There seems to be sufficient evidence to believe that by the time the earliest Buddhist scriptures came into being there was already current in the immediate social environment of the Buddhists a system of dividing time into periods of years, months and days. That this system might have been of considerably long standing may be inferred from the fact that the oldest known of Indian literature, the Rgveda had already these three divisions calculated in terms of numerical units, the year consisting of 12 months of 30 days each, totalling up to 360 days per year.

Griffith in the Hymns of the Rgveda translates the relevant lines:

"Formed with twelve spokes, by length of time unweakned,
Rolls round the heavens these wheels of during order.
Herein established, joined in pairs together
Seven hundred sons and twenty stand" [Rv.i].

The metaphor of ‘the spokes’ obviously alludes to the twelve months of the year. So does the metaphor of the ‘seven hundred and twenty sons’ to the number of days contained in one year, the phrase ‘joined in pairs ‘ suggesting a calculation of two sons per day which makes the number of days 360. A similar metaphorical allusion is made in the lines:

Twelve are the fellers, an the wheel is single,
There are the naves, what man has understood it?
There in are set together three hundred and sixty
Which in no wise can be loosened" [Rv.1,I,164,48]

Neither the word ‘month’[masa] nor ‘year’ [samvastara] are made direct reference in this context, but both of them occur frequently in the Rgveda, the former, for instance, in the lines "preserve us through many months and autumns" [Rv.1.vii,912] While later occurs in Rv.1,i.110, 4; 140,2 etc.].

Even if Rgvedic knowledge remained esoteric, that knowledge would, in its application have percolated down to the ordinary society with which the Buddhists were in contact and from which they would have derived their own methods and practices in time-reckoning. A division of the year into seasons had also naturally been observed and come into pre-Buddhistic literature and the attempt, revealed in the Rgveda itself, to adjust the lunar and solar years by the intercalation of a "later-born" month discloses that the need was already being felt for a conventional calendar in preference to a natural one. These developments and concepts constitute the background to the emergence and evolution of calendrical notions among the Buddhists.

Most of the references in early Buddhist literature, particularly the canonical, to calendrical time-divisions seem to be incidental and fragmentary. Such a reference, which, inspite of being incidental, still enables us to infer the inter-relations between units of time-reckoning in vogue, is in the Digha -Nikaya [D.ii, p.327], which speaks of 100 years, each year containing 12 months of 30 days each, totalling up to 360 days. This passage in the Payasi Suttanta refers to an event probably posthumous to the Buddha, but in all likelihood not later than 50 years from his demise. It places in the mouth of an immediate disciple, Kumara Kassapa, a comparison of human time-divisions with those believed to be obtaining in the celestial word of the Tavatimsa gods, which runs as follows:

"That which in human reckoning, is a century [vassa satam], this to be thirty-three gods, is one ‘night and day’ [rattin-diva]. Of such nights, thirty constitute the month [mass] and of such months, twelve make the year [sammavacchara]."

It is too obvious to point out that the comparison is with the time-divisions obtaining in the India known to Kumara Kassapa and his contemporary Buddhists. There is, however, a similar incidental reference in an Anguttara -Nikaya [A, iv, p.252] passage which places an almost identical comparison of time-divisions, this time with several celestial worlds, in the mouth of the Buddha which indicates that Kumara Kassapa was expressing a view already expressed by the Buddha himself, and that the former probably derived his information from that source. It is therefore, possible to conclude that these time divisions were already accepted by the Buddhist community during the Buddha’s lifetime and that the Buddha made no alteration to the prevailing calendrical notions of the day but merely wished to accept them for his purposes. The Buddhist acceptance, then, was that basically the year was divided into 12 months of 30 days each, i.e., the time-divisions already observed to be accepted as far back as the Rgveda.

It is obvious that these three time-divisions were based on the observation of natural phenomena, the apparent motion of the sun, noon and stars as were the Vedic time-divisions. That the Buddhists were conscious of this connection is supported by a statement in the Agganna Suttanta [D.iii, p.90f], which proposes to explain the origin, among other things of the time-divisions adopted by man. Here, the first inhabitants of the earth, the Abhassara, having become earth-bound, after tasting of the earth, bemoan the loss of their self-luminance, thus:

"As we did so, our self-luminance vanished away. When it was gone, moon and sun became manifest. When the constellations [nakkhattani tarakarupani] became manifest. When the constellations became manifest, night and day were perceived. When night and day were perceived, months and half-months were perceived, seasons and year were perceived."

It will be observed that this passage not only connects the calendrical time divisions with the natural movements associated with the moon and sun, but also notes the connection of the time divisions with the constellations. Not only are the phases of the moon taken account of but also its journey through out the year. The passage also adds two more natural divisions to the three mentioned in the Payasi Suttanta, namely, the half-month natural divisions to the three mentioned [addhamasa] and the seasons [utu]. The tendency towards a lunar bias in Buddhist time-reckoning, already seen, is thus further confirmed, particularly by the admission of the fortnight as a significant time unit as well as the noting of the moon’s annual journey.

Nevertheless, the solar factor too, seems to have been taken account of, as is obvious from the passage including the seasons among its time divisions. The seasons, in fact, seem to have played an important part in the Buddhist indications of time, and their distribution of it. Frequent mention is made in the canonical literature, especially for purposes of indicating time, of one or other of the seasons. Winter, summer, autumn as well as a rainy and a dry season are referred to [e.g., D11, p.183, Mi, pp79, 115;Siii, p.155; V.51; A.V.p.27; Sn.v.233; Vin.i.p.253] in this manner, but a conception of three seasons, hot, rainy and wintry is found in the Mahapadana Suttanta [D11, p.21] in connection with the seasonal abodes of Vipassi Buddha where the rainy season is separated out, as the context demands, to be described as consisting of four months. The possible surmise that the other two seasons wee also divided into equal parts of four months is confirmed by an Anguttara Nikaya passage [A, iv, p.138] which gives the exact distribution of the year in terms of the cycle of the seasons, not only into months but also fortnights, days and the number of meal times, thus:

"…though a man live a hundred years, he lives but three hundred seasons - a hundred seasons of winter, a hundred seasons of summer and a hundred seasons of rain.

And though he lives three hundred seasons, he lives but twelve hundred months, four hundred months of winter, four hundred months of summer and four hundred months of rain."[E.M.Hare, Gradual Sayings, iv.p.93].

The description goes on to show, in similar fashion, that these twelve hundred months imply eight hundred fortnights for each season and altogether consisting of six and thirty thousand days during which a man eats seventy two thousand meal. It is noteworthy that the word "samvacchara" is used in this context to stand for the word "year".

It is plain, therefore, that for the Buddhists of that time the year consisted of three seasons of four months each, with each month containing two fortnights together equalling thirty days, each day containing two meal-times. These three seasons were distinguished as the hot [gimha], rainy [vassa] and wintry [hemanta]. There are, however, references in the canonical literature to a forth season, autumn [sarada: D.ii, p.183] as occurring in the last month of the rainy season and is, rather curiously, described as a time when the sun, springing up in the clear cloudless sky "is hard to look upon and destructive to the eye". There are also some casual references to a dry season of six months. [S.111, p.155].

The order in which the three seasons are mentioned is never consistent and varies according to the context, though the tendency appears to have been to name them in order: hemanta, gimha, and vassa. But the main interest of the Buddhists seems to have in the rainy season whose anticipation was necessitate by the disciplinary rules of the Vinaya [Vin.1, p.137] requiring monks to go into retreat during the rainy season. The greatest concern in fact with time-reckoning in terms of months, seems to have been evinced in the Buddhists in connection with the keeping of the rains-residence [vassa], obviously because of the great emphasis laid on it by the Buddha.

The recognition of the cycle of the seasons as constituting a year, i.e., of three seasons of four months as making up a year and the simultaneous reckoning of the year as composed of 12 months based on the phases of the moon, must have undoubtedly beset the Buddhists with the problem of reconciling the solar and lunar years. But there does not appear to be any direct reference in the canonical literature to any form of intercalation, and the Buddhists seem to have overcome the annual variation in the time of the seasons and their discrepancy with the lunar movements, at least when it became a practical problem to them with regard to the keeping of the rains-retreat [vassa] by offering alternative dates for the commencement of the vassa.

There also does not seem to be any reference in the early literature to the solstices, but definite paths for both the sun and the moon are recognised. The Brahmajala Suttanta [Di, p.10 f] is seen to refer to a usual course [Pathgamana] and unusual course [Upathagamana] of the sun and moon, while the Anguttara-Nikaya [A.i, p.75] speaks of the constellations and stars, as well as these two luminaries, as deviating from their normal courses, the deviation being described as putting ‘the days, nights, months and seasons out of joint’. There also seems to have been an awareness of a regularity in the coldest days of the wintry season in the concept of the antaratthaka, the eight days falling in equal divisions of four on either side of the last day of the month of magha. The Buddha speaking of the futile austerities he practiced in his search for truth, makes use of this concept of illustrate how equanimously he faced dire cold of the climate, as he did in tense heat, during his quest [m.1, p.79].  A similar allusion is made to this period of time by Hatthaka of Alavi in a conversation with the Buddha [A.p.i 136] which demonstrates that the word was in common parlance. But the most suggestive evidence is the test carried out by the Buddha at Vesali [latitude 26 N, Longitude 85 E.], when 'in the cold winter nights, during the eights', he tried to determine, before he laid down the Vinaya rule regarding the maximum number of robes permissible for a monk to keep in his possession, the minimum quantity of clothing an ordinary person would require to keep in his possession, the minimum quantity of clothing an ordinary person would require to keep himself warm. [vin.i, p.297 f.; p.31].

To the natural divisions of time discussed above, the early Buddhists add a few conventional divisions, such as the seven days, sattaha [D.ii, p.248; Vin.i, p.85] and the subdivisions of the day and night, but usually, these conventional divisions were themselves directly or indirectly connected with natural occurrences. Thus, for instance, the day was subdivided into two parts by the noon which in turn was determined by the shadow which depended on the sun rising to its zenith [majjhantika]. The ‘week’, as the one-forth part of a month, was obviously connected with the lunar phases rather than being a more conventional division, sattaha or the seven-day unit not necessarily corresponding to this ‘week’. It is noteworthy, in fact, that there is no separate word for "week", the seven-day unit being applicable to any seven days.

The Buddhists also speak of units of time larger, than the year, such as the decade [vassadasa, J.iv, 397], the century [vassam satam, A.iv, p.138; sn.v.589], and the aeon [kappa] the two former; like the week, being designated by no single word but by one which suggests it to be a multiple of the nearest natural division, while the kappa or mahakappa is defined in terms of the asankeyya kappa, four of which go to make it. Thee assankeyya kappa denote ‘durations of time’ in the involution and the evolution of the world. The century as a time unit was probably suggested by the average duration of human life at the time, which was considered to be, roughly, a hundred years. [S.ii, p.95]