra śukla pratipadā of each year, as an apparent tithi, ends. ² The figures given in cols. 21 to 25 relate to Ujjain mean sunrise on that day.

c7.51

² We have seen before (*Arts. 45 etc. above*) how months and tithis are sometimes added or expunged. Now in case of Chaitra śukla pratipadā being current at sunrise on two successive days, as sometimes happens, the first of these civil days, *i.e.*, the day previous to that given by us, is taken as the first day of the Indian luni-solar year (*see Art. 52*). This does not, however, create any confusion in our method C since the quantities given in cols. 23 to 25 are correct for the day and time for which they are given; while as for our methods A and B, the day noted by us is more convenient.

c7.52

99 When an intercalary Chaitra occurs by the true system (Arts. 45 etc. above) it must be remembered that the entries in cols. 19 to 25 are for the śukla-pratipadâ of the intercalated, not the true, Chaitra.

c7.53

100. The first tithi of the year (Chaitra śukla pratipadâ) in Table I., cols. 19 to 25, is taken as an apparent, not mean, tithi, which practice conforms to that of the ordinary native pañchângs. By this system, as worked out according to our methods A and B, the English equivalents of all subsequent tithis will be found as often correct as if the first had been taken as a mean tithi;—probably more often.

c7.54

101. The figures given in cols. 21 and 22, except in those cases where a minus sign is found prefixed (e.g., Kali 4074 current), constitute a first approximation showing how much of chaitra śukla pratipadâ had expired on the occurrence of mean sunrise at Ujjain on the day given in cols. 19 and 20. Col. 21 gives the expired lunation-parts or tithi-index, and col. 22 shews the same period in tithi-parts, i.e., decimals of a tithi. The meaning of both of these is explained above (Arts. 80 and 81). We differ from the ordinary pañchângs in one respect, viz., that while they give the portion of the tithi which has to run after mean sunrise, we have given, as in some ways more convenient, the portion already elapsed at sunrise. Thus, the entry 286 in col. 21 means that 286 lunation-parts of Chaitra śukla 1st had expired at mean sunrise. The new moon therefore took place 286 lunation-parts before mean sunrise, and by Table X., col. 3, 286 lunation-parts are equal to (14 h. 10 m. + 6 h. 6 m. =) 20 h. 16 m. The new moon therefore took place 20 h. 16 m. before sunrise, or at 9.44 a.m. on the previous day by European reckoning. The ending-moment of Chaitra śukla pratipadâ can be calculated in the same way, remembering that there are 333 lunation-parts to a tithi.

c7.55

We allude in the last paragraph to those entries in cols. 21 and 22 which stand with a minus sign prefixed. Their meaning is as follows:—Just as other tithis have sometimes to be expunged so it occasionally happens that Chaitra śukla 1st has to be expunged. In other words, the last tithi of Phâlguna, or the tithi called amâvâsyâ, is current at sunrise on one civil day and the 2nd tithi of Chaitra (Chaitra śukla dvitîyâ) at sunrise on the following civil day. In such a case the first of these is the civil day corresponding to Chaitra śukla 1st; and accordingly we give this civil day in cols. 19 and 20. But since the amâvâsyâ-tithi (the last tithi of Phâlguna) was actually current at sunrise on that civil day we give in cols. 21 and 22 the lunation-parts and tithi-parts of the amâvâsyâ-tithi which have to run after sunrise with a minus sign prefixed to them. Thus, "—12" in col. 21 means that the tithi-index at sunrise was 10,000—12 = or 9988, and that the amâvâsyâ-tithi (Phâlguna Kṛiṣhṇa 15 or 30) (Table VIII., col. 3) will end 12 lunation-parts after sunrise, while the next tithi will end 333 lunation-parts after that.

c7.56

¹⁰² (a, b. c, cols. 23, 24, 25). The moment of any new moon, or that moment in each lunation when the sun and moon are nearest together, in other words when the longitudes of the sun and moon are equal, cannot be ascertained without fixing the following three elements,—(a) The eastward distance of the moon from the sun in mean longitude, (b) the moon's mean anomaly (Art. 15 and note), which is here taken to be her distance from her perigee in mean longitude, (c) the sun's mean anomaly, or his distance from his perigee in mean longitude. And thus our "a", "b", "c", have the above meanings; "a" being expressed in 10,000ths of a circle reduced by 200.6 for purposes of convenience of use, all calculations being then additive, "b" and "c" being given in 1000ths of the circle. To take an example. At Ujjain mean sunrise on Chaitra śukla pratipadâ of the Kali year 3402 (Friday, 8th March, A.D. 300), the mean longitudes calculated direct from the Sûrya-Siddhânta were as follow: The sun, 349° 22' 27".92.

c7.57

The sun's perigee, 257° 14' 22".86. The moon, 355° 55' 35".32. The moon's perigee, 33° 39' 58".03. The moon's distance from the sun therefore was (355° 55' 35".32—349° 22' 27".92 =) 6° 33' 7".4 = .0182 of the orbit of 360°. This (1.0182) reduced by 0.0200,6 comes to 0.99814; and consequently "a" for that moment is 9981.41. The moon's mean anomaly "b" was (355° 55' 35".32—33° 39' 58".03 =) 322° 15' 37".29 = 895.17. And the sun's mean anomaly "c" was (349° 22' 27".92—257° 14' 22".86 =) 92° 8' 5".06 = 255.93. 1 We therefore give a = 9981, b = 895, c = 256. The figures for any other year can if necessary be calculated from the following Table, which represents the motion. The increase in a, b, c, for the several lengths of the luni-solar year and for 1 day, is given under their respective heads; the figures in brackets in the first column representing the day of the week, and the first figures the number of days in the year.

c7.58

Increase of a, b, c, in one year, and in one day.

c7.59

| Number of days in the year. | a. | b. without bīja. | b. with bīja. | c. |
| 354(4) | 9875.703337 | 847.2197487 | 847.220646 | 969.1758567 |
| 355(5) | 214.335267 | 888.5113299 | 883.512230 | 971.9136416 |
| 383(5) | 9696.029305 | 899.675604 | 899.676575 | 48.57161909 |
| 384(6) | 34.661235 | 935.967185 | 935.968158 | 51.3094039 |
| 385(0) | 373.293166 | 972.258766 | 972.259742 | 54.04789 |
| 1(1) | 338.63193033 | 36.291581211 | 36.291583746 | 2.737784906 |

c7.60

103. Table II., Part i., of this table will speak for itself (see also Art. 51 above). In the second part is given, in the first five columns, the correspondence of a cycle of twelve lunar months of a number of different eras with the twelve lunar months of the Śaka year 1000, 2 which itself corresponds exactly with Kali 4179, Chaitrâdi Vikrama 1135, and Gupta 738. Cols. 8 to 13 give a similar concurrence of months of the solar year Śaka 1000. The concurrence of parts of solar months and of parts of the European months with the luni-solar months is given in cols. 6 and 7, and of the same parts with the solar months in cols. 14 and 15. Thus, the luni-solar amânta month Âshâḍha of the Chaitrâdi Saka year 1000 corresponds with amânta Âshâḍha of Kali 4179, of Chaitrâdi Vikrama 1135, and of the Gupta era 758; of the Âshâḍhâdi Vikrama 1135, and of the Chedi or Kaḷachuri 828; of the Kârttikâdi Vikrama year 1134, and of the Nêvâr year 198. Parts of the solar months Mithuna and Karka, and parts of June and July of 1077 A.D. correspond with it; in some years parts of the other two Christian months noted in col. 7 will correspond with it. In the year Śaka 1000, taken as a Meshâdi solar year, the month Simha corresponds with the Bengali Bhâdrapada and the Tamil Âvaṇi of the Meshâdi Kali 4179, and Meshâdi Vikrama 1135; with Chingam of the South Malayâḷam Siṁhâdi Kollam âṇḍu 253, and of the North Malayâḷam Kanyâdi Kollam âṇḍu 252. Parts of the lunar months Śrâvaṇa and Bhâdrapada correspond with it, as well as parts of July and August of the European year 1077 A.D.; in some years parts of August and September will correspond with it.

c7.61

¹ Calculating by Prof. Jacobi's Tables, a, b, c, are 9980, 896 and 255, each of which is wrong by 1. The above figures were submitted by me to Dr. Downing of the Nautical Almanac office, with a request that he would test the results by scientific European methods. In reply he gave me the following quantities, for the sun from Leverrier's Tables, and and for the moon from Hansen's Tables (for the epoch A.D. 300, March 8th, 6 am., for the meridian of Ujjain). Mean long of sun 345° 51' 47".7, Do. of sun's perigee 253° 54' 58".5, Do. of moon 353° 0' 36".0, Do. of moon's perigee 36° 9' 48".4. He also verified the statement that the sunrise on the morning of March 8th was that immediately following new moon. The difference in result is partly caused by the fact that Leverrier's and Hansen's longitudes are tropical, and those of the Sûrya-Siddhânta sidereal. Comparing the two results we find a difference of 0° 35' 40".9 in "a", 5° 24' 49".69 in "b", 0° 11' 15".87 in "c". The closeness of the results obtained from the use of (1) purely Hindu (2) purely European methods is remarkable. Our Tables being for Indian documents and inscriptions we of course work by the former. [R. S.]

c7.62

² This year Śaka 1000 is chosen for convenience of addition or substraction when calculating other years, and therefore we have not taken into account the fact that Ś 1000 was really an intercalary year, having both an Adhika Jyeshtha and a Nija Jyeshtha month. That peculiarity affects only that one year and not the concurrence of months of previous or subsequent years in other eras.

c7.63

All the years in this Table are current years, and all the lunar months are amânta.

c7.64

It will be noticed that the Tuḷu names of lunar months and the Tamil and Tinnevelly names of solar months are corruptions of the original Sanskrit names of lunar months; while the north and south Malayâḷam names of solar months are corruptions of the original Sanskrit sign-names. Corruptions differing from these are likely to be found in many parts of India. In the Tamil Districts and the district of Tinnevelly the solar sign-names are also in use in some places.

c7.65

104. Table II., Part iii. This portion of the Table, when read with the notes printed below would seem to be simple and easy to be understood, but to make it still clearer we give the following rules:—

c7.66

I. Rule for turning into a Chaitrâdi or Meshâdi year (for example, into a luni-solar Śaka, or solar Śaka, year) a year of another era, whether earlier or later, which is non-Chaitrâdi or non-Meshâdi.

c7.67

(a) For an earlier era. When the given date falls between the first moment of Chaitra or Mesha and the first moment of the month in which, as shewn by the heading, the year of the given earlier era begins, subtract from the given year the first, otherwise the second, of the double figures given under the heading of the earlier era along the line of the year o of the required Chaitrâdi or Meshâdi era (e.g., the Śaka).

c7.68

Examples. (1) To turn Vaiśâkha Śukla 1st of the Âshâḍhâdi Vikrama year 1837, or Śrâvaṇa śukla 1st of the Kârttikâdi Vikrama year 1837 into corresponding Śaka reckoning. The year is (1837—134 =) 1703 Śaka. The day and month are the same in each case. (2) To turn Mâgha śukla 1st of the Kârttikâdi Vikrama samvat 1838 into the corresponding Śaka date. The year is (1838—135 =) 1703 Śaka. The day and month are the same. (3) Given 1st December, 1822 A.D. The year is (1822—77 =) 1745 Śaka current. (4) Given 2nd January, 1823 A.D. The year is (1823—78 =) 1745 Śaka current.

c7.69

(b) For a later era. When the given day falls between the first moment of Chaitra or Mesha and the first moment of the month in which, as shewn by the heading, the later era begins, add to the number of the given year the figure in the Table under the heading of the required Chaitrâdi or Meshâdi era along the line of the year o/1 of the given later era. In the reverse case add that number reduced by one.

c7.70

Examples. (1) To turn the 1st day of Mithuna 1061 of the South Malayâḷam Kollam Âṇḍu into the corresponding Śaka date. The year is (1061 + 748 =) Śaka 1809 current. The day and month are the same. (2) To turn the 1st day of Makara 1062 of the South Malayâḷam Kollum Âṇḍu into the corresponding Śaka date. The year is (1062 + 747 =) 1809 Śaka current. The day and month are the same.

c7.71

II. Rule for turning a Chaitrâdi or Meshâdi (e.g., a Śaka) year into a non-Chaitrâdi or non-Meshâdi year of an earlier or later era.

c7.72

(a) For an earlier era. When the given day falls between the first moment of Chaitra or Mesha and the first moment of the month in which, as shown by the heading, the year of the earlier era begins, add to the given Chaitrâdi or Meshâdi year the first, otherwise the second, of the double figures given under the heading of the earlier era along the line of the year 0 of the Chaitrâdi or Meshâdi era given.

c7.73

Examples. (1) To turn Bhâdrapada kṛishṇa 30th of the Śaka year 1699 into the corresponding Kârttikâdi Vikrama year. The year is (1699 + 134 =) 1833 of the Kârttikâdi Vikrama era. The day and month are the same. (2) To turn the same Bhâdrapada kṛishṇa 30th, Śaka 1699, into the corresponding Âshâḍhâdi Vikrama year. The year is (1699 + 135 =) 1834 of the Âshâḍhâdi Vikrama era. The day and month are the same.

c7.74

(b) For a later era. When the given day falls between the first moment of Chaitra or Mesha and the first moment of the month in which, as shown by the heading, the later era begins, subtract from the given year the number under the heading of the given Chaitrâdi or Meshâdi era along the line of the year 0/1 of the given later era; in the reverse case subtract that number reduced by one.

c7.75

Examples. (1) To turn the 20th day of Simha Śaka 1727 current into the corresponding North Malayâḷam Kollam Aṇḍu date. The day and month are the same. The era is a Kanyâdi era, and therefore the required year is (1727—748 =) 979 of the required era. (2) To turn the 20th day of Simha Śaka 1727 current into the corresponding South Malayâḷam (Tinnevelly) Kollam Aṇḍu date. The day and month are the same. The era is Simhâdi, and therefore the required year is (1727—747 =) 980 of the required era.

c7.76

III. Rule for turning a year of one Chaitrâdi or Meshâdi era into one of another Chaitrâdi or Meshâdi era. This is obviously so simple that no explanations or examples are required.

c7.77

IV. Rule for turning a year of a non-Chaitrâdi or non-Meshâdi era into one of another year equally non-Chaitrâdi or non-Meshâdi These are not required for our methods, but if any reader is curious he can easily do it for himself.

c7.78

This Table must be used for all our three methods of conversion of dates.

c7.79

105. Table III.—The numbers given in columns 3a and 10 are intended for use when calculation is made approximately by means of our method “B” (Arts. 137, 138).

c7.80

It will be observed that the number of days in lunar months given in col. 3a is alternately 30 and 29; but such is not always the case in actual fact. In all the twelve months it occurs that the number of days is sometimes 29 and sometimes 30. Thus Bhâdrapada has by our Table 29 days, whereas it will be seen from the pañchâṅg extract printed in Art. 30 above that in A.D. 1894 (Śaka 1816 expired) it had 30 days.

c7.81

The numbers given in col. 10 also are only approximate, as will be seen by comparing them with those given in cols. 6 to 9.

c7.82

Thus all calculations made by use of cols. 3a and 10 will be sometimes wrong by a day. This is unavoidable, since the condition of things changes every year, so that no single Table can be positively accurate in this respect; but, other elements of the date being certain, calculations so made will only be wrong by one day, and if the week-day is given in the document or inscription concerned the date may be fixed with a fair pretence to accuracy. If entire accuracy is demanded, our method “C” must be followed. (See Arts. 2 and 126.)

c7.83

The details in cols. 3, and 6 to 9, are exactly accurate to the unit of a pala, or 24 seconds. The figure in brackets, or week-day index (w), is the remainder after casting out sevens from the number of days; thus, casting out sevens from 30 the remainder is 2, and this is the (w) for 30. To guard against mistakes it may be mentioned that the figure “2” does not of course mean that the Mesha or Vṛishabha saṅkrânti always takes place on (2) Monday.

c7.84

106. Tables IV. and V. These tables give the value of (w) (week-day) and (a) (b) and (c) for any required number of civil days, hours, and minutes, according to the Sûrya Siddhânta. It will be seen that the figures given in these Tables are calculated by the value for one day given in Art. 102.

c7.85

Table IV. is Prof. Jacobi's Indian Antiquary (Vol. XVII.) Table 7, slightly modified to suit our purposes; the days being run on instead of being divided into months, and the figures being given for the end of each period of 24 hours, instead of at its commencement. Table V. is Prof. Jacobi's Table 8.

c7.86

107. Tables VI. and VII. These are Prof. Jacobi's Tables 9 and 10 re-arranged. It will be well that their meaning and use should be understood before the reader undertakes computations according to our method "C". It will be observed that the centre column of each column-triplet gives a figure constituting the equation for each figure of the argument from 0 to 1000, the centre figure corresponding to either of the figures to right or left. These last are given only in periods of 10 for convenience, an auxiliary Table being added to enable the proper equation to be determined for all arguments. Table VI. gives the lunar equation of the centre, Table VII. the solar equation of the centre. (Art. 15 note 3 above). The argument-figures are expressed in 1000ths of the circle, while the equation-figures are expressed in 10,000ths to correspond with the figures of our "a," to which they have to be added. Our (b) and (c) give the mean anomaly of the moon and sun for any moment, (a) being the mean longitudinal distance of the moon from the sun. To convert this last (a) into true longitudinal distance the equation of the centre for both moon and sun must be discovered and applied to (a) and these Tables give the requisite quantities. The case may perhaps be better understood if more simply explained. The moon and earth are constantly in motion in their orbits, and for calculation of a tithi we have to ascertain their relative positions with regard to the sun. Now supposing a railway train runs from one station to another twenty miles off in an hour. The average rate of running will be twenty miles an hour, but the actual speed will vary, being slower at starting and stopping than in the middle. Thus at the end of the first quarter of an hour it will not be quite five miles from the start, but some little distance short of this, say m yards. This distance is made up as full speed is acquired, and after three-quarters of an hour the train will be rather more than 15 miles from the start, since the speed will be slackened in approaching the station,—say n yards more than the 15 miles. These distances of m yards and n yards, the one in defect and the other in excess, correspond to the "Equation of the Centre" in planetary motion. The planetary motions are not uniform and a planet is thus sometimes behind, sometimes in front of, its mean or average place. To get the true longitude we must apply to the mean longitude the equation of the centre. And this last for both sun (or earth) and moon is what we give in these two Tables. All the requisite data for calculating the mean anomalies of the sun and moon, and the equations of the centre for each planet, are given in the Indian Siddhântas and Karaṇas, the details being obtained from actual observation; and since our Tables generally are worked according to the Sûrya Siddhânta, we have given in Tables VI. and VII. the equations of the centre by that authority.

First 85 of 108 passages. Read on

Cites (4)