in his ante-natal spiritual state, after he had ceased to be a mere Fravashi, or primary idea, and had become an intelligent, moving, and personal existence, but still a spirit. While the later Zoroaster of Pliny, who flourished before the fifth century B.C. must have lived about the same time as the same traditional Zaratüst after he came into the worldly existence, and may reasonably be identified with him, although Pliny had little information to give about him.
76. It will be noticed that this explanation depends entirely upon the peculiarly artificial system of the tra- ditional chronology, in which the whole of time is assumed to consist of twelve millenniums devoted to different pur- poses; and if this particular system had not been in use at the time the statements, quoted by Pliny and Diogenes Laertius, were made, those statements could not have been explained as referring to the same individual. But if they do not refer to the same individual, we have only the options of rejecting all the statements, or believing an im- possible date to be literally correct; neither of which decisions would be altogether satisfactory to a judicious mind. The only reasonable conclusion seems to be that the chronology based upon the twelve millenniums was in use in the fifth century B.C., about which time the earliest quoted statement seems to have been made.
77. It will also be observed that this millennial chronology is inextricably associated with the idea of the primeval existence of all good creations in the state of Fravashi.. These are described as spiritual existences who remained three millenniums unthinking, unmoving, and intangible (Bd. I, 8); and the next three millenniums they still re- mained undisturbed by evil, mankind being represented, for that period, by Gayémard in the world (Bd. XXXIV, 1) and by the spiritual form of Zaratüst in heaven (Dk. VII, ii, 15), while the animals were symbolized by the primeval ox for the same period. Six millenniums, which are half the duration of time, were thus appropriated to Fravashis, spiritual and embodied, probably before the birth of Plato, if we may rely upon classical statements ; and it must have
xlii PAHLAVI TEXTS.
been before this date that the series of millenniums was
arranged for all future history, till Time was expected to
merge once more into Eternity at the renovation of the
universe.
78. As we have seen that Zoroastrian tradition is very consistent in fixing the date of Zaratüst's activity about the end of the seventh century B.C., it may be asked, why have Avesta scholars so strongly insisted upon its greater antiquity? They may have had several reasons, but three, at least, were important. First, they had the classical statements which, as we have seen ($ 71), generally placed Zaratüst as far back as the seventh millennium B.C., on the testimony of persons who lived from two to four centuries after the traditional date of Zaratüst's death. No one, of course, could believe in the literal accuracy of the number of millenniums, which referred, as we have seen, to an imaginary period of spiritual existence, but this number was considered merely as an exaggeration which might be reduced to any amount that seemed reasonable. At the same time, this evidence for antiquity was quite sufficient, in the second place, to discredit the traditional date, of which these old authorities seemed ignorant, though it was a period then comparatively recent. And, if this discredit had not been sufficient to shake the faith of Avesta scholars in the traditional date of Zaratüst, they still had a taird reason for their scepticism, when they discovered that the language of the Avesta was not merely a sister of Sanskrit, but that a large portion of it was sister to the oldest Sanskrit with which they were acquainted, and which appeared to them certainly older than the time of Gautama Buddha, who lived about one generation later than the traditional Zaratüst.
79. How far Avesta scholars were justified in their con- clusions must be left for future ages to determine; at present we have no really historical information about the origin of Zoroastrianism, and must still consider it as decidedly prehistoric; though, it may be admitted that the Parsi calendar, as used in Persia, so far agrees with tradition, that it still bears witness to its own original institution in
INTRODUCTION. xliii
the reign of Darius Hystaspes, as will be seen from the following details.
80. The Parsi year consists of twelve months, each con- sisting of thirty days, with five additional days added at the end of the year, and the total number of 365 days never varies, so that, whenever a leap-year occurs in our calendar, the beginning of the Parsi year retreats one day in the Christian calendar. In this manner, the total number of days which the beginning of the Parsi year has retreated, since the institution of their calendar, records the number of leap-years which would have occurred in the same period of Christian years, if the regular leap-years had existed the whole time; and four times the number of leap-years would be the total number of years. But as leap-years have not been used the whole time, we have to calculate from astronomical data.
81. In the first place, we want to know at what season the Parsi year originally began, and we learn this from Bd. XXV, 7, 20, where we are told that the winter of the rectified year ends with the five extra days, and the spring begins with the first month ; which means that the rectified year begins with the vernal equinox. We also have to observe that, retreating at the rate of one day every four years, the beginning of the year retreats all round the year in 1460 years ; and we know from general history that the period, with which we have to deal, is much more than 1460 years and less than 2920. Then we have to ascertain the exact length of the tropical year, which astronomers say is 365:2422 days, with an infinitesimal decrement, quite inappreciable in the period we have in view.
82. We may calculate back from any vernal equinox which occurs not too far from noon, say that of March 22, 1865, when the beginning of the Parsi year, according to Persian reckoning, had retreated to August 24, 1864, or 210 days, in addition to a previous retreat of a whole year of 365 days, or altogether 575 days since the establishment of the calendar. So that the difference between the Parsi year of 365 days and the correct tropical year of 365-2422 days had then accumulated to a total of 57 5 days. Divid-
xliv PAHLAVI TEXTS.
ing the accumulated error of 575 days by the annual in- crement of 0-2422 of a day, we obtain a quotient of 2374 years, including A.D. 1865, as the time in which this error had accumulated, and this carries us back to B.C. 510 as an approximate date of the establishment of the Parsi calendar, with the first day of the year coinciding with the vernal equinox.
83. This date is, however, liable to some modifications for errors of observation on the part of the ancient astronomers, one of which errors, being constantly in one direction, must be taken into account. These old observers were not aware of the effect of refraction, which always makes the night seem somewhat shorter than it is in reality ; and this would lead them to antedate the vernal equinox by rather more than a day; so that they would observe an apparent equinox in B.C. 505 on the same day in the Parsi year as that on which the real equinox occurred in B.C. 510. The most probable date of the establishment of the Parsi calendar is therefore B.C. 505, with a margin of four to eight years in either direction for accidental errors of observation.
84. A few years before this period we know, from the cuneiform inscriptions of Behistün, that Darius Hystaspes used an older calendar, when recording his early victories over insurgents, which consolidated his empire. It was a time when he was introducing many reforms in the government, and, being a believer in Auramazda, his most induential advisers would probably be Zoroastrian priests. If they thought it necessary to reform the old calendar, the adoption of strictly Zoroastrian names for the new months and days in the Parsi calendar would be fully explained.
85. But, besides this ordinary civil calendar, in which new-year's day was constantly retreating, the Persians had a rectified calendar for religious purposes, which intercalated an extra month from time to time, for the purpose of bringing new-year's day forward again to the vernal equinox, and restoring the festivals to their proper seasons. It is this calendar which is used in Bd. XXV, and its days
INTRODUCTION. xlv
and months are distinguished by the epithet vehifakig
(veh+i£+ak+ig), ‘belonging to what is really good,’ or
vehi&ak, ‘belonging to the really good,’ which, in this
connection, may be best expressed by the word ‘ rectified.’
86. This intercalation is described by Albirüni in various passages! which inform us that, after the new-year's day had retreated more than a month from the vernal equinox, the king would order the priests to arrange for the solemn proclamation of an extra month to be intercalated, between the last month of the year and the five extra days, by merely moving those five days from the end of the twelfth month to the end of the first month of the next year. The effect of this was to put an extra month into the earlier year which, beginning with the first month, would also end with the first month augmented by the five extra days as the usual termination of the year. All following years would begin with the second month, and end with the first month and the five extra days, until the second intercala- tion, when a year of thirteen months and five days would be again obtained, by merely moving the. five extra days to the end of the second month which would thus become the last month of the year, while the third month would begin the year until the third intercalation. By these means, any number of intercalations could be made without any additional month being named, and the position of the five extra days always indicated the end of the rectified year, and that the rectified first month, which followed them, was to become the last month of the preceding year at the next rectification, or intercalation.
87. If the Parsi calendar, as used in Persia, were es- tablished B.C. 505, as we have calculated, it ought to have been rectified by an intercalation of one month about each of the following years :—B.C. 381, 257, 133, 10, A.D. 115, 239, 363, 487, 610. Albirdni (p. 121), however, has re- corded only one intercalation of two months in the time of Yazdakard I (A.D. 399-420), son of Shahpühar, when the
1 Sachau's Albírüni's Chronology of Ancient Nations, pp. 12, 13, 38, 53-56, 121, 184, 185, 220, 221.
xlvi PAHLAVI TEXTS.
five extra days were removed to the end of the original eighth month Ábán, where they remained until Albirüni's own time (Alb. p. 56), about A.D. 1000. The reason for intercalating two months at once, was because the time for the seventh intercalation (A.D. 363) was already long past ; so the eighth was added three or four score years in advance, being due in 487.
88. All that Albirüni says about this double intercalation is quite in accordance with the original establishment of the calendar by Darius Hystaspes, and would render any date more than thirty-seven years later than his reign im- possible’. With regard to the earlier intercalations (which must have occurred to account for the movement of the five extra days) that of A. D. 115 was neither in the reign of Vologeses I, nor in that of Vologeses III, one of whom must have been the Askánian renovator of the Avesta. That of A.D. 239, if carried out punctually, would have been at the extreme end of the reign of Ardashir ; but the intercalations seem to have been usually delayed, as in the case of that of 363 which was delayed for thirty to fifty years, although it ought to have been carried out under the direction of one of those ultra-orthodox high-priests, Átürpád son of Máraspend, or his son Zaratüst, in the reign of Shahpühar II.
