That the two natural units of day and night, distinguished by light and darkness, was emerging, however faintly, into a conventional unit of a ‘day’ as a single concept including both periods, to function as the primary unit of time for calculating the month and the year, is evidenced in the Pâyâsi Suttana passage already cited where rattin-diva seems to have been used as a single unit. However, for all ordinary purposes, the day and the night seem to have remained distinct, with the day divided into three periods and the night subdivided into three watches [yâma].
This natural day apparently commenced with the sunrise [aruna, paccusasamayavin.iii, p.204; Si.p.107; it.p.20] and, although there does not appear to have been any fixed time which marked the beginning of the day except the rise of the sun itself. [e.g., S.i, p.107] the context of a reference [it.p.20] to the osadhi taraka and the description of its brilliance at dawn [Mii.p.34] gives room for the conjecture that this star may have sometime, served as the signal of the morn. Thoma [ERE.Vol.12, p.72], identifies the osadhi taraka with the planet Venus, substantiating his position by the evidence of the Sanskrit recension of a Digha passage in the Mahâvyutpatti  which substitutes Usanas one of the Sanskrit names for Venus, for osadhi taraka.
The end of the day was the natural phenomenon of the sunset [vin.ic.p.273], the evening being referred to usually as the Sayanha or syam, but again having no fixed time to demarcate it, the latter word sometimes being used to define the time even between mid day and sunset. The Buddha is often reported as opening the day with robbing himself and as departing for alms [D.i,p.226], of concluding the natural day with a break in his meditations [S.i, p.146] and of retiring for a noon-day seclusion [divavihara] which is also the practice of his disciples [S.i, pp.129, 132]. The mid-day or noon [majjhana] was obviously marked by the sun being right overhead or “right up” and not only bisected the day into forenoon [pubbanha] and afternoon [aparanha, sayanha] but also became the standard by which the time of the day was measured, the length of the shadow being usually meted by finger-breaths [vin.ii, pp.229, 294] Thus, when majjima was also taken as a unit, the day was considered to be of three divisions. Precision and care in noting the exact time of the moon seems to have engaged the greater monastic attention by the ascendance of the concept of wrong time. [vikala: Vin, ii, p.300; iv, p.84f] which spring up mainly in association with the Buddha’s injunction to the monks against taking of meals after the turn of the shadow, i.e., after noon till the sunrise of the following day [virato vikalabhojana: D.i, p64]. In fact, the Vinaya [ii, p.299] refers to a clarification sought by certain monks as to whether it was permissible to take the mid-day meal once the shadow has passed two finger breaths. Majjhanha, because, its passing could result in either an offence against the monastic discipline or a meal-less day, seems thus to have become a matter of practical importance as far as reckoning of time was concerned, so that in fact, the latter part of the day came to be referred to in terms of the mid-day meal-time as pacchabhatta [literally, after the meal] which, though technically synonymous with sayanha, was probably used to refer to the earlier part of the afternoon, sayanha being applied more to the later part of it. It may be added that purebhattam was similarly used to refer to the forenoon.
The conventional day seems to have been divided into smaller units of time. A passage in the Anguttara Nikaya [A.v, p.137] refers to moments, instants and seconds, the passage concerned running:
“Just as a mountain river, winding here and there, swiftly flowing, taking all along with it, never for a moment or an instant or a second [natthi so khano va layo va muhutto va] pauses, even so is the life of man.”[A.iv, p.92].
Here it appears that the words khana and muhutta are used in the Pali texts to refer to infinitesimal portions of time, perhaps no very different from each other. [Dhp.vv.65, 239]. However, a different system of reckoning seems to give muhurta as one-thirtieth portion of the 24-hour day. [Divy.p.643, 5ff] It is probably the traditional in the Divyavadana that influenced the traveller Hsuan Tsang, who, sojourning in India in the 7th century A.D., records the contents of these three notions and their inter-relationships, thus:
“The shortest portion of time is called a t’sa-na [kshana]; 120 kshana make a ta-t’sa-na [takshana]; 60 of these make a la-fo [lava]; 30 of these make a mau-hu-li-to [muhu-rta]; five of these make a ‘period of time,’ [kala]; six of these make a day and night [ahoratra], but commonly the day and night are divided into eight kalas” [trasl. Beal, Chinese Account of India].
That the muhutta was linked up with the day is also witnessed by the Paramatthadipani [iii, p.198] so that we may, in terms of Hsuan Tsang, deduce that 30 muhuttas made a day-and-night and a khana was the 7080000th part of that day.