To describe the period of 24 hours, regarded as a single unit for purposes of calculation, most modern and also the ancient tongues employ the term that denotes its light part, i. e. ‘day’ etc. Primitive peoples have no term to express this idea and must describe the period by means of expressions equivalent to ‘day and night’, e. g. ‘sun-darkness’ (Malay Archipelago)[2], ‘light and darkness’ (Yukaghir in N. E. Asia)[3]. The day is sometimes described by the concrete phenomenon which it brings, namely the sun. The Bontoc Igorot of north Luzon have the same word for sun as for day, a-qu, and the time is reckoned in suns[4]. The Comanche Indians reckon the days in ‘suns’[5], and in an Indian hieroglyph from the northern shores of Lake Superior the duration of a three days’ journey described is expressed by three circles, i. e. three suns[6]. The western tribe of the Torres Straits reckons time in ‘suns’, i. e. days[7]. We may compare the well-known primitive idea that the sun originates afresh for every new day. The same thing is found in the language of signs. La Billardière in the year 1800 relates of the very low Tasmanians, now long since extinct, that they had some idea of regulating time by the apparent motion of the sun. In order to inform him that they would make a journey in two days, they indicated with their hands the diurnal motion of the sun and expressed the number two by as many of their fingers. This, he asserts, is the only reference that can be found to any knowledge of the movements of the heavenly bodies[8]. So also according to Homfray the natives of the Andamans describe a day by making a circle with the right arm, i. e. a revolution of the sun. We may compare the indication of the time of day by pointing with the hand to the position of the sun, with which we shall shortly have to deal. It is not improbable that the designation of the day by means of an indication of the course of the sun arose in the first place from the indication of the position of that planet. The same method of expression is found in the classical languages as a poetic or hierarchical archaism[9], and also in medieval Latin. But ἥλιος, sol, is also used to denote the yearly revolution of the sun, i. e. a year, and the year is denoted by φάος, lux. Still more striking and more significant for the discontinuous method of reckoning is the Homeric use of ἠώς, ‘dawn’, instead of day, e. g. “this is the twelfth dawn since I came to Ilion”,[10] “this is the twelfth dawn he lies so”,[11] and elsewhere. Aratus follows the Homeric use[12]. The nature of this pars pro toto reckoning will be further explained in the chapter dealing with the year.

The counting of the days from the dawns is unique, and the counting from the day-time is comparatively rare: the Indo-European peoples of olden times, and indeed most of the peoples of the globe, count the days from the nights. For this it will be sufficient to quote Schrader’s statement:—“Moreover it can hardly be necessary to give evidence for this well-known custom of antiquity. In Sanskrit a period of 10 days is called daçarâtrá (:râtrî = ‘night’); nîçanîçam, ‘night by night’ = ‘daily’. ‘Let us celebrate the old nights (days) and the autumns (years)’, says a hymn. In the Avesta the counting from nights (xsap, xsapan, xsapar) is carried out to a still greater extent. As for the Germanic peoples, among whom Tacitus had already observed this custom,[13] we constantly find in ancient German legal documents such phrases as sieben nehte, vierzehn nacht, zu vierzehn nachten. In English fortnight, sennight are in use to-day. That the custom existed among the Celts is proved by Caesar, De Bell. Gall. VI, 18, spatia omnis temporis non numero dierum, sed noctium finiunt (‘they define all spaces of time not by the number of days but by the number of nights’). The Arabians have the same practice. They say ‘in three nights’, ‘seventy nights long’, and date e. g. ‘on the first night of Ramadan’, ‘when two nights of Ramadan have gone’, or ‘are left’[14].”

For primitive and barbaric peoples the evidence is equally abundant. The Polynesians in general counted time in nights. Night is po, to-morrow is a-po-po, i. e. the night’s night, yesterday is po-i-nehe-nei, the night that is past[15]. The New Zealanders, in former times, had no names for days, but only for nights[16], and so with the inhabitants of the Sandwich Islands—and the same is certainly true of the Polynesians as a whole, since they describe the ‘days’, or rather the nights, by the phases of the moon. The Society Islanders reckon in nights; to the question ‘How many days?’ corresponds in their tongue ‘How many nights?’[17] So also do the inhabitants of the Marquesas[18]. In the Malay Peninsula periods exceeding a fraction of a day are reckoned in nights[19]. Among the Wagogos of German East Africa the phases of the moon and the number of nights serve as more exact determinations of time. The third night after the appearance of the moon, for example, is the day following the third night after the moon’s appearance[20]. Sometimes they say ‘day and night’ when they wish to describe the full day of 24 hours. Occasionally they say that they have worked so many days, with reference to the day-time only[21]. Except in the case of this tribe I have found no notes on the African peoples; little attention seems to have been paid to the point in their case. But the material for America abounds. The Greenlanders reckon in nights[22], though certainly we are not told how those who live north of the Polar Circle reckon in summer. So do the Indians of Pennsylvania[23], the Pawnees, who often made use of notches cut in a stick or a similar device for the computation of nights or even of months and years[24], and the Biloxi of Louisiana[25]. Usually however the night is denoted not by this word but by ‘sleep’, ‘sleeping-time’. Of the Kiowas it is expressly stated[26] that they reckon the length of a journey in ‘darks’, kon, i. e. nights, and not in ‘sleeps’. If the question of the distance of any place arises the answer is ‘so many darks’. It may even be doubted whether ‘sleep’ is not sometimes translated ‘night’ by the reporters. The Dakotas say that they will return in so many nights or sleeps[27]. Among the Omahas the night or sleeping time marked the division of days, so that a journey might be spoken of as having taken so many sleeps[28]. The Hupas of Arizona[29], the tribes of the North-East[30], and the Kaigans of the North-West[31] also reckon in sleeps. This mode of reckoning is therefore the common one, that of the Comanches in suns is an exception. Finally the natives of Central Australia also count time in ‘sleeps’[32].

To reckon in nights is therefore the rule among the primitive Indo-European peoples, the Polynesians, and the inhabitants of North America. For Asia, which however is not so important for primitive time-reckoning on account of the old and far-reaching influence of civilisation in that continent, for Africa, and for S. America evidence is wanting or is forthcoming only in isolated instances. The reason probably is that in these continents also time is really reckoned in nights, and our informants have not noticed the agreement. This however is an argumentum ex silentio. Be that as it may, the fact remains that at least half the globe reckons the days in nights.

The current explanation of this striking fact is given by Schrader thus:—“Since the chronometer of primitive times is the moon and not the sun, the reason for counting in nights instead of days becomes almost self-evident”[33]. This statement is a priori not perfectly correct, inasmuch as there is and can have been no people that has not observed the daily course of the sun as well as the monthly phases of the moon: as chronometer neither of the two bodies is older than the other. The difference lies in the development of the time-reckoning. In point of fact an inner connection seems to exist between the counting of the days in nights and the designation of the days, or rather the nights, of the month according to the phases of the moon, to which we recur further on. Even such low races as the tribes of Central Australia already have names for the phases of the moon, from which they reckon time[34], but unfortunately we are not told how many. The Polynesians have very elaborately developed these, so that every day has its separate name. The Wagogos also use the phases of the moon as indications of time. The Arabs speak of ten phases of the moon, combining three days under each name. The Indians know the phases of the moon, but seem to have named and made use of them only roughly: the only tribe that possesses a list of the names of the days of the moon-month is the Kaigans[35], and unfortunately this list is incomplete. Moreover there are no indications that the primitive Indo-European peoples distinguished the phases of the moon otherwise than roughly. The finer distinction and nomenclature of the moon-phases, so that in the end each day comes to have its separate name, is clearly a very far advanced special development: the use of the word ‘night’ to express the period of 24 hours is much older. A causal connection, such as Schrader and others have maintained, must lie in the fact that the period of 24 hours is named after the phases of the moon and consequently the day itself is reckoned in nights. But this is only a comparatively isolated and advanced development, against which must be set the fact that the Indians and so primitive a people as the Australians use not the word ‘night’ but ‘sleep’, which has nothing to do with the moon.

The explanation must therefore be sought elsewhere, and is one which also applies to the use of the word ‘winter’ for year etc. Primitive man knows only concrete indications of time, and in reckoning prefers to use a concrete and clearly visible point of reference. The complete day of 24 hours is unknown to him and so he must reckon according to the principle of pars pro toto, and as a matter of fact it is possible to reckon just as well from a part of the whole as from the whole itself, provided that the part chosen is one that only recurs once every day. The day itself, with its various occupations, offers no such point of reference unless the reckoning is based upon the daily appearance of the sun, which is also actually done in certain cases. However in the daily course of the sun, as we have already seen, two features, its duration and the changing position of the sun, stand out prominently: but it is easier to reckon from points than from lengths, which divert the attention from the number. Now the sleeping-time is necessarily bound up with each day, yet it has no separate parts, or acquires them only later among certain peoples. The time between going to sleep in the evening and waking in the morning appears as an undivided unit, a point. It offers for reckoning a convenient basis in which no mistake or hesitation is possible such as can occur in the various occupations that fall within the period computed. The method of reckoning in nights is merely an outcome of the necessity for a concrete unmistakable time-indication: it is a typical example of the pars pro toto principle and time-reckoning, which, on the psychological grounds just mentioned are especially favoured in the counting.

For the indication of a point of time within the day the reference to the course of the sun is the means that lies nearest to hand, and the indication can indeed be given quite concretely by means of a gesture in the direction of the heavens. This language of signs is especially common in Africa. The Cross River natives of Southern Nigeria indicate the time by pointing to the position in the heavens which the sun occupies at that time of the day[36]. When someone asked a Swahili what time it was, he answered, “Look at the sun”, although this tribe knew other ways of indicating time[37]. The Wagogo in order to shew the time of day indicate with the hand the position of the sun in the heavens[38]. In Loango the people indicate the time satisfactorily enough from the motion of the sun, in divisions of two hours, by dividing the vault of the sky with outstretched arm, often using both arms as indicators[39]. Moreover most peoples have descriptive expressions for parts of the day, as for instance the inhabitants of the Lower Congo[40], the Masai of East Africa, who estimate the time of day from the position of the sun[41], and the Hottentots, who express with certainty and clearness both points and duration of time by referring to the position of the sun[42]. In Dahomey the natives tell the hours by means of the sun; they say that the sun is here or there, in order to give the time of day[43]. The Caffres are able to give the exact time of day by pointing with outstretched arm to the spot at which the sun appears at the time they wish to indicate. So, for example, when the Caffre wishes to shew that he will come at two o’clock in the afternoon of the next day, he will say, “I will be here to-morrow, when the sun is there”,—pointing to the position occupied by the sun at 2 p. m.[44]. The Waporogo of German East Africa estimate the divisions of the day from the position of the sun, which they indicate with outstretched arm. When the arm is vertically raised, that means 12 o’clock noon, and the other hours of the day they are able to give with a sure instinct by means of a greater or lesser inclination of the arm towards the body, corresponding to the position of the sun[45]. In other parts of the world we find the same thing. Thus in the New Hebrides the hours of the day are indicated by pointing with the finger to the altitude of the sun[46]. If a native of Australia is asked at what time anything took place or is going to take place, his answer will take the form of pointing to the position which the sun occupied or will occupy in the sky at that particular time[47]. The Bontoc Igorot of Luzon point to the heavens in order to indicate the position the sun occupied when a particular event occurred[48]. The Kanyans of Sarawak, if asked at what time anyone will arrive, point to the sun and say, “When the sun stands there”[49]. In the Dutch East Indies the time of day is given from the position of the sun[50]. The inhabitants of Java divide the day into ten natural but vague and unequal subdivisions, and for astrological purposes the day of 24 hours is divided into five parts. They also determine the time of day by the length of the shadow and by the working-time, but the most common method is by pointing to the situations of the sun in the heavens, when such and such an event took place[51]. In order to indicate the time the natives of Sumatra also point to the height in the sky at which the sun stood when the event of which they are speaking occurred[52]. The natives of the western tribe of the Torres Straits, though they have learned to tell the time from the clock, also know how to give it very accurately by observing the height of the sun[53]. The Tahitians determine the six parts of their day from the sun’s altitude[54]. Among the Omaha Indians the sun indicates the time of day. A motion towards the zenith meant noon, midway between the zenith and the west, afternoon, and midway towards the east, forenoon[55]. The Karaya of Central Brazil divide up the day according to the position of the sun. Indications of time are given by pointing with the hand to the place occupied by the sun at the time in question[56].

This method of indicating the time of day is quite satisfactory, especially in the tropics and for primitive needs, and only more rarely does it give place to other methods, the chief of which is the observation of the length of shadows. The Javanese know this latter method but do not often use it. In their old writings we find a traveller described as setting out on his journey or arriving at the end of it when his shadow was so many feet long[57]. The Masai usually estimate the time of day from the position of the sun, but more rarely from the length of the shadows[58]. When the shadow measures nine feet, the Swahili say, “It is 9 o’clock (sic!)”[59]. To indicate the time of day or to represent a distance the Cross River natives use the length of shadows. They have however in most of their houses a curious species of sun-dial, a plant about 50 cm. high, with violet-white flowers. The flowers gradually begin to open at sunrise, by noon they are wide open, and they gradually close again between noon and sunset. One of these plants is placed in every garden and enclosed within little stones[60]. To the south of Lake Nyassa the time of day is reckoned either from the position of the sun or from the length of the shadow thrown by a stick, nthawe[61]. The Society Islanders among their numerous expressions for the time of day include two which have reference to shadows, ‘the shadow as long as the object’, ‘the shadow longer than a man’[62]. The Benua-Jahun, a primitive tribe of the Malay Peninsula, indicate the progress of the day by the inclination of a stick. Early morning is represented by pointing a stick to the eastern horizon. Placed erect it indicates noon, inclined at an angle of about 45° to the west it corresponds nearly with three o’clock, and so on[63]. This practice is doubtless connected with the common use of a stick in the Indian Archipelago for observations of time, and is by no means primitive. The ancient Athenians seem to have indicated time by measuring off with the foot the length of the shadow cast by their bodies upon the level ground before them as they stood. At all events the length of shadows served to indicate time, cp. Aristophanes, Ekkles., 652, “when the staff is ten feet, to go perfumed to dinner”[64]. The gnomon which, according to Herodotus II, 109, the Greeks borrowed from the Babylonians was an upright stick the shadow of which was measured: it was also an important instrument for astronomical observations[65]. Here however we are already at a highly developed stage and know nothing about the origins.

The indication of time from the position of the sun is really only satisfactory in the tropics, where the sun always stands very high and the length of its daily course is not exposed to too great variation. Where the sun is much lower in winter than in summer, and the length of the day varies greatly at different times of the year, the method ceases to be practicable. If descriptive expressions of one kind or another are not resorted to, other means must be found. Above all it is important to determine the fixed point which divides the day into two parts, i. e. noon. In the living-room of the houses of the Scanian peasants, which were always built ‘according to the sun’, i. e. facing east and west, there was in the southern window-sill, beside the middle shaft of the frame, a line which was called the ‘noon-line’. When the shadow of the shaft fell parallel with this line it was noon. This device is not exactly primitive, since windows in the room, more particularly in the wall, belong to a quite advanced stage of civilisation. But on the other hand such customs as the determination of noon and other moments of the day from the position of the sun above certain points on the horizon—elevations and hills—are old. In Iceland the divisions of the day were, and still are, determined from the visible course of the heavenly bodies. The people imagined that the sun in the course of a day and a night ran through the eight equal regions of the heavens (ættir, sing. ætt). The time of day was determined from the position of the sun above the horizon by the selection in every house of certain outstanding points within the range of vision to serve as ‘day-marks’ (dagsmǫrk, sing. -mark)—where these were lacking, small piles of stones were erected for the purpose—so that when the sun stood above one of these marks a certain time of day was given. The most important times thus determined were rismál or miðr morgin (6 a. m.), dagmál (9 a. m.), hádegi (12 o’clock noon), míðmundi (1.30 p. m.), nón (undoubtedly originally called undorn and also eykt, 3 p. m.), miðr aptann (6 p. m.), and nattmál (9 p. m.). These indications in hours are however only approximate, since the time varies according to the position of the place in question[66]. The word eykt really designates any of these approximately three-hour divisions; but since the length of the day varies enormously so far north, the business of everyday life leads to an attempt at systematising, e. g. rismál = ‘the time of rising’. The spot which the sun has reached at one of these divisions is therefore called dagmálastað, nónstað, eyktarstað etc. This mode of determining time must be old since it is also found in Scandinavia, where it has given names to many mountain-peaks. In Baedeker I have only noticed:—Middagsfjället in Jämtland, Middagshorn in Norangdal, Middagshaugen in Aardal, Sogn, Middagsnib in Oldendal in the Nordfjord district, Middagsberg on the Nærøfjord in Sogn, Nonsnib above Loen Water in Nordfjord, Solbjørgenut in the Nærøfjord, Sogn. From Fritzner’s Old Norwegian Lexicon (s. v. eyktarstað) I take:—Durmaalstind, Rismaalsfjeld, Nonsfjeld, Natmaalstinden, Middagsfjeld in Tromsö ‘amt’ and in Finnmarken, Eyktargnipa and Undornfjeld in Mule Syssel in Iceland; the peak of the latter lies in the nonstað. Such names are common in Norway. In Sweden there are further:—Middagsberget in Dalecarlia = Gesundaberget, just south of Mora; the name is found again in Härjedalen, in addition to Nonsberget, Nonsknätten and Middagshognan. Lidén[67] instances similar names in S. Sweden and in England, and also those formed with mosse, ‘swamp’, vik, ‘bay’, and åker, ‘field’. It is easy to understand why middag, ‘noon’, everywhere predominates as a nomenclator. The Lapps also indicate time by the position of the sun in relation to the surrounding natural objects[68].