TRANSCRIBER’S NOTE

Footnote anchors are denoted by [number], and the footnotes have been placed at the end of the book.

In the Footnotes a reference to a second or third edition of a book is denoted by ² or ³, for example: Schrader, II³.

This book has many Greek words, which should display correctly on most devices. Some other less common characters are also used. These will display on this device as
ð eth character
Þ thorn character
ǫ o with ogonek
ȱ o with dot and macron
å a with ring above
ă a with breve
ā ī ō a, i, o with macron
ǎ č ř š ž a, c, r, s, z with caron

The cover image was created by the transcriber and is placed in the public domain.

Some minor changes to the text are noted at the [end of the book.]

SKRIFTER UTGIVNA AV
HUMANISTISKA VETENSKAPSSAMFUNDET I LUND

ACTA SOCIETATIS HUMANIORUM LITTERARUM LUNDENSIS

I.

MARTIN P. NILSSON

PRIMITIVE TIME-RECKONING

PRIMITIVE TIME-RECKONING

A STUDY IN THE ORIGINS AND FIRST DEVELOPMENT
OF THE ART OF COUNTING TIME AMONG
THE PRIMITIVE AND EARLY
CULTURE PEOPLES

BY

MARTIN P. NILSSON

PROFESSOR OF CLASSICAL ARCHÆOLOGY AND ANCIENT HISTORY IN THE UNIVERSITY OF LUND
SECRETARY TO THE SOCIETY LETTERS OF LUND
MEMBER OF THE R. DANISH ACADEMY

LUND, C. W. K. GLEERUP
LONDON, HUMPHREY MILFORD PARIS, EDOUARD CHAMPION
OXFORD, UNIVERSITY PRESS LEIPZIG, O. HARRASSOWITZ
1920

LUND 1920

BERLINGSKA BOKTRYCKERIET

PREFACE.

Although in the present study I devote only a few pages to the Greek time-reckoning, and am engaged for the most part in very different fields, yet the work has arisen from a desire to prepare the way for a clearer view of the initial stages of the Greek time-reckoning. In the course of my investigations into Greek festivals I had from the beginning been brought up against chronological problems, and as I widened the circle so as to include the survivals of the ancient festivals in the Middle Ages, more particularly in connexion with the origin of the Christmas festival, I was again met by difficulties of chronology, this time in regard to the earlier Germanic time-reckoning. In the year 1911 I published in Archiv für Religionswissenschaft an article on the presumptive origin of the Greek calendar circulated from Delphi. These preliminary studies led to my taking over myself, in the projected Lexicon of the Greek and Roman Religions, the article on the calendar in its sacral connexions. This article was worked out in the spring of 1914. In it the emphasis was laid not on the historical chronological systems, which have little to do with religion, but on the question of origins, in which religion plays a decisive part. In order to arrive at an opinion it was not enough to work over once more the extremely scanty material for the origin of the Greek time-reckoning; I had to form an idea from my hitherto somewhat occasional ethnological reading as to how a time-reckoning arose under primitive conditions, and what was its nature. This idea obviously required broadening and correcting by systematic research. The war, which suspended the continuation of the Lexicon at its very beginning, gave me leisure to undertake this more extensive research. Certainly it has also imposed some limitations on the work, since I could not make use of the rich libraries of England and the Continent but had to be content with what was offered by those of Sweden and Copenhagen. But I am not disposed to regret this limitation too deeply. The material here reproduced will probably strike many readers as being copious and monotonous enough, and the numerous books of travels and ethnological works which I have ransacked, often to no profit, seem to hold out little prospect that anything new and surprising will come to light. In this conviction Webster’s work has strengthened me.

In two or three instances I have derived material of great value from personal communications. For very interesting details of the time-reckoning of the Kiwai Papuans I am indebted to Dr. G. Landtman of Helsingfors, and Prof. G. Kazarow of Sofia has sent me valuable information as to the Bulgarian names of months. Dr. C. W. von Sydow of Lund has communicated to me details of the popular time-reckoning in Sweden.

An exhaustive examination of all the material obtainable would doubtless lead to a more exact conception of the details of primitive time-reckoning. Above all, large districts with similar peculiarities in time-reckoning could be more accurately defined. The Arctic regions form a district of this nature. South America again differs characteristically from North America; Africa, the East Indian Archipelago, and the South Sea Islands all have their peculiarities. The borrowings which have undoubtedly taken place on a very large scale would be at least in part pointed out. This working up of the material is however the task of the ethnological specialist; my object is simply and solely to attain the above-mentioned goal of a general foundation.

The observation of chronological matters varies greatly in the ethnographical literature; I have gone through many books without result, and in other cases my gains have often been small. It is only in quite recent times that attention has been paid with any great profit to this side of primitive life. Among the English authors Frazer has drawn up a list of ethnological questions (printed in the Journal of the Royal Anthropological Institute, 18, 1889, pp. 431 ff., and also separately), paying due attention to time-reckoning, which has had a lasting and happy result, as can be seen especially in many papers in the JRAI of succeeding years.

Of the works of my predecessors only one has had any more elaborate aims—the ninth chapter of Ginzel’s handbook, which deals with the time-reckoning of the primitive peoples, divided up according to the different parts of the world. The significance of the time-reckoning of the primitive peoples for the history of chronology seems to have been only gradually grasped by the author in the course of his work, since it is not until after he has touched occasionally upon the question of primitive time-reckoning in the course of his account of the chronological systems of the Oriental peoples that he inserts the chapter in question between the latter and the chapters on the chronology of antiquity. Ginzel has in many respects a sound view of the nature of primitive time-reckoning, and makes many pertinent remarks, but on the whole his treatment, as is not seldom the case, is lacking in exactness and depth. I have gratefully made use of the material collected by him, going back, wherever possible, to the original sources. Of other previous works must be mentioned the essays of Andree and Frazer on the Pleiades,—the latter especially distinguished by its author’s usual extensive acquaintance with the sources and by its abundance of material—and the dissertation of Kötz upon the astronomical knowledge of the primitive peoples of Australia and the South Seas, an industrious work which however only touches superficially upon the problems here dealt with, and in regard to the lunisolar reckoning adopts the view of Waitz-Gerland:—“We can here discover nothing accurate, since these peoples have conceived of nothing accurately” (p. 22). I think however that we may fairly say that this is to estimate too meanly the possibility of our knowledge. Hubert’s paper, Étude sommaire de la représentation du temps dans la religion et la magie, is composed throughout in the spirit of the neo-scholastic school of Durkheim. The present work, on the other hand, is based upon facts and their interpretation.

The book was ready in the spring of 1917, but could not be published on account of the war. Later I have only inserted a few improvements and additions. As I was putting the finishing touches to my work, there came into my hands, after a delay due to the circumstances of the time, the Rest Days of H. Webster, whose Primitive Secret Societies has gained him fame and honour. This work deals in detail with a subject akin to mine, but not from the calendarial and chronological standpoint here adopted. Only upon the origin of the lunisolar calendar does the author make a few general remarks (pp. 173 ff.), which however do not advance the subject very far. In the chapters entitled Market Days, Lunar Superstitions and Festivals, Lunar Calendars and the Week he has brought together abundant material which also concerns some of the phenomena treated by me; part of this information will not be found here, since it is compiled from sources inaccessible to me. For the same reason, because I could not collate it for myself, I have not thought it advisable to introduce this material into my book, especially since it adds no new principle of knowledge and does not affect the conclusions I have drawn. Moreover anyone who wishes to go farther into these matters must in any case approach Webster’s careful work.

For the popular month-names of the European peoples I have made use of the well-known extensive collections of Grimm, Weinhold, Miklosisch, etc. In this chapter my object has not been to make contributions to our knowledge of the popular months, but only to bring out, by means of numerous examples, the parallel between the popular names of the Julian months and the names of the lunar months among the primitive peoples. More isolated and disputed names are therefore omitted, and the names are given chiefly in translation. I have made only one exception, namely in the case of the Swedish lunar months, which really hardly belong to my subject since they are a popular development from the ecclesiastical calendar of the Middle Ages. I hope however to be excused for this, in the first place on patriotic grounds, and secondly because little attention has hitherto been paid to the matter. In another place I have dealt fully with the Swedish names of months, which are in the majority of cases not of popular origin.

I have made out a list of authorities so that in the foot-notes reference may be made simply to the name of the author; where an author is represented by two or more works, the work in question is denoted by an abbreviation. This list is to be regarded not as an exhaustive bibliography, but merely as an aid to the quotations. Where so many quotations have been made it has been thought advisable not to use inverted commas, except in a few special cases. The fact that the quotations are nevertheless given as far as possible in the author’s own words must be held to excuse a certain apparent inconsistency in the use of tenses.

Since I was obliged to include in my work the preliminary stages of the time-reckoning of the culture peoples, I had to deal with languages with which I was altogether unfamiliar, or only imperfectly acquainted. I have therefore often availed myself of the expert advice which has been readily given me by friends and colleagues. For help in the complicated questions belonging to the domains of the Semitic languages and Anglo-Saxon respectively I am especially indebted to my colleagues Professors A. Moberg and E. Ekwall. For occasional advice and information I have to thank Docent Joh. Pedersen of Copenhagen (for the Semitic languages), Prof. Emil Olson of Lund, and Prof. H. Lindroth of Gothenburg (for the Scandinavian), and Docent S. Agrell of Lund (for the Slavonic).

The English translation is the work of Mr. F. J. Fielden, English Lector in the University of Lund, who has also read the proof-sheets. I am greatly obliged to him for his conscientious performance of a lengthy and by no means easy task.

Lund, May 1920.

Martin P. Nilsson.

CONTENTS.

INTRODUCTION.

The ancient civilised peoples appear in history with a fully-developed system of time-reckoning—the Egyptians with the shifting year of 365 days, which comes as nearly as possible to the actual length of the year, counting only whole days and neglecting the additional fraction; the Babylonians and the Greeks with the lunisolar, varying between twelve and thirteen months and arranged by the Greeks from the earliest known period of history in the cycle of the Oktaeteris. It has always been clear that these systems of time-reckoning represent the final stage of a lengthy previous development, but as to the nature of this development the most daring hypotheses have been advanced. Thus, for example, eminent philologists and chronologists have believed the assertion of Censorinus, Ch. 18, and have supposed that the Oktaeteris was preceded by a Tetraeteris, even by a Dieteris. It may indeed at once be asserted that such a hypothesis lacks intrinsic probability. To account for the early development hard facts are needed, and unfortunately these, especially in the case of the Greeks, are extremely few. Where they are required they must be sought elsewhere.

Setting aside all ingenious but uncertain speculations, our only practicable way of proceeding is by means of a comparison with other peoples among whom methods of time-reckoning are still in the primitive stage. This is the ethnological method which is so well-known from the science of comparative religion, but the claims of which have been so vigorously contested upon grounds of no small plausibility. Fortunately this dispute need not be settled in order to prove the validity of the comparative method for an investigation into the origin and development of methods of reckoning time. The gist of the dispute may be expressed as follows:—The ethnological school of students of comparative religion assumes that the intellect of the natural man can only master a certain quite limited number of universal conceptions; from these spring more and more abundantly differentiated and complicated ideas, but the foundation is everywhere the same. Hence our authority for comparing the conceptions of the various peoples of the globe with one another in order to lay bare this foundation. The opponents of the school deny the existence of these fundamental conceptions, and maintain that the points of departure, the primitive ideas of the various peoples, may be as different as the peoples themselves, and that therefore we are not authorised in drawing general conclusions from the comparison or from the fundamental conceptions themselves.

In the matter of the indication and reckoning of time, however, we have not to do with a number of conceptions which may be supposed to be as numerous and as various as we please. At the basis lies an accurately determined and limited and indeed small number of phenomena, which are the same for all peoples all over the globe, and can be combined only in a certain quite small number of ways. These phenomena may be divided into two main groups: (1) the phenomena of the heavens—sun, moon, and stars—and (2) the phases of Nature—the variations of the climate and of plant and animal life, which on their side determine the affairs of men; these, however, depend finally upon one of the heavenly bodies, viz. the sun. The claim that the comparative ethnological method can be justified only when we are dealing with a narrowly circumscribed number of factors is therefore here complied with, owing to the very nature of the subjects treated. The comparative method does not shew how things have happened in a special case in regard to one particular people: it only indicates what may have happened. But much is already gained if we can eliminate the impossibilities, since from the complete result of the development, no less than in other ways, we may obtain a certain basis for our deductions.

For the investigation of primitive methods of time-reckoning no special astronomical or other technical knowledge is needed: in fact, such knowledge has rather played a fatal part by causing attention to be paid exclusively to the system of time-reckoning and leading to constant attempts to discover older and more primitive systems. A priori, indeed, we might venture to state that a system is always based upon previous data: unsystematic indications of time precede the system of time-reckoning. These modest beginnings have been obscured from view by the prejudice in favour of the systematic technical and astronomical chronology. The only absolutely necessary thing is a clear idea of the apparent motions of the heavenly bodies, i. e. the sun, the moon, and the most important of the fixed stars, and of the phases of the climate and the life of animals and plants, which give the units of the time-reckoning.

For a statement of the course and phases of the heavenly bodies and the units of the time-reckoning given by these I refer to the article mentioned in the preface, the pertinent sections of which are here quoted:—

“The units of the time-reckoning are given by the motions of the heavenly bodies (expressed according to the Ptolemaic system), and the more intimately these enter into the life of man, the more important do they become. For this reason only those units which depend upon the sun have asserted themselves in our calendar, those depending upon the moon having been dropped, except for the movable paschal term, which has been kept on religious grounds. The units are the year, the month, and the day. Other units more convenient for time-reckoning play no part in the arrangement of the calendar since they are without importance for practical life. The day (= 24 hours, νυχθήμερον) is determined from the apparent motion of the heavenly bodies about the earth, which is caused by the rotation of the earth on its axis; but since the sun also, on account of the annual revolution of the earth about it, runs through the zodiac in an opposite direction to its daily movement and completes the circle of the ecliptic in a year, a day will be a little longer than a complete rotation of the earth. Or to put it otherwise:—The time between two successive upper culminations of a star, i. e. between the moments at which the star passes through the meridian-line of one and the same place (= attains the zenith), represents an axial rotation: that is a stellar day. The time between two successive culminations of the sun is, on account of the annual motion of the sun (really that of the earth), 3 min. 56.₅ secs. longer than a stellar day: that is a solar day. The number of stellar days in a year is greater by one day than the number of solar days. The stellar day does not follow the variations of light and darkness and therefore does not enter into the calendar. The difference between the actual solar day, which is of slightly varying length, and the mean solar day abstracted from it for the purposes of our clock-regulated time-reckoning has no significance for antiquity. The second unit determined by the sun is the year, the period of a revolution of the earth about the sun. In relation to the apparent motion of the sun it may be defined as the time which the sun takes to come back again to the same fixed star. This is a stellar or sidereal year, the length of which amounts to 365 days 6 hrs. 9 min. 9.₃₄ secs. The tropic year is the time which the sun takes to come back to the crossing point of the equator, viz. the vernal equinox. This is the natural year. Its length varies a little; it is about 20 minutes shorter than the stellar year. The lunar or moon-month is determined from the visible phases of the moon. This term will be used only when it is necessary to make an express distinction between the lunar and our Roman month; the latter is a conventional subdivision of the year which has nothing to do with the moon, and has the name ‘month’ only because it historically arose from the lunar month and in its duration comes fairly near the latter. But when in relation to antiquity—apart from Rome and Egypt—we speak of months, lunar months are as a rule to be understood. The moon revolves around the earth twelve times a year and a little more: consequently it moves backwards in the zodiac much more rapidly than the sun. The interval between two successive moments at which the moon culminates at the same spot at the same time as one and the same star is a sidereal month (cp. the sidereal year); its length is 27 days 7 hrs. 43 min. 11.₄₂ secs., but it does not follow the phases of the moon and is therefore of no consequence for the calendar. The phases of the moon are dependent upon the position of the moon in relation to the sun and the earth. When the three bodies are in a straight line (or rather in a plane perpendicular to the plane of the ecliptic) in such a way that the earth is in the middle, the side of the moon turned towards the earth is completely illuminated and we have full moon: when the moon is in the middle, the side turned towards the earth is completely overshadowed, and that is new moon. In between lie the separate phases of the waxing and waning moon. The synodic month is the interval between two new moons and comprises on an average 29 days 12 hrs. 44 min. 2.₉₈ secs. This is the true lunar month: other varieties of month are of no importance for us.

"The risings and settings of the stars. It has already been remarked that the sun in the course of a year runs through the zodiac backwards, so that one particular star culminates 3 min. 56 secs. earlier every day. Hence it is evident that if we indicate the exact interval of time between the culmination of the sun and that of one particular star, or name the star with which the sun precisely culminates, we can determine the day of the solar year. This is the principle of one method of computing time which was very common among ancient and primitive peoples, but has entirely dropped out of use in modern times owing to our paper calendar. The stars are so to speak the stationary ciphers on the clock-face and the sun is the hand. In practice we naturally have to do not with the invisible culmination of the stars but with the position of the sun and certain neighbouring stars on the edge of the horizon, whereby the matter becomes more complicated on the astronomical side. For this observation the so-called circumpolar stars are singled out, that is to say the stars situated so near the pole that they do not set (e. g. the Great Bear). If the star rises or sets simultaneously with the rising of the sun, this is called the true cosmic rising or setting. If the star rises or sets simultaneously with the setting of the sun, this is termed the true acronychal rising or setting. These risings and settings of the star are not visible, since the sun hides them by its light: the rising and setting are perceptible only when the star stands at some distance from the sun, i. e. only the so-called apparent rising and setting are practically observable. We have already seen that the sun every day drops nearly 4 minutes behind a certain star. Assuming that sun and star rise simultaneously on one day (true cosmic rising), then after a few days have passed—the period varying somewhat according to the latitude of the place of observation, the time of the year, the size and place of the star—there will come a day on which the star rises so early that it is visible in the morning twilight, immediately before the sun appears. This is the heliacal or morning rising. From this day the star will rise earlier and earlier, and will therefore remain visible for a longer and longer period. In the course of half a year, commonly a little sooner or later, the time of rising will have been pushed so far back that it will take place in the evening twilight; when it is pushed still farther back the rays of the setting sun eclipse the star and its rising is no longer visible. The last visible rising of the star in the evening twilight is the apparent acronychal or evening rising. After a few more days the star goes so far back that it rises at the very moment in which the sun sets—the true acronychal rising. The rising, which is advanced constantly further into the light of day, is no longer visible, but on the other hand we now see the setting of the star. If it is assumed that the star is situated on the western horizon, i. e. sets, when the sun is on the eastern horizon, i. e. rises—and incidentally it is to be noted that this position, when the star is not situated in the ecliptic, may be divided by an interval of a larger or smaller number of days from the opposite position, viz. star on the eastern, sun on the western horizon—this is the true cosmic setting. The star moves forward, i. e. its setting takes place earlier in the morning, and after a few days it will be noticed in the morning twilight immediately before it sets, and this is the first visible setting in the morning twilight, the apparent cosmic or morning setting. From this day the setting moves further and further forward into the night and approaches the evening twilight. At length it will be so near sunset that the star no longer sets in the night but in the evening twilight. The last visible setting of the star in the evening twilight is the heliacal or evening setting. After a few days the star has approached still nearer to the sun: both set at the same moment, the true cosmic setting. If the star stands in the ecliptic, the true cosmic setting coincides in date with the true cosmic rising, otherwise these are divided by a greater or smaller number of days (see above). As the star moves on, a heliacal rising follows again, and so on. Between the day of the heliacal setting and that of the heliacal rising the star is invisible, since it stands so near the sun that it is eclipsed by the sun’s rays. It has already been remarked that we can determine the day of the year by indicating the true rising and setting of a star at a certain spot. As far as the apparent rising and setting are concerned this indication can only be approximate, since the visibility of a star depends on several variable factors—the size of the star (because a smaller star, in order to be visible, must move farther from the sun than a brighter one), the transparency of the atmosphere, the keenness of vision of the observer, the geographical latitude of the place of observation (since the farther north or south the sun is, the more slowly, because more obliquely, will it sink below the horizon). In this latter respect, for instance, there is a perceptible difference between Rome and Egypt. Only an approximate indication of time, therefore, can be derived from the rising and setting of the stars”.

The phases of the climate and of plant and animal life cannot be particularly described, since they naturally vary so much in different countries. It can only be remarked that though they depend upon the course of the sun, yet in certain cases, owing to the special climatic conditions of the individual years, they may be to some extent advanced or retarded, and further that the climatic phenomena of many parts of the earth, especially in the Tropics but also in the Mediterranean countries, recur with a far greater regularity than in our northern climes, which are subject to such uncertain weather. Instances are the trade-winds and monsoons, the dry and the rainy seasons.

Upon the above-mentioned units the system of time-reckoning will be based. The days are joined into months and the months into years; only more rarely are the seasons interposed as regular units of time. The system is like a chain the links of which run into one another without gaps: each link is equivalent, or as nearly as possible equivalent, to every other link of the same class, and therefore need only be given a name and counted, not necessarily conceived in the concrete, although this is not excluded. This is the only genuine system, a system of continuous time-reckoning, which excludes all gaps in the chain and all links of indeterminate length. The relation between the larger and the smaller units may be treated in various ways, chiefly on account of the fact that the smaller units do not divide exactly into the larger. Sometimes the smaller units may be fitted into the larger as subdivisions of the latter, so that they constitute the links of the chain formed by the larger unit. The inequality referred to shews then that the units vary to some extent in number or size (year of 365 or 366 days, of 12 or 13 lunar months, lunar month of 29 or 30 days). In that case the beginnings of the larger unit and of the first of the smaller units coincide. Thus in our year New Year’s Day and the first day of the first month coincide, but the length of the months varies somewhat. This is an inheritance from the lunisolar year, in which also New Year’s Day and the first day of the first month coincided and the length of the month varied between 29 and 30 days, but in addition the year varied between 12 and 13 months. This mode of reckoning, in which the smaller units are contained in the larger as subdivisions of them, will be termed the fixed method.

But where the smaller units do not exactly divide into the larger, both may also be counted independently of one another without being equalised. A case in point is our week, which is reckoned without reference to the year, so that every year begins with a different day of the week. This method of reckoning we shall term the shifting method. It is less systematic than the fixed method, and we shall therefore expect to find it play a greater part in earlier times than at the present day.

The system of time-reckoning, the continuous counting of the time-units, represents the final point of the development. It is our object to investigate the preceding stages, both systematic and unsystematic. Certain important ideas which frequently recur must however first be clearly set down. The time-reckoning in the proper sense of the term is preceded by time-indications which are related to concrete phenomena of the heavens and of Nature. Since these indications depend upon the concrete phenomenon, their duration fluctuates with the latter, or rather the duration does not stand out by itself but the phenomenon as such is exclusively regarded: the time-indication is not durative, like the link in any system of time-reckoning, but indefinite, or, to borrow a grammatical term, aoristic. And setting aside these finer distinctions we also find that the phenomena to which the time-indications are related are of fluctuating and very unequal duration. Since the duration is indeterminate and fluctuating, and the time-indications are not limited one by the other but overlap and leave gaps, they cannot be numerically grouped together. Here we ought really to speak not of a time-reckoning in the proper sense, but only of time-indications. But since the word ‘time-reckoning’ has become naturalised, this method may be described as the discontinuous system of time-reckoning, because the time-indications do not stand in direct relation to other time-indications but are related only to a concrete phenomenon, and through that to other time-indications, so that they are of indeterminate length and cannot be numerically grouped together.

If the number of dawns, suns, autumns, or snows that has passed since a certain event took place, or will elapse before a certain event is to take place, be indicated, the time that has passed or is to pass will be defined, because the dawn or the sun recurs once in the day, and an autumn or a snow, i. e. winter, once in the year. This is the oldest mode of counting time. It is not the units as a whole that are counted, since the unit as such had not yet been conceived, but a concrete phenomenon recurring only once within this unit. It is the pars pro toto method so extensively used in chronology, and by this name we shall call it[1].

Since it must now be regarded as the natural course of development that the systematic has gradually arisen out of the unsystematic, and that the indication of concrete phenomena following one another in the regular succession of Nature has preceded the abstract numerical indication of time offered by our calendars, the origin of the time-reckoning must be sought not in any one system, however simple, but in the discontinuous or pars pro toto time-indications which are related to concrete phenomena.

Our task is now to make clear the nature of these discontinuous and pars pro toto time-indications, since from them proceeds, as order is ever evolved out of chaos, the continuous time-reckoning, the calendar.

CHAPTER I.
THE DAY.

For primitive man the day is the simplest and most obvious unit of time. The variations of day and night, light and darkness, sleeping and waking penetrate at least as deeply into life as the changes following upon the course of the year, such as heat and cold, drought and rainy seasons, periods of famine and plenty. But for the primitive intellect the year is a very long period, and it is only with difficulty and at a later stage that it can be conceived and surveyed as a whole. Day and night, on the other hand, are short units which immediately become obvious. Their fusion into a single unit, the day of 24 hours, did not take place till later, for this unit as we employ it is abstract and numerical: the primitive intellect proceeds upon immediate perceptions and regards day and night separately.

Evidence for this fact is furnished by most languages, which are as a rule without any proper term for day and night together, the circle of 24 hours. In writing English one sadly misses the Swedish dygn, which has exactly the required significance. The German Volltag is an artificial and not very happy compound. The Greeks also formed a learned and rare (though good) compound, νυχθήμερον. The usual method is to make use of a term according to the pars pro toto principle. This principle, which we meet here at the outset and shall come across more and more frequently in the course of the following pages, is of great importance for the development of time-reckoning since it shews how the original time-indication is discontinuously related to a concrete phenomenon, and only slowly and at a later period develops into a continuous numerical unit of time.

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].

The gestures may be accompanied by descriptive expressions, as among the negroes, or replaced by them, which seems to be the rule among other peoples. The latter practice offers the further advantage of being available in the night-time, when it is necessary to mention a point of time after dark. The Kayans denote the time of day by pointing to the position of the sun, but for morning and evening they also use the expressions ‘when the sun has risen’ or ‘set’[69]. Expressions for the most important divisions, sunrise and sunset (= morning and evening) and noon, are found among all peoples. Even the tribes of Central and Northern Australia have words e. g. for evening and for morning before sunrise[70]. The richness of the terminology however varies exceedingly. The Indians divide the day into three or four rough divisions only. The Seminole of Florida divided up the day by terms descriptive of the positions of the sun in the sky from dawn to sunset[71]: unfortunately we are not told what these words were or how many of them existed. Among the Hopi of Arizona there is every evidence that the time of day was early indicated by the altitude of the sun[72]. The Omahas know no smaller divisions of the day than morning, noon, and afternoon, to which certainly must be added the transitional periods of sunrise and sunset[73]. The Occaneechi of Virginia measure the day by sunrise, noon, and sunset[74]. The Algonquins of the same province mention the three times of the rise, power, and lowering of the sun[75]. Many tribes however had four divisions[76], e. g. the Natchez of Louisiana, who divided the day into four equal parts: half the morning, until noon, half the afternoon, until evening[77]. But there is also a richer terminology, e. g. the Kiowa words for dawn (‘first-light’), sunrise (lit. ‘the-sun-has-come-up’), morning (lit. ‘full-day’), noon, earlier afternoon until about 3 o’clock, late afternoon, evening (lit. ‘first-darkness’)[78]; and in particular among the Statlumh of British Columbia: dawn (‘it-just-comes-day’), early morning (‘just-now-morning’), morning light (‘just-see-things’), full light (‘just-now-day’), sunrise (‘outside-sun’), early morning (midway between sunrise and noon), noon (up till about 2 p. m.), middle of the afternoon, about 4 p. m., ‘three-fourths-of-the-day-have-gone’, ‘sun-sitting-down’, ‘the-sun-gone’,’evening-creeping-up-the-mountain’ (this refers to the line of shadow on the eastern mountains), ‘reached-the-top’, i. e. the line of the shadows, twilight, ‘getting-dark’, night, darkness, pitch dark[79].

Of the Indians of S. America little is reported. ‘The-sun-is-perpendicular’ was the expression for noon on the Orinoco[80]. The Indians of Chile had words for morning twilight, dawn, morning, noon, afternoon, evening, evening twilight, night, and midnight[81].

The terminology for the parts of the day is especially rich in Africa, a fact which is connected with the refinement of the observation of the sun’s position resulting from the custom of indicating this by a gesture in the direction of the heavens. Such simple indications as those of the Babwende for noon, ‘the-sun-over-the-crown-of-the-head’, and for midnight, ‘the-silence-of-the-land’[82], are rare. A number of elaborate time-indications are as a rule employed. The Wadschagga say at six o’clock in the morning ‘the sun rises’, at twelve o’clock ‘the sun rests on his cushion’ (like a tired porter), from twelve to one ‘the sun goes straight on’, about two it ‘bows’, about six it ‘falls down’, or ‘spreads its arms out’, like a man in the act of falling[83]. The terms used by the Bangala are:—about 2 a. m., the lying fowl; 3, the lying bird; 4, the first fowl; 4–5, the sun is near; 5, not translated; 5.₃₀–6, the dawn; 6, the sun is come; 6.₁₅–7, ntete; 12 noon, 2–3, 3–4, not translated; 6, the fowls go in, or the sun enters, or the sun darkens; 6.₃₀, twilight finishes; 11–12, one set of the ribs or one side of a person, meaning that a person turns from lying on one side over on to the other; 12 midnight, second division or second half[84]. In Bornu the expressions for the time of day are formed by the aid of the word dinia = ‘world’, ‘universe’, ‘sky’. From about 4 to 5 ‘the world cuts the aurora’; at 6 ‘the world is light’; at 12 ‘the sun is in the centre of the world’. Afterwards follow ‘it is evening’, twilight, night, midnight. Since the people are Mohammedans they also have expressions for the hours of prayer[85]. The expressions used by the Shilluk of the White Nile are translated[86]:—“The first morning, twilight becomes visible, morning dawn, morning, the earth is morning (it is morning)—the difference here is not evident—noon, the sun is in the zenith, the sun begins to sink (afternoon), it is afternoon, the sun is setting, the sun has set, it is night, at night, midnight.” The Yoruba divide the day into early morning, morning or forenoon, noon (when the day is ‘perpendicular’), shadow-lengthening or afternoon, evening or twilight[87]. The Masai distinguish the following parts of the day:—at 4 a. m. it is ‘not-yet-early’; at 5 it is ‘early’; somewhat later come dawn, twilight (about 5.₃₀, ‘the-sun-is-still-far-off’), and sunrise (‘the-sun-shews-himself-a-little’ or ‘rises’). From 8 to 10 it is ‘still-early’, towards 11 they say ‘the-sun-is-not-yet-perpendicular-overhead’, at 12 ‘the-sun-is-perpendicular-overhead’. The afternoon is usually expressed by ‘the-shadow-is-turned-round’. This phrase is often used for the period from 3 to 5 p. m. In particular, 12–2 = ‘the-sun-is-broken’, 2–4 = ‘afternoon-now’, 4–6 is evening, 5 = ‘the-sun-goes-down’, sunset glow = ‘the-twilight-follows-the-sun’. With the coming of darkness begins the tapa, which lasts until 8 o’clock, when the people usually go to rest[88]. Another authority gives the following list:—Evening, when the cattle return to the kraal just before sunset; night-fall, or the hour for gossip, before the people go to bed about 8 o’clock; then night, midnight, and the time when the buffaloes go to drink (about 4 a. m.), this latter is the hour before the sun rises; then ‘the blood-red period’ or ‘the time when the sun decorates the sky’, this is the hour when the first rays of the sun redden the heavens; after that morning, when the sun has risen. There are also hours called ‘the-sun-stands-(or is-)opposite-to-one’ (midday), and ‘the-shadows-lower-themselves’ (1–2 p. m.)[89]. The Nandi, north-east of the Victoria Nyanza, divide the day into six parts with separate names: 5–6 a. m., 6–9 a. m., 9 a. m.–2 p. m., 2–6 p. m., 6–7 p. m., night. They have moreover a highly developed terminology for the hours of the day, to which we shall return later. The Baganda distinguish the following times of day:—night, midnight, cock-crow, early dawn, morning, ‘little sun’ (early morning from 6 to 9), full or broad daylight (9–2), midday, afternoon, evening[90]. The lower classes sometimes reckon from the meal-times, breakfast at 7 a. m., dinner at noon, and supper at 6 p. m. Women engaged in rough work in the gardens spoke of the time at which such and such an event took place as that of the first or second pipe, the first marking an interval of rest at 8 a. m., the second being smoked when work ceased at 10 a. m.[91]. The expressions for the times of day among the Thonga of South Africa have been translated and explained as follows:—“The dawn is called nipandju; then come tlhabela sana, the time when the rays of the sun (sana) are piercing; hisaka sana, when they are burning; nhlekani, the middle of the sky, or shitahataka, the maximum point of heat; then ndjenga or lihungu, the afternoon; the time when the sun goes down (renga); ku pela or ku hlwa, when it reaches the horizon; and inpimabayeni, the twilight, literally ‘the time when you do not easily recognise strangers coming to your village because it grows dark’”[92]. It is remarkable here that many indications are given from the increasing heat and not from the position of the sun. The Hottentots distinguish morning and evening twilight, morning brightness, i. e. the time of clear day shortly before sunrise (the native name is given because about dawn it is usually most perceptibly cold), and evening brightness, ‘the red twilight’. ‘Little children’s twilight’ was in some places the name given to the time of the first noticeable diminution of light after sunset, in accordance with the belief that at this hour most children were born. Afternoon and morning were only approximate. A distinction was made between evening and late evening, which extended till long after sunset[93]. The author just quoted remarks that in this case one is struck by the fact that while the limits of day and night are elaborately marked out, of the hours of day itself only noon is brought into prominence. The same is the case with most peoples who possess a more highly developed terminology of this nature, and the circumstance is perfectly natural, since the concrete differences in the phenomena of light and of the heavens become so great and so easily visible during the transition from day to night and night to day. As soon as the sun has risen a little in the heavens these differences consist chiefly in the position of the sun and in the increasing heat. Here the language of signs is really more expressive.

The aboriginals of the Andaman Islands have terms for the following times of day:—dawn, the time between this and sunrise, sunrise, the time between sunrise and 7 a. m., morning (three different expressions), noon, the time from noon to 3 p. m., from 3 to 5, from 5 to sunset, sunset, twilight, from night-fall to midnight, midnight[94]. In Busang (the common commercial language of the Bakau) as spoken by the Mendalam Kayan of Borneo the different times of day are named:—dow (day) bekang (open, split) = 6 a. m.; dow njirang (to shine) mahing (powerful) = about 9 a. m.; dow negrang (upright) marong (real) = about 12 noon; dow njaja (great) = about 4 p. m.; dow lebi (little) = about 6 p. m.[95] The terms used by the Islamite Malayans of Sumatra are mingled with Arabic loan-words, which I indicate by (Ar.):—6 a. m. (Ar.) dawn, 9 ‘half of the rising’, 11 ‘close to noon’, 12 ‘middle of the day’, 12–1 p. m. (Ar.), 1–3 ‘mid-descent’, 3 ‘the time of the long sinking’, 4 (Ar.) afternoon, 5.₃₀ ‘time of twilight’, 6 (Ar.) sunset, 8 (Ar.) evening[96]. The Javanese speak of morning, forenoon, noon, afternoon, fall of the day, sunset, evening[97]. The Achenese of Sumatra, who have a fully developed calendar influenced by Arabic, keep the old names for the times of day but with Arabic words and the Moslem hours of prayer intermingled. About 6 a. m. = with the breaking forth of the sun; 7–7.₃₀ = the sun a pole high, referring to the poles used in propelling craft; 9 = rice time, i. e. meal time; 10 = the loosening of the ploughing-gear; 11 = the approaching of the zenith; 12 = the zenith; 12.₃₀ = the falling from the zenith; 1.₃₀–2 = the middle of the period devoted to obligatory noon-day prayers; 3 = the last part of this; 3.₃₀ = the beginning, 4.₃₀–5 = the middle, and 5.₃₀ = the last part of the time of afternoon prayers; 6 = sunset; 7.₃₀ = evening, especially referring to the time of commencement of the evening prayer; then come midnight and the last third of night, 3 a. m. = the single crowing of the cock, 4–4.₃₀ = the continuous crowing of the cocks, nearly 5 = the streaks of dawn[98]. For the Malays of the Peninsula the following list is given:—just before dawn = before the flies are astir; after sunrise = the heat begins; about 8 a. m. = when the dew dries up; about 9 = when the sun is half-way above. Then follow:—when the plough rests; noon = just noon, right in the middle, when the shadows are round; afternoon = when the day turns back; about 1.₃₀ p. m. = after (Friday) prayer; about 3 = when the buffaloes go to water; about 10 = when the children have gone to sleep[99].

The natives of the Solomon Islands have a rich terminology. In Buin the following degrees of brightness in the daylight are distinguished:—4 a. m., ‘it gradually begins to get light’; 5, ‘the brightness is coming on’; 6, ‘the sun shews himself’; 7, ‘it is getting sun’, ‘the sun is there’; 10, ‘the sun is over the side-rafters of the roof’ (i. e. not yet quite overhead); 12 noon, ‘the sun has come overhead’; 2 p. m., ‘with westerly inclination’, ‘turning’; 3.30, ‘it has come to the tying of the knot’ (on the Gazelle Peninsula they say of this time ‘the sun has sat down to glow’); 5, ‘darkness is drawing near’; 6, ‘it has begun to get dark’; 7, ‘it has grown dark’[100]. Moreover there are words and expressions which mean ‘middle of the heavens’, ‘the sun is over the ridge’, ‘the sun stands below 70° from the horizon’, ‘the sun is on the entrance-beam’[101]. A feature of special note here is that the houses (which must all be built facing the same direction) and their parts serve as aids in indicating time. The inhabitants of New Britain (Bismarck Archipelago) divided up the day according to the position of the sun, and had words for sunrise, noon, afternoon, the time of the declining sun, nearly sunset, sunset, and presumably some others[102].

The Polynesians mingle the time-indications based on the position of the sun with others which are derived from the life of men and nature. We are told that the Hawaiian day was divided into three general parts, 1, breaking the shadows, 2, the plain, full day, 3, the decline of the day. But this must be completed by what follows:—The lapse of night, however, was noted by five stations: 1, about sunset; 2, between sunset and midnight; 3, midnight; 4, between midnight and sunrise; 5, sunrise[103]. A native Hawaiian writes:—“When the stars fade away and disappear, it is ao, daylight; when the sun rises, day has come, la; when the sun becomes warm, morning is past; when the sun is directly overhead it is awahea, noon; when the sun inclines to the west in the afternoon, the expression is wa ani ka la. After that come evening, ahi-ahi (ahi, fire), and then sunset, napoo ka la, and then comes po, the night, and the stars shine out”. Other expressions are translated:—‘there comes a glimmer of colour on the mountains’, ‘the curtains of night are parted’, ‘the mountains light up’, ‘day breaks’, ‘the east blooms with yellow’, ‘it is broad daylight’[104].

These are, poetically regarded, very fine examples of the rich terminology for the time of transition between night and day. In Tahiti the day has six divisions which are fairly accurately determined by the height of the sun. Names are given for midnight, midnight to daybreak, daybreak, sunrise, the time when the sun begins to be hot, when it reaches the meridian, evening before sunset, the time after sunset[105]. The names for the times of day among the Society Islanders were particularly well developed. For the day there were two expressions according to its extension either from morning to evening twilight or from the rising to the setting of the sun. No division into regular periods was known, nor any means of establishing these; nevertheless the islanders distinguished a varying number of points of time, according to recurring physical changes, at unequal distances from each other. Thus:—the time of cock-crow, the first breaking of clouds, twilight, the stirring of the flies, the time at which a man’s face can be recognised, daylight, the dipping forward of the sun’s edge, sunrise, the sun above the horizon, the rays broadening over the land, the rays falling on the crown of the head, the same a little oblique, the shadow as long as the object, the same longer than the man, the sun near the horizon, sunset, the time at which the houses are lit up, twilight, night, midnight[106]. For the Marquesas are given:—daybreak, twilight, dawn, (‘the day or the red sky, the fleeing night’), broad day—bright day from full morning to about ten o’clock—, noon (‘belly of the sun’), afternoon (‘back part of the sun’), evening (‘fire-fire’, the same expression as in Hawaii, i. e. the time to light the fires on the mountains or the kitchen fire for supper)[107]. The Samoans divided the day into first dawn, dawn, cock-crowing, day-break, the time when the bird iao was heard (i = call, ao = day-break), morning, the time to feed the tame pigeons (about 9 a. m.), the sun upright (= noon), half-way down (about 3 p. m.), sunset. After that the night was divided into:—the crying of the cricket (about 20 minutes after sunset), fire-lighting (about half-an-hour later), the extinguishing of the lights (about 9 p. m.), midnight, and tulna o pa ma ao, ‘the standing together of night and day’[108].

Indications of this nature are convenient only in countries in which the sun is neither too often nor too long hidden by clouds. When the sun is hidden the inhabitants have to manage as best they can. A very interesting statement in this connection is made by a Swahili native. In rainy days his tribe observed the crowing of the cock. At the first cock-crow they knew that it was 5 or 6 a. m.; when the cock failed to crow all sense of a division of time was lost to them[109].

The phenomena of Nature afford little basis for the naming of the times of day, since there is hardly one of them which recurs regularly every day at a definite time, with the exception of cock-crow, which is in great favour as an indication of the time before sunrise. Other exceptional cases are such names as that mentioned for the Society Islands, ‘the stirring of the flies’; one given for the Mahakam Kayan of Borneo, tiling (a cricket which is only to be heard at sunset) duan (to sing)[110]; a couple of expressions of the Wadschagga, ‘the cry of the partridge’ in the evening, ‘the turning of the smoke down the mountain’[111]; and one of the Nandi, ‘the elephants have gone to water’[112]. But a people which devotes itself to cattle-rearing or to agriculture may borrow from its regular daily occupations expressions for the times of day. Thus the Mahakam Kayan, besides the above-mentioned name for late afternoon and the term for noon (beluwa dow, ‘half-day’), have an expression for about 4 p. m.—dow uli, i. e. ‘the time of the home-coming from work in the fields’. The Javanese are strongly influenced by civilisation and have, especially for astrological purposes, a fully developed chronological system; not seldom, however, the times of day are given in relation to the rural labour. So they say ‘when the buffalo is sent to the pastures’, ‘when the buffalo is brought back from the pastures’ or ‘is housed’ etc.; but for the time of the occurrence of any event the position of the sun is usually indicated[113]. The Achenese and the Malays of Sumatra have an expression exactly corresponding to the Greek βουλυτός[114]. The Wadschagga have expressions for the position of the sun, but also others[115], among which may be mentioned ‘the first going of the oxen to the pastures in the morning’. This kind of terminology seems to have been developed into a system among the Banyankole, a cattle-raising tribe of the Uganda Protectorate. The day is divided up in the following way:—6 a. m., milking-time; 9 a. m., katamyabosi, not translated; 12 noon, rest for the cattle; 1 p. m., the time to draw water; 2 p. m., the time for the cattle to drink; 3 p. m., the cattle leave the watering-place to graze; 4 p. m., the sun shews signs of setting; 5 p. m., the cattle return home; 6 p. m., the cattle enter the kraal; 7 p. m., milking-time[116]. This terminology is of especial interest since it remains in various Indo-European languages as a relic of antiquity, and affords a hitherto little observed piece of evidence for the life of antiquity which agrees well with others. Compare Sanskrit sagavás, the time when the cows are herded together; βουλυτός, the time when the oxen were unyoked in the Homeric phrase ἦμος δ’ ἠέλιος μετενίσσετο βουλυτόνδε[117]; and Irish im-buarach, morning, ‘at the yoking of the oxen’. With rest or meal-times are associated Old High German untorn, ‘noon’, the time of the mid-day rest, Sanskrit abhipitvam, ‘evening’, and Lithuanian piëtus, ‘noon’, which goes back to Sanskrit pitus, ‘meal-time’[118].

Time-indications of various kinds are, as we have seen, used alongside of one another; when they are fully employed a very highly organised terminology for the times of day may be arrived at. The names for the times of day among the Nandi seem almost artificial:—2 a. m., the elephants have gone to the waters; 3, the waters roar; 4, the land (sky) has become light; 5, the houses are opened; 5.₃₀, the oxen have gone to the grazing-ground; 6, the sheep have been unfastened; 6.₃₀, the sun has grown; 7, it has become warm; 7.₃₀, the goats have gone to the grazing-ground; 9, the goats have returned from the grazing-ground; 9.₃₀, the goats sleep in the kraal; 10, the goats have arisen, the oxen have returned; 10.₃₀, the oxen sleep; 11, untie the cattle, i. e. let the calves get their food, the goats feed; 11.₃₀, the oxen have arisen; 12 noon, the sun has stood upright, the goats sleep in the wood; 12.₃₀, the goats have drunk water; 1 p. m., the sun turns, i. e. goes towards the west, the cattle have drunk water; 1.₃₀, the drones hum; 2, the sun continues to go towards the west, the oxen feed; 3, the goats have been collected; 4, the oxen drink water for the second time, the goats have returned; 4.₃₀, the goats sleep; 5, the eleusine grain has been cleaned for us, take the goats home, shut up the calves; 5.₃₀, the goats have entered the kraal; 6, the sun is finished, the cattle have returned; 6.₁₅, milk (sc. the cows); 6.₄₅, neither man nor tree is recognisable, cattle-fold doors have been closed; 7, the heavens are fastened; 8, the porridge is finished; 9, those who have drunk milk are asleep; 10, the houses have been closed; 11, those who sleep early wake up; 12, the middle of the night[119].

As a last example I give the most detailed list of all, from the neighbourhood of Antananarivo, the capital of Madagascar. The times given are naturally to be taken on the average. 12 midnight, centre of night or halving of night; 2 a. m., frog-croaking; 3, cock-crowing; 4, morning also night; 5, crow-croaking; 5.₁₅, bright horizon, glimmer of day, reddish east; 5.₃₀, the colours of cattle can be seen, dusk, diligent people awake, early morning; 6, sunrise, day-break, broad daylight; 6.₁₅, dew falls, the cattle go out; 6.30, the leaves are dry (i. e. the dew disappears); 6.₄₅, the hoar-frost disappears, the day chills the mouth (this applies only to the two or three winter months); 8, advance of the day; 9, (the sun is) over (at a right angle with) the purlin; 12 noon, over the ridge of the roof.—In the forenoon the position of the sun nearly square with the eastern purlin of the roof marked about 9 o’clock; and as noon approached, its vertical position about the ridge-pole, or at least its reaching the meridian, clearly indicated 12 o’clock. In regard to the terms for the afternoon we must bear in mind that the houses in former times were always built with their length running north and south and with the single door and window facing the west; the sunlight coming in after midday at the open door by its gradual progress along the floor gave a fairly accurate measure of time. The house therefore served, as among the Dyaks, as a kind of sun-dial.—12.₃₀ p. m., day taking hold of the threshold; 1, peeping in of the day (into the room), day less one step; 1.₃₀—2, slipping of the day, decline of the day, afternoon; 2, (the sun) at the rice-pounding place (i. e. the sunbeam falls on the rice mortar), at the house-post (there were in the house three posts supporting the ridge: in the southern one there were notches, jinja andry, from which the advance of the sunlight and of the day was observed); 3, at the place of tying the calf (as the rays reached the one of the posts to which the calf was tied at night); 4, at the sheep- or poultry-pen; 4.₃₀, the cow newly calved comes home; 5, the sun touching (i. e. when the declining sunshine reached the eastern wall of the house); 5.₃₀, the cattle come home; 5.₄₅ sunset flush; 6, sunset (lit. ‘sun dead’); 6.₁₅, the fowls come in; 6.₃₀, dusk, twilight; 6.₄₅, the edge of the rice-cooking pan is obscure; 7, people begin to cook rice; 8, people eat rice; 8.₃₀, finished eating; 9, people go to sleep; 9.₃₀, everyone in bed; 10 gun-fire; 12, midnight[120].

Finally I collect the Homeric expressions for the parts of the day. They are far from being so elaborately organised as the examples quoted above, and many are incidental periphrases; the terminology is still at its beginnings. Its character is quite primitive also in the juxtaposition of terms of different kinds. The day is divided into the familiar three parts. ‘It will be a dawn, or an afternoon, or a noon when I am to be killed’, says Achilles[121]. The meaning of ἠώς, ‘dawn’, is also extended so that the word can denote forenoon or at least morning. Cp. the following phrases:—‘I slept the whole night and to the dawn and to the noon’,[122] ‘as long as it was dawn and the holy day increased’[123]; of this the phrase already quoted, ‘as the sun turned over to the unyoking of the oxen’, is the counterpart. In this sense appears also the derivative ἠοίη. When Menelaus wishes to surprise the Old Man of the Sea he goes to the seashore ‘as the dawn appeared’[124]: the Old Man is said to come ‘as the sun ascends the middle of the heavens’[125]. Thus ‘we waited the whole dawn’ until ‘the Old Man came up from the sea at noon’[126]. The afternoon, in which the suitors amuse themselves with dance and song, is also called eventide[127]; when evening, ἕσπερος, comes, they go home to sleep[128]. Besides these larger divisions smaller ones were also indicated, e. g. the morning twilight, ‘when it was not yet dawn but still the twilight of the ending night’[129]. Before dawn there appears the morning star, ἑωσφόρος, Il. XXIII, 226, Od. XIII, 93. ἠώς, ‘dawn’ in the proper sense of the word, is often used as a time-indication, sometimes in the well-known periphrastic expressions of Il. XI, 1, XIX, 1, Od. V, 1. XXIII, 347, XXII, 197, sometimes alone, e. g. ‘at dawn’, ‘at the appearance of dawn’[130]. Sunrise is always indicated by verbal and often periphrastic expressions, simply by ἀνιέναι, ‘rise’[131], further ‘the sun, leaving the fair sea, rose into the all-brazen heaven to shine for the immortal ones’ etc.[132], and ‘neither as he ascends to the starry heaven nor as he again turns back to the earth from the heavens’[133], similarly Od. XII, 380 ff., Il. XI, 735 ‘as long as the shining sun rose above the earth’[134], and Il. VII, 421 ff. ‘the sun thereafter once more struck the fields, ascending in the heavens from the deep and soft-flowing ocean’[135]. The expression can therefore also include the time immediately following after sunrise, but is not applied to the whole period of the sun’s ascension, i. e. the forenoon. The culmination of the sun is mentioned in Od. IV, 400 (cp. above) and in Il. VIII, 68. The decline of the day is thus described, ‘the day was for the greater part gone’[136]; for the sinking of the sun see Od. XI, 18, XII, 381 (cp. above), and the already quoted expression ‘the sun turned over to the unyoking of the oxen’. Sunset (Il. XVII, 454, XVIII, 241, Od. II, 388) is described by the common word δύνειν, ‘set’, or by ‘goes under the earth’[137], or ‘the bright light of the sun sank down in the ocean, drawing after himself the dark night’[138]. The evening star has the same name as evening, ἕσπερος[139]. The Homeric Greeks therefore do not seem to have observed the position of the sun in any but the most general fashion. We may add certain indications taken from the business of daily life. The word βουλυτός (cp. above p. 31) appears in the twice-recurring verse ‘as the sun turned over to the unyoking of the oxen’[140]. It is not the sun but the ploughman that unyokes the oxen: the word has therefore become established as a chronological terminus technicus which is significant on account of its antiquity. About the expression ἐν νυκτὸς ἀμολγῷ there has been much dispute. It occurs:—Il. XI, 173 and XV, 324, where lions surprise a herd, XXII, 28, in the simile of the morning rising of Sirius, 317, of the shining forth of the evening star, Od. IV, 841 ‘so clear appeared the dream to her’[141]: it is a well-known fact that we dream for the most part shortly before waking. The sense ‘beginning or end of night’ is therefore fully confirmed. As for the etymology I do not hesitate to pronounce in favour of that lying nearest to hand, viz. ἀμέλγειν, ‘to milk’, and therefore ‘milking-time’. Compare the terms of the Banyankole for early morning at 6 o’clock and evening at 7—‘milking-time’—and those of the Nandi: 6 p. m. ‘the sun is over, the cattle have come back’; 6.₁₅, ‘milk’ (sc. the cows). That only these two expressions have settled into termini technici admits of a not unimportant conclusion in regard to antiquity. The meal-hour as an indication of time occurs Il. XI, 86, ‘when a wood-cutter prepares his meal after having fatigued his arms by felling large trees’[142], and Od. XII, 439, ‘when a man rises from the market-place to go home to the meal after having judged many quarrels’[143],—in the latter instance in connexion with the market. This time-indication was destined to have a great future as the social life of the Greeks developed. Phrases such as the following are of common occurrence:—‘when the market-place is full’[144], ‘before the market-place has filled itself’[145], ‘the breaking up of the assembly of the market-place’[146], etc. The night was divided into the familiar three parts (although the expression μέση νύξ, ‘middle of the night’, first occurs in the smaller Iliad) and was judged according to the position of the stars:—‘Let us go, for the night draws close to an end and the dawn is near. The stars are far gone. The greater part of night is gone, the two parts, only the third part remains’[147]; ‘when it was the third part of the night and the stars had passed’[148]. The morning star serves as a time-indicator at the nocturnal home-coming of Odysseus[149].

The Latin expressions I merely copy from Censorinus, Ch. 24, and insert in brackets the additions made by Macrob., Sat. I, 3, 16 ff. Tempus quod huic—i. e. nox media—proximum est vocatur de media nocte (media noctis inclinatio), sequitur gallicinium, cum galli canere incipiunt, dein conticinium, cum conticuerunt; tunc ante lucem, et sic diluculum, cum sole nondum orto iam lucet. Secundum diluculum vocatur mane cum lux videtur sole orto, post hoc ad meridiem, tunc meridies, quod est medii diei nomen, inde de meridie (inde—i. e. a meridie—tempus occiduum), hinc suprema ... post supremam sequitur vespera ante ortum scilicet—this must be before the appearance of the star—eius stellae, quam Plautus vesperuginem ... appellat. There are also ortus and occasus solis, crepusculum. This terminology is poor and applies almost exclusively to the daylight. In ancient Rome the edifices of the Forum are said to have served as sun-dials. A servant of the consul proclaimed noon “when the sun peeped between the Rostra and the Graecostasis; when the sun sank from the Maenian column to the prison he proclaimed evening, but only on clear days”[150]. With the advance of civilisation the Greek terms for the twelve hours of the day, each of which varied in length according to the time of the year, became customary, a fact which is connected with the spread of sun- and water-clocks[151]. Hence arises in the Middle Ages the terminology derived from the daily mass (hora canonica)[152]. In daily life there was often a recurrence to primitive methods. I borrow a few examples of a quite primitive character from the early medieval tract Peregrinatio Aetheriae:—‘the hour when people can recognise each other’[153], ‘when the crow of cocks begins’[154], ‘from the first cock-crow’[155], etc., but also hora tertia, quinta, sexta (noon).

An obviously isolated method is the determination of the times of day from the daily twice-recurring ebb and flow of the tides; the method is also very unsuitable, since the period of a complete tide is 12 hours 25 minutes, so that the two periods together exceed the day by nearly an hour. In fact the Eskimos of Greenland are the only people who reckon by the tides. They divide up the day according to ebb and flow, although they must always reckon differently on account of the variations of the moon[156]. Dalsager[157] also points this out and remarks that their reckoning cannot last for two consecutive days, so that they have to make a fresh division every day. The rudiments of this method are however seen among some of the tribes of Polynesia. Immediately after the above-quoted divisions of the day among the Society Islanders are mentioned “the longer periods before noon and midnight during which the sea rises, and the others following these, in which it falls”[158], and “night or the light quite gone, when the sea begins to flow towards the land, about 11 at night”[159]. The Hawaiians called the rising of the tide by such names as the rising, big, full, and surrounding sea; when the water neither rose nor fell it was called the standing sea; the ebbing sea they spoke of as the parted, retiring, and defeated sea[160].

The night is the time of complete darkness and rest, and therefore the frequently mentioned expression, ‘sleeping-time’, corresponds to night. Seldom is the whole time during which the sun remains below the horizon to be understood by it. On the Society Islands there were two expressions for day according to its extension from morning to evening twilight or from sunrise to sunset[161]. The Hawaiian judge, Fornander, follows this mode of speech when he distinguishes five periods of night, (1) about sunset, (2) between sunset and midnight, (3) midnight, (4) between midnight and sunrise, and (5) sunrise[162]. For the times between sunset and night-fall and between night and sunrise there is a rich terminology which has already been illustrated. During the night itself time-indications are for obvious reasons scanty. Often the only point distinguished is midnight, e. g. by the Kiowa[163], the Masai[164], the Shilluk[165]; ‘the silence of the land’ among the Babwende[166], ‘the back of night’ among the Hottentots[167], ‘the time of sleep’ among the Hawaiians[168]. Hence arises of itself a threefold division in which the periods of night before and after midnight are distinguished, as e. g. by the Hawaiians[169]. The usual method is to start from the day, i. e. the limit of the day, and then to proceed on both sides in the direction of midnight, as in the late evening of the Hottentots, which extends till long after sunset[170], and the ‘not yet early’ and the tara (beginning at dusk and extending till the time of rest) among the Masai[171], etc. The Tahitians are credited with six divisions of the day and as many of the night, this more accurate division of night being of course determined by the stars[172]; the only expressions reported however are those for midnight and the time from midnight to daybreak[173]. On the Marquesas Islands the first night-watch was ‘the hour of ghosts’; the advanced night was termed ‘black night’, and midnight ‘great sleep’; the last watch of night was ‘the coming of day’[174]. The Wadschagga have three night watches:—the awakening in the evening, that in the middle (midnight), and that in the morning twilight[175]. The Javanese have night, midnight, and waning of night[176]. Where the cock is kept, its crow serves as a sign that the night is drawing to an end, as for instance among the Swahili[177], and in the Dutch Indies[178]; the Yoruba distinguish other cock-crowings, such as ‘the cock opening the way’, i. e. the first cock-crowing, ‘the time of the cock-crowing immediately before sunset’[179]. Quite exceptional however is the device ascribed to the inhabitants of the New Hebrides. In order to denote the hours of the night they make a gesture in the direction of the spot where the sun would be at the corresponding hour of day[180].

There is only one means of accurately indicating the times of night, and that is by the observation of the stars. Many peoples judge from the position of the morning-star the time that has yet to elapse before sunrise: but this cannot always be done, and in any case the method is only of use in the early morning. But the fixed stars are always there. The difficulty however arises that every day the stars gain about four minutes on the sun; the stars must therefore be accurately known, and the observer must either be acquainted with their positions at definite times of the year or else be constantly choosing a new star as his chronometer. Not many peoples have got so far as that. Although the science of astronomy was very well developed among the Polynesians, we are told of the Tahitians that to distinguish the hours of night by means of the stars was a science with which very few of them were acquainted[181]. On the Society Islands the advance of night was determined from the stars[182]; and so in Hawaii, with as great accuracy as the hours of the day from the sun[183]. “When the Milky Way passes the meridian and inclines to the west, people (in Hawaii) say ‘the fish has turned’”[184]. Among the Indians of South America the knowledge of the stars is very wide-spread. E. Nordenskjöld, who visited the border districts where Brazil, Bolivia, and the Argentine meet, says repeatedly that the stellar heavens are the Indian’s clock and compass. When sitting in their huts they can, without looking out, indicate the positions of the more important constellations in the sky. If one is out with an Indian at night he will point to Orion or some other constellation and shew how far it will have moved on before the end of the journey is reached[185]. The Eskimos of Greenland, when it is dark, indicate the time from nelarsik (Vega)[186], or from the Pleiades[187]. Among them the observation of the stars is uncommonly well developed. The Lapps, who have to tend their reindeer during the long winter nights, determine the course of time by certain stars. Sarvon is the largest star in the heavens: when in winter it stands in the middle of the sky it marks midnight; it is called the night-clock of the Lapps. The Great Dog, the Old Man, and the Old Woman are three stars that pursue sarva. They rise when the people go to sleep, and set a little before daybreak. They ascend the heavens obliquely in front of sarva, in the morning they dip downwards. Another authority states that sarva is the Great Bear; the first couple of stars in it are the Old Man and the Old Woman, the second the Dog and the Elk. The reindeer herdsman decides from it how far night is advanced, and when he may expect to be relieved. Lovosj or suttjenes is the name given to the Pleiades. The constellation indicates midnight, when the weather is good. A fable tells how this constellation saved a servant who had been driven out by his master into the great cold of a winter night. The young men wish the maidens to tend the reindeer by night and say:—“Go and kiss the suttjenes young men”, but the maidens answer:—“Go yourselves and kiss the suttjenes maidens”[188]. Of the old Icelanders Kålund writes:—“At night the moon and certain stars, especially the Pleiades, afford them the same aid” (i. e. as the signs of day)[189]. The Homeric Greeks—at least in a general fashion—also judged of the advance of night by the position of the stars[190]. This more accurate method is therefore peculiar to a few primitive peoples specially gifted in astronomy.

From the investigation of the modes of naming and reckoning the day and its parts it follows for primitive time-reckoning in general that the time-indications refer to concrete phenomena, and therefore either they indicate a point of time or, if they are related to periods, these periods are of different and fluctuating length. They are accordingly of no use in calculating, they cannot simply be added together even when a number of such periods together make up the period of a complete day, i. e. they are fundamentally discontinuous. When several days are to be counted the pars pro toto method is used: instead of the whole day a part is counted. Within the day two phenomena chiefly recur with such unfailing constancy as to be of use in counting: they are the daily reviving sun and the night or sleeping-time. The word for sun is often the same as that for day. Within the day fall a number of occupations which chiefly turn the attention to its length and varying phenomena, and this is the case also with the sun itself, for the varying position of the sun in the heavens affords the most usual mode of indicating the time of day. For the counting a point of time is best suited, or, which comes to the same thing, a unit without subdivisions, a blank period. This is the reason why the counting by ‘sleeps’ or nights predominates. On the same grounds the quite isolated pars pro toto counting of the days from the dawns in Homer may be explained. To indicate the duration of time primitive peoples make use of other means, derived from their daily business, which have nothing to do with time-reckoning; in Madagascar ‘rice-cooking’ often means half an hour, ‘the frying of a locust’ a moment[191]. The Cross River natives say:—‘The man died in less than the time in which maize is not yet completely roasted’, i. e. less than about 15 minutes; ‘the time in which one can cook a handful of vegetables’, i. e. an hour[192]. The Malays, the Javanese, and the Achenese use the following expressions for a period of time:—a blink of the eyes (literally), the time required for chewing a quid of sirih (about 5 minutes), the time required for cooking a kay of rice (about half an hour), for cooking a gantang of rice (about an hour and a half), half a day, a ‘sun-dark’, i. e. a complete day and night[193]. The natives of New Britain (Bismarck Archipelago) measure the time between sunset and the moon-rise by the smouldering of a torch or the time occupied in cooking yams, taro, or wild taro. Short divisions of time were also expressed by comparative terms, e. g. the throwing of a stick for a short distance, ‘a woman’s crossing’, or the distance a woman would paddle[194]. Very often duration of time is indicated by reference to the time needed to traverse a well-known piece of road between two places. Examples are superfluous. But all these indications of periods of time are found among more developed peoples: the primitive peoples pay little or no attention to them.

Both in the case of the day and in that of the other time-units this clinging to a natural basis long proved a hindrance to a rational system of time-reckoning, which could only be achieved by breaking away from natural phenomena. For there are no fixed natural limits of day, but if morning and evening, or still more clearly sunrise and sunset, are chosen as the limits, these must change every day and the days will vary in length. Here the midnight period proved of assistance, since it facilitated the establishing of a fixed point of divergence. This was done in Rome, and the practice had its root in daily life, where in order to indicate the time of occurrence of events which took place in the night-time the calculation was pushed forwards on both sides towards midnight, until this became the limit of divergence. It is however an artificial epoch that must be found by calculation[195].

In the second place the hour of antiquity is a twelfth part of the whole time of daylight, and this duodecimal division was also transferred to the night, which had commonly been divided into four watches according to the practice borrowed from military life. This hour therefore varied in length according to the time of the year. The inconvenience of a varying division of this nature must have made itself felt in daily life, although in the south it was not so insupportable as it must have been in the north. It rendered the construction of the clock difficult, and above all was impracticable for scientific astronomy. Hence alongside of it appeared even in antiquity the hour of constant length or the double hour, viz. a twelfth or a twenty-fourth part respectively of the complete day. The double hour, notwithstanding Bilfinger’s assertion to the contrary, arose in Babylon (kasbu), and is connected with the duodecimal division of the zodiac[196]. This hour of constant length was not generally adopted until very late: the varying hour remained almost up to the end of the Middle Ages. Our modern hour has only been in general use since about the 14th century, when it was first spread by the construction of the striking-clock[197]. Its convenience for the business of practical life and the construction of the clock together secured the victory of the hour as 1/24th of the day, originally a numerical and astronomical division. A condition for its use was the fusion of day and night into one unit, since as long as these were kept separate the constant hour could not thrive. Both the complete day and its regular divisions however only won their way after a very long time, because men were unwilling to depart from the natural basis in time-reckoning. The substitution of the artificial for the natural time-reckoning has also, as far as the day is concerned, created a rational system of reckoning which has borrowed from the natural system only one feature, viz. the average length of the complete day.

CHAPTER II.
THE SEASONS.

The year is for us a numerical quantity of 365 or 366 days. But we speak of the year in two senses, first as the calendar year beginning on New Year’s Day, and secondly as the current year, a period of the same number of days beginning at one chosen day, as for instance in giving a person’s age. The word ‘year’ may however also represent the highest chronological unit even independently of the seasons, as in the case of the Egyptian shifting year of exactly 365 days, and the Islamite lunar year of 354. These however are exceptional cases. At the basis lies the natural year conditioned by the course of the sun and by the natural phases dependent thereon, which penetrate closely into the life of man. This connexion has necessitated the agreement of the numerical year with the sun, whence arises a situation very inconvenient for reckoning, namely that years of a varying number of days have to be accepted, since the natural year does not contain a whole number of days.

The year as a numerical quantity is only the tardily attained summit of development, and the connexion with the natural year has always been so strongly felt that, except in certain cases such as the Egyptian and Islamite years, the chronological year has had to adjust itself accordingly. Here also we see the point of departure, the natural phenomena which are in the end dependent upon the course of the sun, such as the variation between heat and cold, verdure and snow, rainy season and drought, the blooming and withering of vegetation, between the different trade-winds or monsoons, between abundance and scarcity of food. With these and similar concrete phenomena the time-reckoning is from its origin bound up, and is at first discontinuous, i. e. it fixes the attention solely on the phenomena in question, and not on the year as a whole. The fusion of the various seasons into the circle of the year is arrived at only by degrees: the year is at first counted by the pars pro toto method. The process is therefore similar to that already found in the discussion of the day.

It must be granted as a premise to our investigation that when we speak of ‘seasons’ not only the larger divisions of the year are to be understood by the word—those which alone of all the natural epochs of the year are current among us to-day—but also smaller divisions which might perhaps be called seasonal points; for instance the times of cherry-blossoming and hop-picking are also seasons. Such short—often very short—seasons are not distinguished in any important feature from the longer: the difference only arises from the longer or shorter duration of the phenomena in question. The Hidatsa Indians describe any period thus marked by a natural occurrence, be it long or short, the hot season or the season of strawberries, by the same word, kadu, ‘season’, ‘time’ (of the occurrence), and the longer seasons include shorter[198].

We begin with these shorter seasons since they are more foreign to us: to primitive man however they are of extreme importance, since in the absence of a regular calendar they afford the only means he knows of determining the shortest periods of the natural year, in so far as they are connected with this. A time-determination of this nature is important not so much for giving the date of any occurrence as for establishing beforehand the time of certain occupations, e. g. sowing or a festival.

The classical instance is afforded by the peasants’ maxims of Hesiod. The cry of the migrating cranes shews the time of ploughing and sowing[199]. If one sows too late, the crop may still thrive if Zeus sends rain upon it on the third day after the cuckoo has called for the first time in the leaves of the oak (486). Before the appearance of the swallow, the messenger of spring, the vines should be pruned (568). But when the snail climbs up the plants there should be no more digging in the vineyards (571). When the thistle blossoms and the shrill note of the cicada is to be heard, summer has come, the goats are at their fattest, and the wine is at its best (582). The sea can be navigated when the fig-tree shews at the end of its branches leaves which are as big as the foot-prints of the crow (679). Especially well-known and beloved as a sign that the hard winter was over was the swallow: evidence is afforded by the famous procession of the Rhodian swallow-youths[200], and by a vase-decoration clearly expressing the delight felt at the appearance of the herald of Spring[201]. The observation of the birds of passage was very useful for this kind of time-determination: Homer already knows it, ‘when the cranes flee the winter’, he says[202], so also Theognis: “I hear, son of Polypais, the voice of the shrill-crying crane, even her who to mortals comes as harbinger of the season for ploughing”[203]. Aristophanes makes his birds boast of it:—

“All lessons of primary daily concern

You have learnt from the Birds, and continue to learn.

Your best benefactors, and early instructors,

We give you the warning of seasons returning.

When the Cranes are arranged, and muster afloat

In the middle air, with a creaking note,

Steering away to the Libyan sands,

Then careful farmers sow their lands;

The crazy vessel is hauled ashore,

The sail, the ropes, the rudder, and oar

Are all unshipped, and housed in store.

The shepherd is warned, by the Kite reappearing,

To muster his flock, and be ready for shearing.

You quit your old cloak at the Swallow’s behest,

In assurance of summer, and purchase a vest”[204].

Similar time-determinations from natural phenomena are still not entirely neglected by the modern peasant. In Bohuslän (W. Sweden) the sowing-time was at hand when the swallow had come, it was the right sowing-time when the juniper flowered. In northern Scania (S. Sweden) the barley was to be sown when the hawthorn was in bloom. Older people could not give their birthdays but only knew that they were born e. g. at the rye- or potato-harvest, when the cattle were first driven out to pasture (in the spring), etc. My father knew quite well that his birthday was the fifth of September, but when anyone asked him when he was born he would generally answer: ‘When they pick hops’. The Eskimos said that such and such a person was born when eggs were collected or seals caught[205]. From modern Palestine a bond is quoted in which a sum of money was to be paid when next the fakûs (a kind of cucumber) was ripe[206].

We return to the primitive peoples and give first a few examples in which a natural phenomenon serves as the sign of the beginning of one of the longer divisions of the year or of some occupation, generally agriculture. Of the Bushmen we are told that they paid particular attention to the time at which the first thunder-storm broke. They hailed it with great joy since they counted it a sure sign that summer had commenced. In the midst of their excessive rejoicing they tore in pieces their garments of skins, threw them into the air, and danced for several nights in succession. The Garieb Bushmen made great outcries accompanied with dancing and playing upon their drums[207]. The Banyankole of Uganda used the euphorbia trees to guide them as to the nearness of the rainy season: when these trees began to shoot out new growth they knew that the rains were near[208]. The Indians of the Orinoco took great pains to determine the approach of the rainy season, as Gilij relates in a chapter entitled: De segni, che precedon l’inverno[209]. The signs were:—The scream of the Araguato monkeys at midnight or at the approach of day; the sudden bursting into blossom of certain trees; the swelling of the brooks, which almost dry up in summer but swell a few days before the rainy season; the yams which in summer have lost their leaves suddenly grow green again when the rainy season is at hand; finally the heliacal setting of the Pleiades. The tribe of the Bigambul in S. E. Australia reckon the seasons from the blossoming of certain trees. Yerra, for example, is the name of a tree that blossoms in September: this time of the year is therefore called yerrabinda. The apple-tree blossoms at Christmas time, which is called nigabinda. The iron-bark tree blossoms about the end of January, and this time is called wobinda. The height of summer however is named by them ‘the time when the ground burns the feet’: at this time no trees blossom[210]. The natives of New Britain (Bismarck Archipelago) determine the planting-season from the buds of certain trees and from the position of certain stars[211]. In Alu (Solomon Islands) one division of the year is determined from the bloom on the almond, another from the Pleiades[212]. The time for the sun-dance of the Kiowa Indians is determined by the whitening of the down on the cotton-plant[213]. One of the annual festivals of the Society Islands is regulated by the blossoming of the reed[214].

Instances are numerous in which phenomena like those mentioned by Hesiod serve as signs for agricultural labour. The Indians of Pennsylvania say that when the leaf of the white oak, which comes out in spring, is as large as a mouse’s ear it is time to plant maize: they note that the whippoorwill has come by then, and is constantly fluttering round them calling out his Indian name wekolis in order to remind them of planting-time, just as if he were saying ‘hacki heck’, ‘go and plant maize’[215]. Among the Thonga the period in July when the warm weather begins is called shimunu, ‘the little heat’: the mahogany and sala trees become covered with leaves, certain flowers blossom. Winter has passed away, soon the summer will come. When the Thonga woman notes these signs she picks up her hoe and sets off for the hills or the marshes to make the fields ready. In January comes nwebo, the time for the first ears of maize to ripen[216]. Among the Ba-Ronga January is nuebo, the time of the first ripe ears: great pains are taken to keep away the birds from the sorgho fields, and therefore one period is known as ‘the time when the birds are driven away’[217]. When a certain mushroom named kulat bantilong appears in large quantities the Dyaks of S. E. Borneo regard it as a sign that the time for rice-planting has come[218]; among the Malgassi the blossoming of the shrub Vernonia appendiculata in November is regarded in the same way[219]. In New Zealand plants and birds which appear at regular seasons give signs of the approach of the time to begin agricultural labours. Two kinds of migratory cuckoo, Cuculus piperatus and nitens, which appear at Christmas-time on the coasts, mark the period of the first potato-harvest. The flowering of the beautiful Clematis albida reminds the people to dig over the soil for the planting of potatoes, which is done in October[220]. According to the communication of a native, the Basutos reckon time by the changing of the seasons, the birth-times of animals, the annual variation and growth of plants, but also by the stars and the moon[221]. The most curious method is one common among the Hidatsa Indians, who reckon from the development of the buffalo calf in utero[222]. Such signs may also serve to mark off the longer seasons: the Tunguses begin summer with the time when the grayling spawns, and winter with the time when the first good squirrel is caught[223].

The examples hitherto given are only single instances intended to make clear the manner and signification of this method of indicating time. Similar starting-points for reckoning are afforded the whole year through, and as their times are fixed in regard to each other, they may form a sort of calendar. The statements made for the extremely primitive Andamanese give a very characteristic circle of occupations throughout the year, though here we have to do not with names of seasons but with the phenomena and business of the year, which our authority gives according to the European calendar. January: much honey; two kinds of wild fruit ripen and are gathered. February: two other kinds of wild fruit, also a tuber; the inhabitants of the coastal districts catch the dujong and also a few turtles; the older folk make out of bark turtle-nets, cables, and lines for harpoons. March: still another two kinds of wild fruit ripen, wild honey is abundant. April: many visits of neighbouring tribes; fruit is scanty, there is only one kind ripe, the honey is finished, the bread-fruit has not yet ripened. From May to August the ripe bread-fruit forms the principal food. In June many cases of death occur since the men in their boar-hunting expeditions in the forest sleep without shelter. In August certain white caterpillars which live in the decaying tree-trunks are a favourite dish. From August to October boats are built. In November the people are particularly merry. The turtle-catch is productive, the weather is pleasantly cool, there is little rain, and shelter is not necessary. Different tribes visit one another and feast and dance together[224].

How upon such a foundation a number of seasons may be built up is shewn by a comparison with an instructive account referring to the Eskimos of the Ungava district of Labrador. The seasons have distinctive names and are again sub-divided into a great number of shorter seasons. There are more of these during the warmer weather than in winter. The reason is obviously that the summer offers so many changes, and the winter so few. The chief events are the return of the sun, always a sign of joy to the people, the lengthening of the day, the warm weather in March when the sun has attained sufficient height, the melting of the snow, the breaking up of the ice, the open water, the time of birth of various seals, the advent of exotic birds, the nesting of gulls, eider, and other native birds, the arrival of white whales and the whaling season, salmon fishing, the ripening of salmon-berries and other species of edibles, the time of reindeer crossing the river, the trapping of fur-bearing animals, and hunting on land and water for food. Each of these periods has its special name applied to it, although several may overlap each other. The appearance of mosquitoes, sandflies, and horseflies is marked by dates anticipated with considerable apprehension of annoyance[225]. The Eskimos of Greenland reckon from the winter solstice five moons until the time when the nights become so bright that it is impossible to reckon any longer from the moon. Then they reckon by the increasing size of the young of the eider-duck and by the ripening of berries, or along the sea-coast by the departure of the tern and the fatness of the seals; when the reindeer shed the velvet from their horns they know that it is time to move into the winter houses[226].

These smaller seasons have seldom developed into an annual cycle otherwise than among some agricultural peoples[227], unless they have been fitted into the larger seasons. This is the case with the western tribes of the Torres Straits, who also determine the seasons from the stars. In the counting of the seasons they commonly begin with surlal (mid-October to the end of November). This name is given to the turtles when copulating: while in this state they float on the sea and are readily caught. The constellation known as the Shark arises. Everything is dried up, the yams are ripe. The sounding of the first thunder is the sign for planting yams. Raz (December to February) is described as ‘the time of death’, i. e. the season when the leaves die down. The first part of this season is called in Mabuiag duau-urma, ‘the falling of the cashew nuts’. There is an interval of fine weather and the wind is shifty: this coincides with Christmas-time. This is the time when the yams which have been planted begin to sprout. In Muralug this period is called malgui, which is the exact equivalent of our word ‘spring’. The next division is called dob, ‘the last of growing things’, or kusikuki, ‘medusae of the north-west’, the latter name being due to the large numbers of jelly-fish that float on the sea. The runners of the yams now grow. The time immediately after this is called purimugo, in Muralug apagap or keme. The longer season following raz is kuki, (March to May), the time when strong winds blow intermittently from the north-west, accompanied by deluges of rain, and the time of the damp heat. The appearance of the constellation dogai kukilaig (Altair, together with β, γ aquilae) heralded the beginning of this season. It has the sub-divisions kuki, kupa kuki, and gugad arai. The dry season, aibaud, forms the remaining part of the year. The south-west wind, waur, blows steadily: for this reason the first part of this period is known as waur and perhaps merits a distinctive name as much as raz. It is marked by the appearance of the constellation magi Dogai (Vega with β, γ lyrae). Food is abundant and festivals are celebrated. The divisions of aibaud are sasiwaur (‘child’, i. e. lesser south-east), piepe, tati waur (‘father’, i. e. greater south-east), and birubiru, a bird which at this time migrates from New Guinea to Australia[228].

The Kiwai Papuans who dwell on the opposite coast of New Guinea have the same star myths as the inhabitants of the Torres Straits Islands: for them, however, no smaller but only two greater seasons are mentioned[229]; on the other hand they have months[230]. The smaller seasons have clashed with the reckoning by moons, and have surrendered their names to describe the latter. They have therefore in great measure become merged in the counting of the months, which will be dealt with later. The greater seasons on the other hand, on account of their length, could not be merged in the reckoning by months, and these have therefore everywhere remained. The number of the longer seasons varies considerably, and is of course connected not only with the climatic conditions but also with the fundamental phenomena which for one reason or another attract attention; a larger season may also be divided into two or three smaller ones.

It may be taken for granted that all peoples outside the tropics, even where it has not been thought necessary expressly to mention the fact, know the two larger divisions of the year, the warmer and colder seasons. Where the plants die in winter and the trees lose their leaves, or where the snow covers the ground, this great difference becomes especially pronounced and determines the whole mode of life: but even in the sub-tropical regions it is obvious enough. To it corresponds in many parts of the tropics and sub-tropical zones the natural division into a dry and a rainy season. For the division into the summer period of vegetation and winter with its snow and ice it is superfluous to give examples: the above-quoted description of the year of the Labrador Eskimos is a typical instance. Swanton and Boas state that certain Indian tribes of N. W. America divide the year into two equal parts of six months each, summer extending from April to September, and winter from October to March[231]. The Comanches reckon by the cold and the warm seasons[232]. I give a few instances from districts in which a winter of this nature does not exist. Among the Hopi of Arizona the year has two divisions—there seems to be no equivalent to our four seasons—which may be termed the periods of the named and the nameless months: the former is the cold period, the latter is the warm. They may also be called the greater and the lesser periods, since the former begins in August and ends in March[233]. The Zuñi of western New Mexico also divide the year into two periods of six months each[234]. The Chocktaw of Louisiana have the same number of seasons[235]. The natives of Central Australia have names for summer and winter[236].

In the tropics there is often only one rainy and one dry season, with two divisions of the year corresponding to these. On the Orinoco there are summer and winter, i. e. the dry and the rainy seasons. In Maipuri the dry season is called camoti, ‘the glowing splendour of the sun’, and the rainy season canepó. Among the Tamanacho winter is called canepó, ‘rain’, ‘rainy season’, summer is vannu, ‘crickets’, since these insects chirp incessantly to the end of the season[237]. The Tupi have expressions for dry and rainy seasons but not for the year as a whole. The Bakairi reckon by the semesters of the dry and the rainy seasons[238]. The Karaya of Central Brazil reckon the year from one fall of the river to another. They thereby distinguish two seasons, the dry season when they live on the sand-banks, and the rainy season when they live on the upper banks of the river[239]. The Wagogo of E. Africa divide the year into two halves: kibahu, the dry season, about May-October, and kifugu, the rainy season, November to April[240]. So also the Nandi: iwotet, rainy season, March-August, and kement, dry season, September-February[241]; further the tribes of Loango[242], the Bantu tribes of the Congo State[243], and the Cross River negroes of the Cameroons[244]. The Tshi-speaking peoples divide the year into two periods: the smaller hohbor, from May to August, and the larger from September to April[245]. Among the Akamba the year consists of two rainy seasons separated by two dry periods: ambua anzwa, ambua ua[246]. Where this natural division prevails, however, the half-year is often put in the place of the year[247].

The Javanese have a dry and a rainy period which include six of their seasons[248], and so have the Islamite Malays of Sumatra[249]. The Polynesians divide the year throughout into two greater periods. Their seasons were in general two, the rainy season or winter, and the dry season or summer, but varied according to the situation of the particular group of islands north or south of the equator. On the Society Islands they embraced the months of May-November and November-May respectively. On the Sandwich Islands the rainy season, hooilo, lasted from about Nov. 20 to May 20, the dry season, kau, from May 20 to Nov. 20[250]. We shall find later that both seasons were named and regulated according to the visibility or invisibility of the Pleiades. Other writers also give information for Hawaii. When the sun moved towards the north, the days were long, the trees bore fruit, and the heat was prevalent: it was summer; but when the sun moved towards the south, the nights became longer and the trees were without fruit: it was winter[251]. Kau was the season when the sun was directly overhead, when daylight was prolonged, the trade-wind prevailed, days and nights alike were warm, and the vegetation put forth new leaves. Hooilo was the season when the sun declined towards the south, the nights grew longer, days and nights were cool, and the herbage (lit. vines) died away: each had six months. On Kauai Island the seasons were called mahoe-mua and mahoe-hope[252]. In Tahiti the bread-fruit can be gathered for seven months, for the other five there is none: for about two months before and after the southern solstice it is very scarce, but from March to August exceedingly plentiful. This season is called pa-uru (uru = ‘bread-fruit’)[253]. The recurring scarcity of bread-fruit shewed the changes in the course of the year, but the Pleiades afforded a surer limit[254]. In Samoa one authority gives the wet season, ending in April, and the dry season, which comes to an end with the palolo fishing in October[255]; another vaipalolo, the palolo or wet season from October to March, and toe lau, when the regular trade-winds blow, embracing the other months[256]; a third the season of fine weather—in which however much rain falls in some localities—and the stormy season, when it rains heavily[257]. The importance of agriculture is so great that the seasons in following it may sometimes depart from the changes of the climate. The Bontoc Igorot have two seasons which however do not mark the wet and dry periods, as might be expected in a country where these two periods occur: cha-kon is the season of rice or ‘palay’ growth and harvesting, ka-sip the remaining portion of the year[258]. In the New Hebrides the year is divided into two parts, the periods of yam-planting and harvesting[259].

In certain localities the atmospheric conditions are such that two divisions of the year may be distinguished according to the winds, as for instance in the Marshall Islands, where there are the months of calm and the months of squalls[260]. More commonly two seasons are given by the variation of the monsoons, as on the island of Bali, east of Java: in each case there were six homonymous months. The Kiwai Papuans have uro, the comparatively dry season of the south-east monsoon (April-December), and the time of the prevailing north-west wind, hurama, a period of alternating calms, storms of wind and rain, and thunder[261]. A native judge from the island of Vuatam in the Bismarck Archipelago remarked that the north-west trade blew throughout the time when the sun was southerly, that is from November to February, but during the time in which the sun moved in a northerly direction, May to August, the south-east monsoon prevailed. On Valam it is said that the south-east monsoon blows as long as the sun sets WNW, i. e. from May to August: from the month of November to February, when it sets WSW, the north-west trade blows[262]. In Rotuma or Granville Island near the equator periods of six months are reckoned. The west wind, which blows from October to April, serves to distinguish these two periods, although it does not affect the vegetation[263]. The people of the Nicobar Islands reckon by the south-west monsoon (November to April)[264]. The Benua-Jahun of the Malay Peninsula distinguish the half-year of the north monsoon and that of the south monsoon[265].

It would seem that the whole year might easily arise through the fusion of these two larger periods: that this is not the case will be shewn in the following chapter.

These half-years are as a rule well defined, but the natural conditions upon which they depend are subject to fluctuation, and in particular there are transitional periods the position of which cannot be certainly decided. Moreover smaller characteristic periods arise within the larger, and hence more seasons appear. Elsewhere the natural conditions are such that they directly lead to more than two seasons, e. g. where there are two different rainy seasons in the year. From these circumstances it becomes plain that a fluctuation between a larger or smaller number of seasons is possible, and indeed it often actually occurs. The seasons that adhere to natural phenomena are never clearly defined like a division of the calendar: the limits are uncertain, different seasons may be merged into one another or in part overlap one another, as has been shewn in the case of the Eskimos of Labrador.

Among the Eskimos of the Behring Strait the year is often divided into four seasons corresponding to the usual occupations, but these divisions are indefinite and irregular in comparison with the reckoning by months[266]. Of the Indians in general it is said that as a rule four seasons are recognised and have specific names applied to them (apart from the tribes that have two). In many cases however the latter may split up both summer and winter into two subdivisions: this is stated e. g. for the Chocktaw of Louisiana[267]. The Siciatl of British Columbia however have three: spring, summer, and winter[268]. The Thompson Indians of the same province group their months into five seasons, winter beginning with the first snow that stays on the ground, and lasting until its disappearance from the valleys, generally the 2d, 3d, and 4th months, spring beginning with the disappearance of the snow, and embracing the period of frequent Chinook winds, 5th and 6th months, summer 7th, 8th, and 9th months, early autumn (Indian summer) 10th and 11th months, and late fall which takes up the rest of the year[269]. The neighbouring tribe of the Shuswap recognise five seasons exactly corresponding to those of the Thompson Indians[270].

The natural phenomena from which the seasons are determined and named vary according to the geographical latitude, the nature of the country, and the mode of life, i. e. according as the tribe lives by hunting or by agriculture. Certain writers state that the Indians of Virginia divided the year into five seasons: the budding of spring, the earing of corn or ‘roasting-ear time’, summer or ‘the highest sun’, corn-gathering or ‘fall of the leaf’, and winter[271]. The Maida of northern California say that the seasons—the rainy season, the leafy season, the dry season, and the season of falling leaves—were instituted by Kodoyampeh, the Creator[272]. The Kiowa distinguished only four seasons: saigya or säta, considered to begin at the first snowfall; asegya, spring (the etymology of the word is unknown, a more recent name is son-pata, ‘grass-springing’), which begins when grass and buds sprout and the mares foal; paigya, summer (pai, ‘sun’), which begins when the grass has ceased to sprout and lasts until fires become necessary in the tipis at night; paongya, autumn (the thickening of the coat or fur, pa, of the buffalo and other animals), sometimes called ‘the time when the leaves are red’, begins when the leaves change colour[273]. It is to be noted that these seasons must be of very different length. In the same way the Dakota reckon five months each for winter and summer and only one month each for spring and autumn, but it is expressly mentioned that this reckoning is not strictly followed[274]. The Pawnee divided the year into a warm and a cold period, and also into the four seasons, each of which however was normalised to three months[275]. The account of the Comanches is somewhat indefinite: they have no computation of time beyond the seasons, which are reckoned by the rising height of the grass, the fall of the leaves, and the cold and the hot season. They very seldom reckon in new moons[276]. They have the four seasons therefore. The Indians of Chile have words for our four seasons[277].

The above-mentioned names of the five seasons are those of the Algonquins of Virginia[278]; the Occaneechi of the same district call them:—the budding or blossoming, the ripening, midsummer, harvest or fall, winter[279]. Certain agricultural tribes of the east divided autumn into early autumn, when the leaves change colour, and late autumn, when they fall, but denoted the two periods by entirely different names[280]. Agriculture is responsible for the adding of a fifth season to the four arising from the warm and the cold periods and the times of transition between these[281]. But other transitional periods between the longer seasons also arise independently[282]. The Lapps have names for the four ordinary seasons, but their language also contains compounds like ‘spring-winter’, i. e. late winter,—a compound also known in Swedish (vår-vinter)—and ‘autumn-summer’, i. e. late summer[283]. The Lapps of Västerbotten divide the year into sjeunjestie, the dark period, and tjuoikestie, the bright period. They also have four seasons:—dalvie, winter, from the freezing of the lakes till the melting of snow; geira, spring, time of snow-melting and spring floods; gese, summer, from the time when the earth becomes visible to the fading of the grass; tjatj, autumn, from this time until the lakes begin to freeze again. The Lapps speak also of talve-qvoutel, mid-winter, kese-qvoutel, midsummer, and tjaktje-kese, late summer[284].

The Yukaghir of N. E. Sibiria use more often the names of periods or the seasons of the year than the names of the months. They have six seasons. The limits of these seasons can hardly have corresponded in former times to fixed dates. Being at present baptized, they reckon the seasons of the year according to the Greek-Orthodox holidays; and thus we have the following seasons:—1, puge, summer, from St. Akulina to Mary’s Day, 13th June to 8th September; 2, nade, autumn, from the 8th of September to St. Michael’s Day, 8th of November; 3, cieje, winter, from the 8th of November to Purification, 2d of February; 4, pore, first spring, from Purification to St. George’s Day, 23d of April; 5, cille, the second spring, from the 23d of April to the beginning of snow-melting, usually to St. Nicholas’ Day, 9th of March; the name denotes the icy surface forming during the night on the snow, after having melted during the day, and is also given to a month; 6, conjile, the third spring, from the snow-melting period to St. Akulina’s Day[285].

Africa offers good examples of the fluctuation and further sub-division of the seasons. The Wagogo of East Africa divide the year into the dry season, about May to October, and the rainy season, November to April. In the latter they further distinguish the little rainy season, songola, November and December, and the greater one, itika, about February and March[286]. In the neighbourhood of Mombasa the great rains begin in April and last approximately for a month, mwaka or masika: mchoo is a week in August, and vuli a fortnight in November, with showers. Beyond the seasons the natives have little idea of the lapse of time[287]. The Wa-Sania of British East Africa have three periods of four months each, gunu, adolaia, and huggaia, but no explanation whatever of these names is given[288]. The Masai divide the rainy season into three periods, and also have four seasons of three months each:—(1) ol dumeril, the time of the lesser rains, preceding that of the great rains. The latter fall in (2) en gokwa, named after the Pleiades, which at that time rise low on the western horizon (sic!). Then follows (3) ol airodjerod, the season of the gentle after-rains, and then (4) ol ameii, the time of hunger and drought[289]. Hollis begins the list with the months of the showers, and calls the season of the great rains l’apaitin le-’l-lengon, ‘the months of plenty’, stating that the latter season, in which the setting of the Pleiades takes place in the evening, is called from these loo-’n-gokwa[290]. Among the Ewe tribes the year has three periods:—adame, March to June; keleme, July to October; pepi, November to February. In the first two much rain falls, so that work in the fields is greatly hindered. Inland the year begins in March with the yam-sowing, and ends in February. The three principal seasons include four months each. Inland keleme also includes another period, masa, September and October, the second maize-sowing. Hence the name ‘masa-corn’. Pepi is the harmattan time, in which fall yam-harvesting, grass-drying, and hunting[291]. The Yoruba divide the year into the dry season, the season of the harmattan wind, and the rainy season, the last-named being further divided into the time of the first rains and that of the last rains or ‘little rainy season’[292]. In Loango a dry and a rainy season of about 6 months each are distinguished. In many districts there is also a third season, tschimuna, the time of the ripening of favourite fruits etc., and the hot seasons are then often simply called bimuna[293].

Where two rainy seasons separated by dry seasons occur, a fuller division of the year presents itself. The Babwende have five seasons:—ntombo, from the first rains at the end of September or beginning of October to the ceasing of the great rains at the end of January; kianza, the lesser dry season, to the beginning of the great rains in February; ndolo, the latter part of the rainy season up to sivu, the dry season, which begins in June; and mbangala, in August and September, when the grass withers and is burnt up[294]. The Wadschagga count:—the great rainy season, 4 months; the time of dew, 2 months; the season of heat, about 2 months; the so-called lesser rainy season, 1–2 months; the great heat, about 3 months[295]. The seasons of the Banyankole are determined by the rains. The longer period is termed kyanda and usually has six months: the lesser, akanda, has four, and there are two months called itumba. During the six months very little rain falls, then come a few days of rain followed by four months of dry weather, and after that two other months of rain[296]. A very striking example of the crossing and overlapping of the seasons is afforded by the Bakongo. They have sivu, the cold season, at the beginning of the dry season which commences about May 15; mbangala, the dry season with little or no dew, July to the middle of October, including also mpiaza, the grass-burning season, second half of July, August, and September; masanza, early light rains, latter part of October, November, and December; nkianza, short dry season, most of January and the early part of February; kundi, nsafu, fruit season, end of February to May, including kintombo, heavy rains, March, April, and nkiela, the time when the rains cease, from the beginning to the middle of May[297].

In the inland districts of Madagascar, in the neighbourhood of Antananarivo, there are properly only two seasons, a hot rainy period from the beginning of November to the end of April, and a cold dry period during the other months. However four seasons are distinguished:—lohataona, ‘head of the year’, September and October, when the rice is planted and a few showers fall; fahavaratra, ‘the thunder-time’, from the early part of November to the end of February or into March; fararano, ‘the last rains’, from the beginning of March to the end of April; and ririnina, ‘time of bareness’, when the grass becomes dry, June to August. Rice is planted twice, first before the end of October and again in November or December; the first crop is ripe in January or early in February, the second about April; the two crops however are not clearly distinguished and together last about four months[298]. One name for winter is maintang, ‘the earth is dry’[299].

The Hottentots seem to keep in view the vegetation rather than the climate. Their seasons are four in number. First, early spring. When with increasing warmth, independently of the rain-fall, trees and bushes break into leaf, and in good years winter or early spring rains have revived the grass, spring or blossoming-time has come; it begins in August and ends in October. The following season, which in the upland Damara dialect is called ‘the sun-time’, embraces the first half of the hot period in which, when the year is good, the so-called lesser rains fall. If these are wanting, or, as is usually the case, are scanty, the land is for the most part desolate, without grass or herbage. This time of drought is described by the same word as the drought itself: it prevails from October to December inclusive. The season upon the productiveness of which the welfare of the Hottentots in the main depends may be called the pasture-season: it includes the period of the greater rains and the time immediately after this, when the fodder has not yet lost its freshness. It fills, loosely speaking, the period January-April, and constitutes summer and early autumn. Winter, or the cold season, May to August, embraces two-thirds of autumn and the first half of winter[300]. The Herero also have four seasons:—spring (from September onwards), summer, autumn or the rainy season, and winter[301].

In Burmah there are three seasons, though certainly they are regulated by the months: the cold season, the hot season, and the rainy season[302]. The Polynesians usually have two long seasons, but three are not unheard of. A native of the island of Molokai, in the Sandwich group, states that there the year was divided into three seasons:—maka-lii, kau, and hoo-ilo. Maka-lii was so called because the sun was then less visible, being obscured by clouds, and the days were shortened. Kau was so termed because tapa could then safely be spread out to dry. Hoo-ilo meant ‘changeable’[303]. The two main seasons are called kau and hoo-ilo. It is to be observed however that in a notice from Hawaii they are called hoo-ilo and maka-lii[304]. This shews that the number is not fixed. On the Society Islands besides the two seasons regulated by the Pleiades there were also three seasons: (1) tetau, autumn or season of plenty, the harvest of bread-fruit, commencing with December and continuing until faahu, which corresponded to January and a part of February, the time of the most frequent rains, comprising three months; (2) te tau miti rahi, the season of high sea, November to January; (3) te tau poai, the longest season, winter, the season of drought and scarcity of food, which usually extended from July to October[305]. It will however be seen that these seasons do not fill up the year, and that the second partly covers the first. Their names are taken from different phenomena of Nature. The New Zealanders distinguish four seasons:—spring, te aro aro, mahaua, te toru, ‘the time of growth’, both toru and aro aro signify ‘the shooting or springing forth of plants’, mahaua is the season of warmth; summer, raumati, waru, rehua,—raumati means ‘dead leaves’, and the summer is so called because all the trees with one exception are evergreen and shed their leaves in summer; autumn, ngahura matiti; winter, hotoke, puanga, the season when the earth is damp and gives forth her worms, which were formerly highly prized as food[306]. The seasons are regulated by the stars, puanga is the great winter star, rehu the great summer star.

The names of the greater seasons are therefore taken for the most part from the varying phases of the climate, but very often refer also to the phenomena of natural life accompanying these. The climatic phases, on account of their fluctuating duration and their limited number, afford no means of distinguishing and naming a greater number of smaller seasons: the phases of plant and animal life may be used as an equivalent and are much better adapted to this purpose, especially when to them are added the regular occupations of agriculture. In the above examples terms referring to natural life have already been found mingled with those borrowed from the climate. Where the seasons are numerous this is always the case: direct references to the climate may even be entirely lacking. These facts shew moreover that between the largest and smallest seasons there exists no difference in the main: they pass into one another without interruption through a series of intermediate stages. Such smaller seasons may be run together into the circle of the year; but this seldom occurs, since the ordinary reckoning according to lunar months has absorbed the smaller seasons, which, on account of their varying and indeterminate length, are inconvenient for reckoning, whereas the regular and definite length of the months makes them easy to reckon. It is however sometimes the case.

The Indians in general have lunar months named from natural occurrences, but not so the Luiseño of Southern California. According to P. S. Sparkman in his unpublished Dictionary of their language the Luiseño year was divided into 8 periods, each of which was again divided into two parts, distinguished as ‘large’ and ‘small’ or ‘lean’. These divisions did not represent periods of time but merely indicated when certain fruits and seeds ripened, grass began to grow, and trees came into leaf in the valley or on the mountain. The native names are given but are unfortunately not translated. Du Bois, to whom we are indebted for this information, names the parts ‘months’ (in inverted commas), and adds that the names are all taken from the physical features of different seasons. Tausunmal, about August, means that everything is brown and sear. Tovukmal refers to the little streams of water washing the fallen leaves. Tasmoimal means that the rain has come and grass is sprouting. In nemoimal the deer grow fat. The ‘months’ are marked by the rising of certain stars. The seasons have here developed into a regular calendrical cycle[307].

In reality this cycle is in no way distinguished from the succession of seasons given above: it has only been improved and regulated. This happens more particularly under the influence of agriculture; one can speak of an agricultural year the seasons of which are determined and named in accordance with agriculture. Of the Fanti of the Gold Coast it is said that they divide the year, according to the changes of the climate, into nine parts with distinct names, beginning with the harmattan wind in January and ending with the small tornadoes in December[308]. The periods however are related to agriculture, as appears from a detailed description for the countries around the Niger. The end of the rainy season and the beginning of the dry (about November) forms a kind of season by itself, and is called odun (year). The farmers go on weeding their farms to give the crops of their second harvest a chance. The dry season is divided into two sections of two months each. During the day it is very hot. The cold wind blowing from the east is called harmattan by Europeans, oye by the natives. The second crops of corn, beans, and guinea-corn are now gathered. The land is cleared for the next season’s crops, and the bush already felled is burnt. This is also the fishing season. The dry season (erun) continues for the next two months, but during the latter part of the second month the rumbling of thunder is heard and small rains fall. The preparation of the ground is continued and yam-planting begins. The rainy season may be divided into two parts separated by a little dry season: the first section consists of five lunar months of rain, the latter of two lunar months, one nearly dry month intervening. The first two months of this section of the rains are called asheroh ojo: it is the tornado season. At the beginning of this season ground-nuts and the first crop of corn are planted. In the next two months the rain-fall reaches its maximum. Towards the end of the second month it becomes possible to eat new corn. The main crop however is left standing in the fields until it becomes quite dry, which happens when the next season, the little dry season, sets in. This sub-division of the rainy season is called ago, probably because the corn has grown tall during the last month. The season called awori consists of one month of rain and the little dry season. The first crop of yams, the corn, the ground-nuts, and the gourds are gathered in. Before long the rains have ceased, the seed for the second crop of corn is sown. The two following months are called the arokuro season, and like the first two months of the rains they are tornado months. Bushes are felled in order to prepare the land for next year’s sowing, and weeding is continued[309]. The months mentioned are lunar months. An interesting feature is that the names of the seasons do not altogether coincide with the natural divisions of the climate, as the following comparison clearly shews:—odun, end of rains, beginning of dry season; erun, dry season I, II, 4 months; asheroh ojo, season of rains (tornadoes), 2 months; ago, rainy season, maximum, 2 months; awori, 1 month rain and little dry season; arokuro, season of rains (tornadoes), 2 months. The deviations are brought about, as the description shews, by the business of agriculture.

The Shilluk know the months but also divide the year into the following nine seasons:—yey jeria, about September, harvest of red dura; anwoch, about October, end of the harvest, people are waiting for white dura to ripen; agwero, about November-December, harvest of white dura begins; wudo, December to January, harvest of white dura continues; leu, January-February, the hot season, dodin, about March, in these two there is no work in the fields; dokot, about April, ‘mouth of rain’, beginning of the rains; shwer, about May-July, time for planting red dura; doria, about July-September, beginning of harvest[310]. A similar but more indefinite mode of reckoning seems to exist among the Bakairi of S. America, of whom it is said that they reckon by dry and rainy seasons, and also distinguish ‘months’ not by the moon but quite vaguely by the rain and the heat and the phases of the maize-culture[311]. Their months are given as follows:—‘hardest rain’, about January; ‘less rain’, February; ‘rain ceases’, March; ‘it (the weather) becomes good’, April; ‘wood-cutting’, May and June; July, nameless; ‘end-of-the-day-time’, August; ‘the rain is coming’, September and October; ‘the maize ripens’, November; December, nameless[312].

The agricultural year is most clearly defined among the rice-cultivating peoples of the Indian Archipelago, by whom the seasons are determined according to the state of the rice. It is said, for example, in speaking of an event, that it happened at the blossoming or harvesting of the rice[313]. Among the Bahau, a Dyak tribe of Borneo, the year is divided into eight periods according to the various kinds of labour carried on in the rice-field:—the clearing of the brushwood (to prepare the fields for cultivation), the felling of the trees, the burning of the wood felled, the sowing or celebration of the seed-time festival, the weeding, the harvest, the conclusion of the harvest, the celebration of the new rice-year[314]. The Bontoc Igorot, as has been mentioned, divide the year into two parts, the period of rice-culture and the other period. There are however other periods which vary in different villages as regards name, number, and duration, but are everywhere called after the characteristic occupations that follow one another in the course of the year. Eight of these together make up the calendar, and seven of them have to do with the rice-cultivation. Each period receives its name from the occupation which characterises its beginning, and keeps this name until the beginning of the next period, even when the occupation that characterised it had ceased some time before. To cha-kon belong:—(1) i-na-na, the first period in the year, the time, as it is said, of no more work in the rice sementeras, when practically all the fields are prepared and transplanted; in 1903 it began on Feb. 11 and it lasts about 3 months, continuing until the time of the first rice-harvest in May, in 1903 till May 2; (2) la-tub, the time of the first harvests, lasts about four weeks and ends about June 1; (3) cho-ok, the time when most of the rice is harvested, fills about 4 weeks, in 1903 till July 2; (4) li-pas, the season of ‘no more palay-harvest’, lasts for about 10 or 15 days. To the half-year ka-sip, belong:—(5) ba-li-ling, which takes its name from the general planting of camotes and is the only one of the calendar periods not named from the rice industry: it lasts about 6 weeks, or nearly to the end of August; (6) sa-gan-ma, the time when the sementeras which are to be used as seed-beds for the rice are put into condition, the earth being turned three several times, lasts about 2 months: on Nov. 15, 1902 the seed was just peeping from the kernels; the seed is sown immediately after the third turning of the earth, which thus ended early in November; (7) pa-chog, the period of seed-sowing, begins about Nov. 10; although the seed-sowing does not last many days, the period continues for 5 or 6 weeks; (8) sa-ma, the last period, in which the sementeras are prepared for receiving the young plants, and in which these seedlings are transplanted from the seed-beds, lasts nearly 7 weeks, from about Dec. 20 to Feb. 10. The Igorot often say e. g. that an event occurred in la-tub or will take place in ba-li-ling; they therefore keep these periods in mind just as a European thinks of some particular month in which an event has happened[315]. The greatly varying length of the periods is once more to be noted, and also the fact that a vacant season is made into a period (see e. g. under (7)), it being necessary to fill in the gaps so that the circle shall be continuous.

How such seasons and the year formed out of them may be developed under the influence of the improved calendar into periods of definite numbers of days is shewn by the Javanese peasant calendar which is still used in Bali and Java. The year is an embolimic year of 360 days and is divided into 12 periods of unequal length. These are:—koso, 41 days; karo, 23; katigo, 24; kapat, 24 (25)[316]; kalimo, 26 (27); kanam, 41 (43); kapitu, 41 (43); kawolu, 26 (in leapyear 27); kasongo, 25; kasapuluh, 25 (24); dasto, 23; sodo, 41. The first ten of these names are the ordinal numerals of the Javanese vernacular, the last two, according to Wilken, are corruptions of Sanskrit words. In Bali the year begins with the eleventh season (April), in Java with the winter solstice. The different divisions correspond to the following occupations and natural events:—1, the falling of the leaves, burning of dry grass, and cutting of trees for the cultivation of mountain rice; 2, beginning of vegetation; 3, blossoming of wild plants, planting of yams and other secondary crops; 4, rutting season, high winds, the rivers swell; 5, preparations for rice-planting; 6, ploughing and rice-sowing; 7, rice is planted, the canals are repaired; 8, rice grows and flowers; 9, the seeds form in the rice-plants; 10, rice turning yellow; 11, the rice-crop is ripe, harvest begins; 12, cold weather begins, the harvest is finished and the rice housed. This is almost literally translated from the language of the natives[317]. Wilken gives to certain periods a different number of days (see [note 1]); according to him the year has 365 days, but every fourth year is a leapyear with 366 days. The calendar was regulated in 1855 by Pakoe Boewånå III, naturally according to the Gregorian calendar: hence the variation from Crawfurd’s statements. This is the only instance of an attempt to bring a natural calendar into agreement with the demands of a modern one; it is however unpractical and inconvenient on account of the varying length of the divisions. It is still used in eastern Java and in the Tengge mountains[318].

In China, besides the lunisolar type of year, there is a division of the year into 24 parts, the names of which correspond to the climatic phenomena but are also borrowed from the phenomena of natural life. They are:—rain-water, 15 days; moving of snakes, 15 days; spring equinox, 15 days; pure brightness, 15 days; sowing-rain and dawn of summer, together 31 days; little fruitfulness (Ginzel) or little rainy season (d’Enjoy), corn in the beard, together 31 days; summer solstice, 16 days; beginning of heat, 16 days; great heat, signs of autumn, together 31 days; end of heat, white dew, together 31 days; cold dew, 15 days; autumn equinox, 15 days; hoar-frost, 15 days; signs of winter, 15 days, beginning of snow, great snows, together 29 days; winter solstice, 15 days; little cold, 15 days; great cold, 15 days; dawn of spring, 15 days[319]. Of this division Ginzel says that among the Chinese the seasons are expressed by a division of the ecliptic: they are therefore astronomical, the Chinese have no special names for the physical seasons. In former times they took the length of the astronomical year to be 365¼ days, and assumed an equal period for the course of the sun in the ecliptic; but they afterwards learnt to calculate the beginning of the divisions directly. It would be surprising however not to find underlying the present divisions old seasons which the astronomical knowledge has drawn within its scope, and which have thus been systematically developed and regulated. To decide the matter would require special knowledge which the present writer does not possess. It is to be noted moreover that the periods are connected in pairs, the odd numbers (according to Ginzel’s scheme) are called tsie, the even k’i, the joint name being tsie-k’i.

As far as the Indo-European period is concerned it seems now to be agreed that there were then three seasons: for only the roots occurring in the words hiems, ver, and summer recur in a greater number of the Indo-European languages. The much criticised statement of Tacitus about the Germans is therefore corroborated: “They know and name winter and spring and summer, but are ignorant of the name and the goods of harvest”[320]. Spring however is not equivalent to the other two seasons, for Indo-European antiquity certainly also divided the year into two parts, the cold and the warm seasons. The question whether the primitive Indo-European tribe had two or three seasons is therefore pointless, and that this is so will be readily understood by anyone who has become familiar with the overlapping and the instability of the seasons of the primitive peoples. The same phenomenon repeats itself in the addition of a fourth season. The Greeks complete the circle of the year with the three seasons winter, spring, and summer (χειμών, ἔαρ, θέρος), but in Homer the fruit-harvest, ὀπώρη, already appears with the pretensions of an independent season. Alkman has these four[321]. The principle of nomenclature is however different: the first three names are derived from climatic phenomena, ὀπώρα from the fruit-harvest. Now since four climatic periods are naturally to be distinguished—cold, warmth, and two transitional periods—the logical consequence is that the fourth season should also be referred to the climate, and indeed to the still unnamed period of transition between summer and winter. This period however does not coincide with ὀπώρα, but follows it. The latter term is therefore corrected to φθιν- or μετόπωρον; the ὀπώρα naturally persists as the fruit harvest, and Theophrastus[322] counts it in addition to the other four and thus gets five seasons. The same thing seems to have happened in the case of the Latin autumnus, although the process cannot be demonstrated. If the small seasons are included the circle may be still further extended. Thus the pseudo-Hippocratean treatise Περὶ ἑβδομάδων[323] gives seven seasons:—1, seed-time, σπορητός, from the early rising of the Pleiades to the winter solstice; 2, winter, until the late rising of Arcturus; 3, tree-planting, φυταλιά, up to the spring equinox; 4, spring; 5, summer, from the early rising of the Pleiades up to that of Sirius; 6, fruit-harvest, ὀπώρα, until the early rising of Arcturus; 7, autumn. This arrangement is certainly affected by the septenary system which pervades the treatise, but is founded on a popular basis: the smaller seasons, which otherwise pass into the greater, are given an independent position by the side of these. The system has not prevailed, it is true, but it affords a typical example of the instability of the seasons.

Exactly the same process recurs in the Indian seasons. The natural division of the North Indian year is into three periods—a warm, a rainy, and a cold season. Three corresponding seasons are the most usual in the Vedic period, and these are still the popular divisions in the Punjab. Later two transitional periods are interpolated, one of an autumnal character between the rainy season and the cold season, and a warm period between the cold season and the hot. These five seasons often occur in the Brahmanas. The well-known six seasons—vasanta, spring; grishma, hot season; varsha, rainy season; śarad, autumn; hemanta, winter; śiśira, cool season: the cold season is divided into two periods—are the result of a systematic comparison with the months, the latter being distributed in pairs among the seasons. By this arrangement the rainy season is the loser, since it embraces at least three months. There is also a second sexpartite division of the year, not indeed mentioned in the Vedic literature but better corresponding to the course of the seasons, in which the rainy season is divided into two periods[324].

The splitting up of the seasons persists to this day among the Germanic peoples; but a systematising of these small seasons is only found when they are referred to the Julian months. This point will be dealt with below, in chapter XI. The phenomenon is known to me from my own native district. The word höst, ‘autumn’, still persists there in the old literal sense of harvest, mowing, and indeed höhösten is particularly the hay-harvest. Hence the designation of the autumn season as höst is felt to be insufficiently accurate and the term is replaced by efterhöst, literally ‘after-harvest’, late autumn. Between summer and efterhöst appears the skyr (dialect for skörd), the harvest, as a fifth season; sometimes there is added a sixth season, sivinter, late winter. Little attention has been paid to this phenomenon, though it is common enough. The periods of the rural occupations in particular give rise to such terms. Any period of this nature is described by the old Swedish word and (ann), now obsolete except in dialects. For the other districts I add from the Dialect Dictionary of Rietz:—hobal, the period on the one hand between the tillage in spring and the hay-harvest, and on the other between the hay- and the corn-harvest, the former period being the greater, the latter the small hobal. Elsewhere the word has the form hovel, summer being divided into hoveln, mellan-anna and ann (which is here used pregnantly to mean harvest). Compounds with and are vår-, säs-, gödsel-, hö-, slått-, skår-, skyr- and sädes-and (periods of spring, sowing, manuring, hay, hay-harvest, harvest, corn). The North Frisians of Amrum and Föhr for instance mark events by the periods um julham (‘at Christmas’), um wosham (‘in early spring’), pluchleth (ploughing-time), meedarleth (hay-harvest), kaarskörd (corn-reaping). In Norway there are current as general time-indications:—fishing-time (fiskja), springtime (voarvinna or voaronn), ploughing-time (plogen or plogvinna), midsummer (haavoll or haaball), ‘between time’, i. e. between ploughing and hay-making, (mellonn), early summer (leggsumar), haymaking-time (høyvinna, høyonn, or slaatt), harvest-time (haustvinna or skurd), ‘shortest-days-time’ (skamtid)[325]. In Iceland, where the sheep-farming is the principal industry, we find:—Lamb-weaning time or Pen-tide, stekk-tid, in May; Parting-tide, fra-faerar, when the sheep are driven to the hills; Market-tide, kaup-tid, when all purchases for the year are made; Home-field hay-time and Out-field hay-time (July and August); Folding-tide, rettir (September), when the sheep are driven off the hill pastures into folds to be separated into flocks and marked. Again from wild birds and eider-ducks one calls the spring Egg-tide. The fisherman uses such seasons as ver-tid, Fishing-tide; of these there is a spring, an autumn, and a winter fishing-month. Flitting-days, fardagar, come in the spring, and skil-dagar in summer, when servants leave.[326] In the old German laws and elsewhere similar time-indications are common, e. g. at plough-time, at the second plough-time, at autumn-sowing, at harvest, at hay-making time, at hemp-gathering, after harvest and hay-making, at the bean-harvest, at plough-time, at the grape-harvest, at sowing-time, at harvest-time, fall of the leaves, sprouting of the leaves, oat-cutting or harvest[327]. In Anglo-Saxon a similar expression occurs in a law of King Vihtraed in the year 696, sexton dæge rugernes (rye-harvest). These periods are in themselves indefinite, they fail to achieve a definite length or quite fixed position in the year. Where they do so, this is due to the comparison with the Julian months, of which more later.

However over the number of the seasons among the Germans or, what has often been regarded as the same thing,—and this is an evidence of the false methods by which the problem has been attacked—over the German division of the year, a long and vigorous dispute has been carried on. That the year was divided into two parts, summer and winter, is well known. I refer to the Scandinavian half-years[328], to the testimony of Bede[329] that the Anglo-Saxons reckoned six months for winter and six for summer, and to the German expressions for a year: ‘in bareness and in leaf’, ‘bare and leaf-clad’, ‘in straw and in grass’[330]. No less a scholar than J. Grimm has cast doubt on the statement of Tacitus that the Germans had only three seasons, but later he withdrew his doubts in view of the consideration that the Germans at the time of Tacitus were acquainted with grain-culture but not with fruit-culture, and that the word autumn, harvest, referred to the fruit and vine-harvests and therefore naturally did not appear among the Germans of that time[331]. In view of the linguistic phenomenon mentioned [above, p. 71], it seems now to be agreed that the account of Tacitus is in the main correct. Weinhold has given the treatment of the question its direction. According to him the tripartite division to which reference has been made crowded out the older division into two parts, the points of division, he maintains, doubtless coinciding in the first instance with the three Lauddinge or ungebotene Gerichte (regular courts), which are found as early as the time of Charlemagne. The beginnings of the four seasons—determined from saints’ days—in February, May, August, and November are of foreign origin: on the other hand the quadripartite division of the year, arising from the fact that mid-winter and midsummer were added to the beginning of winter and summer as interpolations in the time-reckoning, is German. This Weinhold tries to prove from the popular festivals associated with these dates. The attempt however is a complete failure. No season begins with any of the solstices, on the contrary these fall right in the middle of a season. His thesis rests on an erroneous conception of the festivals, viz. that they are in general calendar-festivals. Under primitive conditions a festival (the harvest-home in particular) may certainly conclude a division of time and may thus also indicate the beginning of a new season, but as a rule the festivals, though regulated by the calendar, are not so ordered that they coincide with the beginning of a season. We are therefore not authorised in drawing conclusions as to the beginning of a division of the year from the existence of an old festival. Support has been lent to the idea of Weinhold by the fact that in later times the beginnings of the seasons were indicated by festivals and saints’ days. The fact of the matter is that the common medieval calendar was composed of a series of festivals and saints’ days from among which suitable and well-known days were chosen in the dating of the beginnings of the seasons also. For the general understanding it was necessary throughout to bring in popular saints’ days[332]. Tille attacks Weinhold very sharply but remains throughout under the influence of the method indicated by the latter: his work, however, has its good points, inasmuch as it refers to economic conditions, agriculture, the payments of rent, etc. The bipartite division, he asserts, is primitive Indo-European, the tripartite is of foreign (Egyptian) origin: both existed for a long time side by side. This fact is explained by an old sexpartite division of the year, since the six seasons could be run together either in twos or in threes. The beginnings of the half-years are given by natural phenomena, those of the three annual divisions are placed by Tille at March 13, July 10, and Nov. 11, old style: in the north on account of the climatic conditions they are pushed back a month. Hammarstedt[333] remarks very pertinently that the beginning of winter in November, in the north in October, belongs to the reckoning in half-years, and that hence arises the absurdity that Tille has to give Feb. 10 as the date for the beginning of spring in the north. But to assign Dec. 13 with Hammarstedt as the beginning of one of the three seasons agrees just as little with the natural seasons of the year.

The principal error lies in the systematising, the seasons being regarded as periods of a definite number of days. This is not the case even to-day, and still less was it so, as we have seen, among primitive peoples. Still more clearly does the same error of method appear in Tille’s assumption of a sexpartite division of the year, or of sixty-day periods, as they are expressly termed. He refers to the six old Indian seasons, which are a comparatively late and artificial product called forth by the adoption of the names of the seasons in the reckoning by months[334], and to the pairs of months of the Syrian and Arabian calendar. He regards as 60-day divisions not only the smaller seasons mentioned [above, p. 75], the duration of which was originally no less indefinite than it is to-day, but also the Germanic pairs of months, which owe their origin to an adaptation of the Roman months (for this see [below, ch. XI]). The 60-day periods are so far from being primitive that they first took their origin under the influence of the reckoning in months.

In Iceland there still exists a curious calendar, the ‘week-year’. The year is divided into two halves, misseri; the people reckon in so many misseri, not years; it consists of whole weeks, in the ordinary year 52 (= 364 days), in leapyear 53 (= 371 days). Until midsummer (or mid-winter) they reckon forwards, so many weeks of summer or winter have elapsed, after that backwards, so many weeks of summer (winter) remain[335]. Bilfinger in a penetrating study has tried to shew that this curious calendar is an outcome of the ecclesiastical calendarial science of the Middle Ages. He does not however prove his case: rather, the calendar, as tradition shews, reaches far back into heathen times[336].

The reckoning in weeks was once common to all Scandinavia. The Lapps have special names for every week of the year, borrowed from festivals and saints’ days falling within the weeks; they have therefore taken from the Scandinavians the reckoning in weeks and adapted it to the uses of a primitive time-reckoning. From the same source they have also derived the special significance of the summer night (April 14, Tiburtius) and of the winter night (Oct. 14, Calixtus), from which also two weeks are named. The system is better preserved in certain parts of South Sweden[337]. The people count in räppar, quarter-years—in Öland they are called trettingar, thirteenths, i. e. 13 weeks—beginning with the räppadagar: these are Lady Day, Midsummer Day, Michaelmas Day, and Christmas Day, old style. Just as in Iceland, they reckon backwards, not however in the same quarters as there, but in the quarters before Midsummer and Christmas: in the other two quarters they count forwards. In northern Scania I have met with a relic of the same type of reckoning, the ‘number of weeks’ (ugetalet), which begins on April 6 (Lady Day, old style), and is reckoned backwards as far as the thirteenth week. The duration of both rural occupations and natural phenomena is determined in so many weeks. As the starting-point of this reckoning in weeks the four great festivals which come nearest to the four points of the solstices and equinoxes are chosen. There can be no doubt that these have made their appearance under the influence of the Christian calendar instead of the four Old Scandinavian points of division of the year. The people call Calixtus’ day (Oct. 14) the first day of winter, and Tiburtius’ day (April 14) the first day of summer; many rune-staves have this division of the year, and almost all describe the former by a tree without leaves, the latter by a tree in leaf. They fall in the same weeks as the initial days of winter and summer in Iceland, which vary there on account of the peculiar arrangement of the calendar. In Scandinavia, however, they have been transformed into fixed days under the influence of the Julian calendar.

It is a natural conclusion that the reckoning in weeks had its origin in the use of the rune-staff. Since the week-day letters on these are repeated the whole year through, the weeks offered an easy means of reckoning. This conclusion is certainly correct, but still we may venture to ask why the week-day letters were admitted into the national calendar by the North especially, and why the reckoning in weeks should be adopted in popular use only there. The reason can only be that the counting in weeks was already in use before the rune-staff was introduced. This mode of counting, which in Iceland had been developed into a curious form of year, was in Scandinavia adapted to the Julian calendar and remained bound up with this. The leap-week was therefore unnecessary. The old basis is however still preserved in the points of departure, the summer and winter nights. It is the same system as the Icelandic, built up on the week and the year, but differently modified: the idea of any borrowing cannot be entertained. The basis of this calendar, therefore, was once common to all Scandinavia, and the calendar must go back to heathen times.

Under the influence of the popular lay astrology the week was early spread among the Germanic peoples: on it and on an approximate knowledge of the length of the year, such as could easily be acquired in the lively intercourse with Christian lands during the Viking period, the system of the Icelandic calendar is built up. An indigenous element however appears, the half-year reckoning, and indeed the great probability is that the limitation of the half-year to a fixed number of days was first achieved as a result of this systematising of the calendar. Winter and summer, like all natural seasons, had at first no fixed limits. The quarters arose in the course of the reckoning, the people counting forwards in the first half of the half-year and backwards in the other half. The middle points of the half-year, mid-winter and midsummer, fell where both reckonings met. This agrees with the popular objection to high numbers. The Germanic tribes of the south, in accordance with their milder climate, commonly reckoned five months for winter. In the north the dead season is longer, about six months, and this fact has contributed to the half-year reckoning which, as has already been remarked, is widely characteristic of northern peoples. That the limits between both seasons were unstable and could be moved forward according to circumstances is in my opinion shewn by the names of the initial days of the half-year—sumarmál (plural) and vetrnaetr, ‘the winter nights’. Where a definitely determined day is in question the plural is out of place: it is used to describe a period, for instance jol (plur.) denotes Christmas-time[338].

With the two opening days of the calendar and the one division in the middle are often combined the three great sacrificial feasts, the autumn festival at the winter nights, the Yule festival at mid-winter, and the spring festival at the summer nights. It is true that the first of these festivals, which was celebrated at the beginning of a period of rest after the completion of the harvest and agricultural labour, denoted, as such festivals often do, the conclusion of the old year and the beginning of the new. That it was fixed for a definite day cannot be demonstrated any more than that the festival of victory in spring, celebrated before the Vikings went forth on their voyages, fell exactly on the summer night. On the contrary the time probably varied according to circumstances: the expression of Snorre lacks calendarial accuracy and remains indefinite:—“They should sacrifice against the winter to get a good year, and at mid-winter sacrifice for germination; the third sacrifice in summer, and this was a sacrifice of victory”[339]. In historical times the Yule festival is regulated by the Christian calendar; Snorre says that in heathen times it was celebrated at the hökku night, but of this we have no certain knowledge. Things happened as in the Middle Ages and later: after a calendar has arisen the festivals are regulated by this, but they are not calendar-festivals, and in reconstructing the scheme of the calendar from the festivals very great caution must be exercised.

Our conclusion is that the Germanic seasons, like the seasons in general, were not in themselves definitely limited divisions of time, and that alongside of the greater seasons smaller ones arose without there being any numerical determination of the relationship between the two. Seasons only become divisions consisting of a definite number of days when in the regulation of the calendar they are taken over as calendar divisions, as winter and summer were in Scandinavia. Where a calendar has arisen directly out of the seasons, the divisions, like the seasons, are of varying length[340]. This also shews that the Germanic seasons first attained a definite number of days through the calendar-regulation introduced from abroad. Further, when a calendar existed, the beginning of the seasons could be given with reference to this: the day varied according to circumstances, but the choice was limited in this manner, viz. that only a popular festival or saint’s day was appropriate as a distinguishing day. Here also, therefore, the calendar was the starting-point for the regulation of the seasons. A division of the year in the more accurate sense also first arose through the regulation of the calendar, since, owing to the method of calculation, the middle days of the half-year divisions became distinguishing days in the calendar. When the calendar came, the old festivals were also regulated by it.

By way of supplement two or three curious exceptional cases may be noted. A completely isolated instance is offered by the Bangala of the Upper Congo, who count in lunar months, and, since there is no dry season, reckon for longer periods by the rise of the rivers[341]. In the monsoon districts however it is frequently a peculiarity to distinguish the seasons by the winds. Of Sumatra it is reported:—The principal seasons are named after the quarters of the heavens from which the wind blows. At the time when we were in Taluk, April to mid-June, the south monsoon was blowing; the east, the west, and the north monsoons also come under consideration for the seasons. Moreover the people also distinguish a dry and a rainy period. The seasons 4. tahun djin, 5. tahun wou, 6. tahun sai were regarded as falling within the rainy period, while the dry season set in with 1. t. ali, and continued with 2. t. dal awal, and 3. t. dal akhir. In the two seasons 7. t. ha and 8. t. ‘am dry and wet weather alternate[342]. In New Britain (Bismarck Archipelago), between the two greater seasons of the south-east and the north-west monsoons, each consisting of 5 months, there were two smaller intermediate seasons of one month each, the period of variable winds and the period of calm[343]. In Songa (Vellalavella), one of the Solomon Islands, various seasons are distinguished according to the direction of the wind:—the time of the west wind, nanano; the time of the almond-ripening, tovarauru (the time of the north wind); rari, the time of the south wind—during this period calm prevails at night but there is wind in the day-time; sassa nanamo, time of the east wind; mbule, time of calm, lasting about a month. After mbule follow tovaruru, lasting about 2 months, and sassa nanamo, one month. In Lambutjo the matter is still further complicated. The following winds are distinguished:—south wind, west wind, good wind at the time of almond-ripening, lasting about one month. Further the east wind, strong or quite weak with squalls, not good. Three months afterwards comes the west wind, lasting about 2–3 months. After the east wind a south-west wind, very strong, at that time one cannot sail on the sea: it often comes 5 months after the east wind. After the south-west wind a SE wind, lasting only 1–2 weeks. Then strong E wind, lasting 1–2 months, during which time navigation in canoes is impossible. Then again a time of ‘clear water’, i. e. calm, lasting two months. After this, S wind, NW wind, and NE wind. Each of these lasts only a short time, altogether they occupy 3–4 months. Then begins a lighter E wind, lasting 3–4 weeks. Then about one month of light W wind, then again stronger E wind for 1–2 months. Afterwards S wind for 1½-2 months, lighter SE wind for 1–2 weeks, and then again stronger E wind for 2–3 months. At the time of the west wind there is much rain, at the time of the east wind much sunshine[344]. It is very interesting to see how accurately primitive peoples observe Nature, but these are not indications of time. On the Gazelle Peninsula it has been observed that when the SE monsoon blows the sun comes up in the east, and when the NW monsoon blows it rises in the south: the wind comes from the opposite direction to that in which the sun rises[345].

CHAPTER III.
THE YEAR.

Following the practice of my authorities I have often in the foregoing pages made use of the expression that the year is ‘divided’ into so many parts. From a genetic stand-point this expression is incorrect, because the time-indications, which relate to a concrete phenomenon of Nature, are older than the year, and, since they are connected only with the single phenomenon, are discontinuous or even indefinite. Only through their union does the complete year arise. Every natural year however offers on the whole the same phenomena following one another in definite succession, and thus the circle of the year has its prototype in Nature herself. Nevertheless the uniting of the different seasons into a complete year only takes place gradually by means of a selection, systematising, and regulation of the seasons. It must be carried out according to a principle—we shall see that this is as a rule the lunar reckoning—but the occupations of agriculture also serve as a handle. The present chapter will shew how the uniting of the seasons into the year is only a late and incomplete development, how originally the year does not exist as a numerical quantity, the pars pro toto counting being resorted to, and finally how the years are not reckoned as members of an era but are distinguished and fixed by concrete events.

The difficulty of struggling through to the conception of the year is exemplified by certain peoples who know two seasons but reckon in half-years without joining them together. Naturally this happens in the rare case in which there is very little difference—or none at all—between the two halves of the year. Thus of the Akikuyu of British East Africa it is reported:—The equatorial year has no winter or summer. Its passage is marked by two wet seasons, which occur in what are our spring and autumn. The planting is done in all cases at the first commencement of the rains, and harvesting as soon as the crop has ripened after the cessation of the rain. There are therefore two seed-times and two harvests in twelve months, and when the native speaks of a year he means six months[346]. This is very natural, since by ‘year’ a vegetation-period is often to be understood: the half-year reckoning however also appears where a difference between the two seasons does exist. In Rotuma or Granville Island the inhabitants reckon in periods of six months or moons. The west monsoon, which blows from October to April, doubtless serves to distinguish these seasons: otherwise the difference between the seasons is hardly perceptible, the island lying near the equator. The half-years each contain six months, to which the same names are given in both halves[347]. The people of the Nicobars reckon in monsoon half-years, shom-en-yuh, the SW monsoon, sho-hong, blowing from May to October, and the NE monsoon, ful, from November to April, so that two of these form one of our years[348]. The half-years are also said to contain seven months each[349]: in reality they must vary between 6 and 7 months, as the year varies between 12 and 13. In New Britain (Bismarck Archipelago) there are monsoon years of five months: the two intervening periods of the variable winds and of the calms, each lasting one month, are not counted[350]. It is said that the Benua-Jahun of the Malay Peninsula have no other division of the year than the natural one of the north and south monsoons, each of which they call a ‘wind-year’, satahun angni; however a word for year, sa taun, is also ascribed to them[351]. In Bali the year is divided into two seasons or monsoons, each of which includes six months; since the months of both halves have the same names it is evident that originally only half-years existed[352]. The greatest unit of time among the Orang Kubu of Sumatra is the six-month mussim (season), which is of Malay origin[353]. The Samoans have a name for a period of twelve months, but they formerly reckoned years of six months (tau-sanga); each of these corresponded to one of the two six-month periods, the palolo or rainy season and the monsoon season[354]. The Moanu of the Admiralty Island name the division of the year according to the position of the sun. When it stands north of the equator, the season in question is named morai in paiin (sun of war), since wars are chiefly fought in this season. When it stands over the equator, the season is called morai in houas (sun of friendship), the season of friendship and mutual visits. When the sun turns towards the south, the cooler season begins[355]. Of the Kiwai Papuans of the islands in the delta of the Fly River in New Guinea, Torres Straits, Landtman writes to me that he cannot say if the people are clear whether they reckon in years or in half-years[356]. The former supposition is really only supported by the fact that they are aware that the same natural conditions recur after the lapse of the two half-years. There is no word for year. On the whole it may be said that they count only the months, and hardly conceive of so great a unit as the year, nor even (at least not everywhere) of the half-year, although there may be a hint of this in special cases.

Not seldom the dry and the rainy seasons are counted without being combined into a year. This is expressly stated of the Tupi of Brazil and certainly applies also to the Bakairi[357]. In Loango there are dry and rainy seasons, and in many districts a third season also, the fruit-ripening. Commonly the people reckon by the two main seasons. A centenarian is therefore fifty years old[358]. In Uganda there are in the course of twelve months two rainy and two dry seasons, although there is hardly a month in which no rain falls at all. The rainy season from February to June is called togo mukazi, since the rain then falls without much thunder: the second, from August to November, is called dumbi musaja, because of the thunder and the frequent deaths from lightning. The dry season about December is more intense than that about June. However the year, mwaka, is composed of one rainy season together with the following dry season, and consists of six moons or months[359]. Their year, corresponding to a half-year, consists of five moons, and a sixth in which it rains[360]. In north Asia the common mode of reckoning is in half-years, which are not to be regarded as such but form each one separately the highest unit of time: our informants term them ‘winter year’ and ‘summer year’. Among the Tunguses the former comprises 6½ months, the latter 5, but the year is said to have 13 months; in Kamchatka each contains six months, the winter year beginning in November, the summer year in May; the Gilyaks on the other hand give five months to summer and seven to winter. The Yeneseisk Ostiaks reckon and name only the seven winter months, and not the summer months[361]. This mode of reckoning seems to be a peculiarity of the far north: the Icelanders reckoned in misseri, half-years, not in whole years, and the rune-staves divide the year into a summer and a winter half, beginning on April 14 and October 14 respectively. But in Germany too, when it was desired to denote the whole year, the combined phrase ‘winter and summer’ was employed, or else equivalent concrete expressions such as ‘in bareness and in leaf’, ‘in straw and in grass’[362].

‘Years’ with less than twelve months are to us the strangest of phenomena. The Yurak Samoyedes and probably the Tunguses of the Amur reckon eleven months to the year, the Kamchadales only ten, of which one is said to be as long as three[363]. The natives of southern Formosa reckon about eleven months to the year[364]. The inhabitants of Kingsmill Island, which lies under the equator, reckon periods of ten months, which are numbered but, in contradistinction to the other examples, are reckoned in cycles[365]. In the Marquesas 10 months formed a year, tau or puni, but the actual year, i. e. the Pleiades year, was also known[366].

The Yoruba reckon in 16-day divisions. Fourteen of these form their old year, of 224 days, i. e. in former times attention was paid to the rainy season only. The first thunder was the signal for the fishers and hunters to come back to their huts and begin farming again.[367] The Toradja of the Dutch East Indies reckon in moon-months: two to three months however compose a vacant period in which they do not trouble about time-reckoning[368]. The Islamite Malays of Sumatra distinguish tahun basar, the great year, or tahun musin, the year of the seasons, both reckoned as 12 months, from tahun padi, the rice-year, which among them counts only eleven months[369]. The Dusun of British North Borneo have two methods of reckoning their longest divisions of time. If the native be a hill-man he will reckon by the taun kendinga or the hill-padi season, six months from planting to harvest, if a plain-dweller by the taun tanau or wet padi season, 8 to 9 months[370]. This incomplete year is therefore a vegetation year in which the vacant period of no work is simply passed over. In this manner may be explained the much discussed ten-month year of the Romans[371], if it really depends upon old tradition and is not a mere creation of spurious learning. It is not a cyclical year like ours: a complete explanation will be given below in the investigation of the manner in which the years were counted.

It is true indeed of most primitive peoples, as is said of the Hottentots, that they are well acquainted with the conception (sic! I should have said rather: the concrete phenomenon) of the year, guri-b, as a single period of the seasonal variation, but do not reckon in years in this sense[372]. That is to say the year is by them empirically given but not limited in the abstract: above all it is not a calendarial and numerical quantity. Of the Waporogo it is said:—Somewhat more difficult (than the times of day) is the conception of the year. Only older, more intelligent people have a clear idea of it, the sowing-time and the rainy seasons constituting their points of reference. But they too can only reckon up a few years (though they certainly do this by counting the seasons, cp. [below, p. 92]), and for the great mass of the people the conception of the year does not exist[373]. The Bontoc Igorot has no idea of a cycle of time greater than a year, and in fact it is the rare individual who thinks in terms of a year[374]. The length of the year consequently varies. Among the Banyankole it begins with the first heavy rains and lasts until the next heavy rains, so that a year may be longer or shorter by a few days: it is a matter of no consequence whether it is a week or even three weeks that are taken off or added to the length[375].

With the agricultural year it is just the same. For the Dyaks of Borneo the rice-harvest is a main division of the year (njelo); in September after the conclusion of the harvest the year is at an end; a definite beginning, a New Year’s Day, is unknown[376]. The translation of a Ho text runs:—“When the inhabitants of the interior begin to cultivate the yam-fields they begin a new year: when the yams are dug up and the dry grass is burnt away, a year has passed”[377]. Among the Thonga the notion of the year (lembe, dji-ma) is extremely vague: the year begins at two different periods, that of tilling and that of harvesting the first-fruits. They do not make any difference between a lunar and a solar year[378]. A very significant account comes from Dahomey. The word for year does not denote any definite number of months: the sense is rather ‘to plant maize and eat, to plant it again and harvest it’. At the end of the harvest the year also is at an end[379].

Here therefore we have a natural year quite concretely and empirically given. Chronologically it is of no use nor indeed is it used: what method is resorted to will be shewn below. Attention must first be called, however, to an important point. The purely natural year is a circle which has no natural division, i. e. no beginning or end, the seasons following upon each other immediately; not so the agricultural year, which has both beginning and end. Here therefore there is a natural point of division, a new year, which appeared in some of the examples just given, and this is an extremely important point for time-reckoning. The vacant period between harvest and sowing presents some difficulty, and so both of these periods can be used as the beginning, as is done among the Thonga: otherwise the beginning of the year varies considerably, just because it can be arbitrarily determined[380].

The contradiction between length or duration of time and time-reckoning evidently here becomes apparent. The counting is not performed by means of these fluctuating empirical years, but the pars pro toto method is employed, the years are counted by a season. As soon as it is said that some event took place at a definite time of the previous year, or will take place at some point in the following year, a counting of the years is thereby implied, although for an enumeration of this kind the conception of the year is not necessary. When it is said that something happened at the previous harvest, or will happen at the next dry season a counting of the years is no less implied, although seasons are reckoned instead of years, i. e. the pars pro toto method is used. Thus it is, in fact, with all primitive and many highly developed peoples, and that not only when an event that took place at a definite time is spoken of, but also where the number of years alone is in question: in the latter case the reckoning is only performed from a favourite, conventionally selected season. The statement made for the Hottentots is significant for the kind of reckoning just mentioned. They keep in mind the age of their cattle from the calving and lambing periods[381]. Similarly we are told of the modern Arabians that the female camel is covered for the first time when she is four rabi old (rabi = the pasture-season in spring, when the camel foals), so that she foals in the fifth rabi[382].

As a basis for the counting either a longer or a shorter season may serve, or indeed any popular natural phenomenon of regular annual occurrence. Thus of the Chinhwan of Formosa it is stated that they have no calendar: they only know that a new year has come when a certain flower blooms again[383]. The Paez of Columbia have a word enzte, ‘fishing, summer, year’, since a great fishing is only engaged in once a year, in January or February[384]. In the language of the Tupi of S. Brazil the year is always called akayú, cashew-tree, which blossoms once a year, and produces a much-prized reniform stone-fruit which is also often used in the preparation of wine: the word also means ‘season’. This tree bears fruit only once a year, whence it comes that the Brazilians reckon their age by the stones, laying aside one for each year, and keeping them in a small basket reserved for this purpose[385]. The Algonquin of Virginia reckoned in cohonks, winters; the name refers to the wild geese, and shews that these have come back to them so many times[386]. In medieval Swiss charters time is often reckoned in louprisi, ‘leaf-fall’; dri, nün louprisi = when the leaves have fallen three, nine times, etc.[387].

In a later section on the beginning of the year we shall find that the appearance of a certain constellation, in particular the Pleiades, gives the signal for the beginning of the agricultural labour, whence is developed the importance of this date as the opening of the year. The time between two like appearances of the same constellation, e. g. between two heliacal risings, is a year. In this manner the name of the constellation itself can come to denote ‘year’. In many parts of S. America the same word means both ‘Pleiades’ and ‘year’[388]. The inhabitants of the Marquesas call the year of 12 months, as distinguished from the 10-month fruit-year, by the name of the Pleiades, mata-iti[389]. How easily this comes to pass is shewn by a statement made for the Bangala of the Upper Congo. The culmination of the constellation kole gave the principal planting-season. This was so familiar to the natives that the informant used the word kole as equivalent to the word ‘year’[390]. This is in its very nature a pars pro toto designation, since it refers to an annually recurring phase of the stars.

More often the years are reckoned by one of the greater seasons. It is a well-known fact that in Old Norse generally, in Gothic, and often in Old German and Anglo-Saxon time was reckoned in winters. We find traces of the same practice in Greek (χίμαρος, ‘a one-year-old goat’, from the same root as χειμών, winter) and in Latin (bimus, trimus = ‘of two, three years’, from hiems): poets often reckon in hiemes[391]. It is almost the rule among all peoples who live under a climate that has a winter with snow and ice. The Ostiaks reckon in winters, and so do the Eskimos of Greenland[392] and of the Behring Straits[393], and the N. American Indians in general, for instance the Kiowa[394], the Pawnee[395], and the Omaha[396]. The common method of reckoning is not by the season, ‘the cold time’, but by the concrete phenomenon that distinguishes it, viz. the snow. So with the tribes of the N. W. interior[397], the Hupa[398], and the Dakota, who say that a man is so many ‘snows’ old, or that so many ‘snow-seasons’ have passed since an occurrence[399]. The Siciatl of British Columbia reckon either by summers, ‘fine seasons’, or by winters, ‘snows’[400]. For the Algonquin [see p. 93]. In the tropics to reckon by the cold season is rare: the Guarini of Paraguay however reckon in roi, i. e. ‘seasons of coolness’, ‘winters’[401], and the Bakongo occasionally by sivu, the cold season, though more often by mou, ‘season’[402]. The reason for the reckoning of the years in winters is the same as that for the counting of the days in nights. Winter is a time of rest, an undivided whole, which practically becomes equivalent to a single point: it is therefore more convenient for reckoning than summer, which is filled up with many different occupations. In the south of N. America, in the states on the Gulf of Mexico, where the snow is rare and the heat of summer is the dominant feature, the term for year had some reference to this season or to the heat of the sun[403], e. g. among the Seminole of Florida the name for the year was the same as that used for summer[404]. Here the summer is the time of rest, but in Slavonic also time is reckoned in summers (leto = ‘summer’, plural = ‘years’). We may compare here the English expressions ‘a maiden of 18 summers’, etc. The reckoning in springs is only exceptional. The Basuto word selemo means ‘spring, ploughing-time, year’[405]. At the southern end of Lake Nyassa time is reckoned by ‘rains’, i. e. rainy seasons[406].

Ever since the principal food of man has been the produce of fruit-trees or the corn, the fruit- and corn-harvests and the whole period of vegetation in general have been of decisive importance for his well-being. We have already seen how this circumstance has left its mark upon the indications of the seasons, and in the same way the second most important method of counting years is to reckon by harvests or vegetation-periods. The fellahs of Palestine still do this. Their usual method is to reckon from one harvest to another, or, as they put it, ‘from threshing-floor to threshing-floor’[407]. In modern Arabia rents are hardly ever reckoned for a whole year, but only until the next spring, rabi, when the young animals are sold, or, as by the fellahs, until the next threshing-time, bedar, when the farmer can realise upon his corn[408]. The Negrito of Zambales determine the year by the planting or harvesting season, but their minds rarely go back farther than the last season[409]. In Bavaria in the Middle Ages the years used to be reckoned in autumns. The ceremonial reckoning in the Sanskrit ritual texts is in autumns, Sanskrit çarad, ‘autumn’[410]. The subjects of the Incas had a word huata, ‘year’, which as a verb meant ‘attacher’: but the lower classes reckoned in harvests[411]. This is also done in the district around Mombasa[412]. The Arabs sometimes reckon the years as e. g. 40 charif, charif being the time of the date-harvest[413].

We have already spoken of the rice-year in the East Indian Archipelago as a combination of the agricultural seasons; the period of vegetation of the rice also serves, although seldom, for the counting of the year. Among the Toradja the time needed for a plant to come to its full development up to maturity is called ta’oe, and santa’oe accordingly means ‘a year ago’. Sampae is the rice-year of six months, but santa’oe has practically the same meaning, since the rice is the most important cultivated plant. In general, however, the word is seldom used as a time-indication, but the years are reckoned by well-known events (on this see [below, pp. 99 ff.]); nevertheless expressions like the following are heard:—santa’oe owi, ‘when last year’s rice-crops still stood on the field’, roeanta’oe owe, ‘two harvests ago’[414]. In the South Sea Islands the bread-fruit is the most important article of food: the people, as we have seen, know a time of abundance of food and a time of scarcity. We are told:—The Malay word for ‘year’ is taun or tahun. In all Polynesian dialects the primary sense of tau is ‘a season’, ‘a period of time’. In the Samoan group tau or tausanga, besides the primary sense of season, has the definite meaning of ‘a period of six months’, and conventionally that of ‘a year’, as on the island of Tonga. Here the word has the further sense of ‘the produce of a year’, and derivatively ‘a year’. In the Society group it simply means ‘season’. In the Hawaiian group, when not applied to the summer season, the word keeps its original sense of ‘an indefinite period of time’, ‘a life-time, an age’, and is never applied to the year: its duration may be more or less than a year, according to circumstances[415]. So far our authority. It seems however to be questionable whether the original sense is not the concrete ‘produce of the seasons’, rather than the abstract ‘period of time’. It is significant that on the Society Islands the bread-fruit season is called te tau, and the names of the other two seasons, te tau miti rahi and te tau poai, are formed by adding to this name[416].

Of great significance are the accurate reports for the Melanesians. They have no conception of the year as a definite period of time. The word tau (a Polynesian loan-word), or niulu, which corresponds most nearly to ‘year’, signifies a season, and so (now) the space of time between recurring seasons. Thus the yam has its tau of five moons, from the planting—when the erythrina is in flower—until the harvest, after the palolo has come and gone. The bread-fruit has its tau during the winter months: bananas and cocoa-nuts have no tau, since they always bear fruit. The notion of the year as the time from yam to yam, from palolo to palolo, has been readily received, but it is very doubtful if such a conception is anywhere purely native[417]. The Melanesians are only interested in the concrete phenomena of the year, and not in time-reckoning as such, and therefore do not in practice combine the period from yam-planting to harvest with that from harvest to planting to form a year. When it is pointed out, however, it is quite clear to them that this is a single period of the variation of the seasons. The Polynesians have themselves noted this fact, and accordingly the sense of the word tau has been extended from ‘season’ to ‘year’.

Whether the conception of the year was known in the Indo-European period is not certain: it is however significant that all the words for ‘year’ of which the etymology is fairly certain either refer to the produce of the year—as ὥρα and its cognates, and also the word ‘year’ itself, Old Scand. ár—or else come from the pars pro toto counting of the year. Thus the Slavonic leto means ‘summer’ and ‘year’. Sanskrit çarad means ‘autumn’: that the corresponding Avestic sared means ‘year’ is explained by the fact that the years were reckoned in autumns. The Greek ἐνιαυτός is unexplained, but in Homer, in the law of Gortyn, and in the inscription of the Labyades it has also the little observed sense of ‘anniversary’[418], which may be the original sense. Further evidence of the lack of an acquaintance with the conception of the year is afforded by the fact that the Germanic peoples render it by periphrases like ‘winter and summer’, etc.[419].

The pars pro toto counting of the year from shorter or longer seasons does not however extend beyond the years immediately following or preceding. It is stated of the tribes living at the southern end of Lake Nyassa that the years are reckoned in ‘rains’ up to three or four years: everything beyond that is kale, ‘some time ago’[420]. In the district around Mombasa, in periods not exceeding five years, the date is usually fixed by the number of harvests which have been gathered[421]. In general the primitive peoples reckon only where an immediate practical interest requires them to do so. The Kiwai Papuans have no word for year, but only for the monsoon periods: they cannot as a rule state how many years have elapsed since a certain event, but only whether it took place recently or long ago[422]. The inhabitants of the islands of the Torres Straits never count years[423]. Individuals belonging to tribes at a low stage of civilisation keep no account of their own age. Among the Waporogo no one can say how old he is[424]. The Edo-speaking tribes have a calendar, but an enquiry as to the age of a man or the number of years since a given event will meet with no answer, or a random one[425]. In Dahomey no negro has the slightest idea of his age[426]. The Hottentots have no interest in their own age, but are interested in that of their cattle, which they reckon by the calving and lambing periods[427]. Few of the Chinhwan of Formosa know their age[428]. The Negritos of Zambales have no idea of their age[429]. No Marquesas Islander, no Oceanian in general, can give either his own age or the time of any event[430]; even the Maoris do not know their age, although they know that the man of forty years is older than the man of thirty[431]. The statements here made obviously refer to the absolute age of a man, not to the relative age; for either it is immediately seen or else easily remembered from childhood who is older and who younger. The Babwende, for instance, never know how old they are, but do know quite well who is the oldest[432]. Since the relative age is thus known, the age of the people and the time of events can be determined by reference to the speaker’s own relative age or to that of someone else. On the same page as that from which the above quotation for the Marquesas Islands is taken, it is stated that in order to determine the time of any event the people indicate how tall a person was, or how long his beard was, at the time when the event took place. The Indians of Pennsylvania temporarily determined an event by referring to their own age at the time of its occurrence[433].

From these indications of relative ages there arises of itself a familiar chronological expedient usually found at the point where history begins, viz. the reckoning by generations, which is common e. g. among the Polynesians[434] and in the older Greek historians. Among the Masai an elaborate system for classifying ages has exceptionally developed. The circumcision takes place in four-year periods with intervals of three and a half years. The circumcisions are known alternately as ‘right-hand’ and ‘left-hand’. Those who have been circumcised at the same time have a special name, such as ‘those who fight openly or by day’, ‘those who are not driven away’, etc.; one ‘right-hand’ and one ‘left-hand’ period combine to form a generation. The ‘those-who-fight-openly’ period is a ‘right-hand’ period, and those who belong to it were circumcised in 1851–5; the ‘those-who-are-not-driven-away’ period is a ‘left-hand’, and its members were circumcised in 1859–63. The two periods or ages together form a generation composed of persons born from 1834–1850. Each age has three divisions, first those known as ‘the big ostrich feathers’, secondly those called ‘the helpers’, and thirdly those known as ‘our fleet runners’[435]. It is evident that an excellent basis for the determination of relative time is hereby given. With time-reckoning per se the system is not concerned.

Common bases for reckoning are afforded by important and striking events which have been impressed upon everyone and are present to all men’s minds: through their relation to the age of some person they serve as a guide to the chronology. The Aino, for example, do not count the days, but always refer to events; if it is asked how old anyone is, the answer will be that he was born after the catching of the very big fish, or perhaps in the year when there was so much snow[436]. Here once more we see how concrete time-indications always precede the abstract numerical counting of time. And where numbers are known they are not willingly used, but the year is referred to as one distinguished by a certain noteworthy event, instead of being regarded as a member of a series. From a year of this kind the natives can only reckon for a few years at most in either direction. Where there are many such noteworthy years the time-relationship is so far recognised that the succession of the events is known, and perhaps in certain cases also forms the basis of calculation.

In the neighbourhood of Mombasa wars, famines, the arrival of white men form epochs of this kind: it is impossible to detect the age of any adult[437]. It is mentioned that the Toradja of the Dutch East Indies sometimes reckon nearly approaching events or events of recent occurrence by the rice-sowing: dates at a more distant past are indicated by mentioning events of most note, such as the death of a great man, an epidemic of small-pox, an important military expedition, a conclusion of peace, the payment of a tax, etc. The people do not reckon their own age, but count that of their children, saying: “When he was born I had my rice-field there, the next year there”, and so on[438]. It is amusing and at the same time instructive to note that precisely the same mode of reckoning was found in Scania at the beginning of the last century. It was a very common thing, says a well-known authority on the folk-lore of this district, for a peasant, when asked how old e. g. his little girl was, to give some such answer as: “She must be four years old, for she is the same age as my brown mare, and she was born when our southern field was a grazing meadow”[439].

The Batak of Sumatra think that a small-pox epidemic returns at intervals of from nine to twelve years, and make use of this belief in reckoning time. On questioning a chief, says a traveller, how old his house was, I was told: “It has existed only for two small-pox epidemics”, by which he meant that it was somewhat more than 24 years old[440]. In Borneo there have occurred two eclipses of the sun during the last half-century. The first of these served as a fixed date in relation to which other events were dated[441].

The Eskimos of Greenland knew up to about the twentieth year how many winters a person had lived, but beyond that they could not go. Sometimes however they used as epochs from which to calculate pellesingvoak, ‘the little priest’, i. e. the arrival of Egede in the country, or the arrival or departure of other well-known Europeans, or the founding of Godthaab and other colonies; they would say that this or that person was born at the coming or departure of such and such a person, or when eggs were collected, seals caught, etc.[442].

The Caffres rarely give the proper length of past or future periods of time, and when they do so the period is never of more than a few months’ duration. Otherwise it is their custom to determine the date at which this or that event took place by reference to a contemporaneous event of greater importance[443].

The Lapps of Västerbotten reckon their age by the reindeer, e. g. when this or that aldo (= female with calf) was born. Formerly they never went farther back in counting than the previous year. When they had to give the date of an important event they referred to the time at which some specially fine female reindeer was born[444].

The Hottentots, as has been said, have no interest in their own age, but keep in mind that of their cattle from the calving and lambing periods. When they wish to date back somewhat farther, well-known events such as the outbreak of cattle-plague, hostilities with neighbouring tribes or with the whites, immigrations, etc. furnish them with satisfactory general indications from which, coupling them in particular cases with the birth of their children or the stature of these at the time, they can arrive at a date[445].

Where the political development has advanced so far that a stable monarchy exists, the succession of rulers offers an excellent means of chronological orientation, and within every reign certain years can be distinguished by special events. But this brings us to the beginning of history, and I desist from following the subject further. One example only:—The Baganda reckon by the reigns of the kings and by certain wars in one particular reign. They say ‘It was in the reign of such a king’, or ‘I was still in arms when such and such a war was fought in so and so’s reign’[446].

Where no reigns furnish a system of chronological reckoning, the concrete references may be systematised until each year is named and distinguished by a definite event. This was the practice of the Arabians before Mohammed. Mohammed is said to have been born in the year of the elephant, or, according to other sources, some years after the year in which the viceroy of Yemen marched against Mecca with an army in which there were elephants[447]. Another year is called the year of treason or outrage, because certain garments which a Himjarite king had sent that year to Mecca were stolen, whence arose a conflict at the feast of pilgrims, in which the young Mohammed is said to have taken part[448].

The Wagogo count the years by important events, e. g. ‘the year when the cattle died’, or ‘two years after the building of Boma (Kilimatinde Station)’[449]. The Masai do not count the years, but rather denote them by referring to the most important events that took place in them, e. g. a murrain, a drought, the death of the chief, an expedition particularly rich in booty, etc.[450]. A fully developed calendar of this nature is possessed by the Herero, and has been published from the year 1820[451]. I give a few years as examples:—1820, ojo (= year of the) tjekeue: from the name of the Matabele chief who in 1820 came to Okahandja with a white peace-ox and made peace with Katjamuaha. 1842, ojohange, ‘year of peace’, the Nama and Herero made peace. 1843, ojomaue, ‘year of the stones’: the Herero as the slaves of Jonker Africander had to build for him a stone wall; or ojovihende, ‘year of the stakes’: the Herero had to build a palisade around Jonker’s dockyard. 1844, 1845, ojomukugu or ojombondi, ‘year of vomiting, of nausea’: the Nama had poisoned Katjamuaha, and the latter vomited and purged. And so on up to 1902 inclusive. There are lacking only the years 1854, 1855, and, significantly, 1891, 1895, 1899, and 1900, towards the end: the reckoning fails under growing European influence. Several years have two descriptions, e. g. 1844 and 1845 (see above); these and 1887–8 are run together, the latter as the ‘year of the red murrain among the cows’.

The same mode of reckoning appears, strongly developed and fixed by the aid of picture-writing, among the Indians of N. America. Heckewelder says of the Indians of Pennsylvania:—“They reckon larger intervals of time by some noteworthy event, e. g. a very severe winter, a very deep snow-fall, an unusual inundation, a general war, the building of a new town by the whites, etc. Thus I have heard more than fifty years ago:—‘When their brother Miqaon talked to their fathers they were so old or so tall, they could catch butterflies or hit a bird with an arrow’. Of others I have heard that they were born in the hard winter (1739–40), or could then do this or that, or already had grey hair. When they could not refer directly to any such distinguishing epochs they would say: ‘So many winters after that’”[452]. This method of reckoning seems to have existed among the Pawnee at an initial stage. Sometimes they referred to a year that had been marked by some important event, e. g. a failure of crops, unusual sickness, a disastrous hunt: this was referred to as a year by itself, but after only a few years’ remove this mark became indistinct and faded away[453]. Among the Dakota and the Kiowa detailed descriptions were given in picture-writings, which are well-known and have been published, for the Dakota by Mallery and for the Kiowa by Mooney. They are painted on buffalo hide, later also on paper, and represent in painting the history of the tribe. They were executed by a specially gifted Indian and were handed down from father to son. When worn out and obliterated by use they were renewed. In winter they were often produced before the fire, and the events recounted. Everyone knew them, however, so that anybody could shew when he was born or when his father died, and some also knew the meaning of the pictures. Four copies belonging to the Dakota are known, which go back to 1800, 1786, 1775, and the mythical period, respectively. Every year is denoted by a picture, without distinction between winter and summer. Some of the terms used are:—1794–5, the ‘Long-Hair-killed’ winter; 1817–8, the ‘Chozé-built-a-house-of-dead-logs’ winter; 1818–9, the ‘small-pox-used-them-up-again’ winter; 1821–2, ‘the star (meteor)-passed-by-with-a-loud-noise’ winter; 1825–6, the ‘many-Yanktonais-drowned’ winter (through an inundation); 1833–4, the ‘storm-of-stars’ winter (so called from the abundance of shooting-stars), etc. Four Kiowa calendars are known, one of which is arranged in months, of which it gives 37; two of the others refer to the years 1833–93, one to the years 1864–93. In the first each month is indicated by the crescent of the moon, and above is the picture characteristic of the month. The Kiowa annual calendars are clearer than the Dakota in that they indicate winter by a thick black stroke signifying that the vegetation has died, and summer by the medicine lodge with its figures, which form the central feature of the religious ceremonies of the summer. Above and by the side of these signs are the pictures, giving the principal events of the seasons, so that the reckoning of the year becomes the history of the tribe. The Indians however were also acquainted with simpler modes of reckoning. Among the Nahyssan of S. Carolina time was measured and a rude chronology arranged by means of strings of leather with knots of various colours, like the Peruvian quipos[454]. The Dakota use a circle as the symbol of time, a smaller one for a year and a larger one for a longer period: the circles are arranged in rows, thus: ȱȱȱ or o-o-o[455]. The Pima of Arizona make use of a tally. The year-mark is a deep notch across the stick. The records of early years are memorised, and there are a few minor notches to aid in recalling them. The year-notches are alike, yet when a narrator was asked to go back and repeat the story for a certain year he never made a mistake. Taking the stick in his hand, he would rake his thumb-nail across the year-notch and begin:—‘This notch means etc.’[456].

The development is clear. Often an important event has been impressed upon the memory and now serves as a landmark from which the few years that it is possible to count are reckoned. Such events multiply, and when their succession is known, a longer period can be mastered. Finally the process is systematised, so that every year has its event (necessarily even if it be an unimportant one), and is named from that: hence the reckoning of the years becomes also the history of the people. This kind of time-reckoning is really used by every one of us. Whoever looks back over his past life sees chiefly the more important events, not the dates of the years, and to these he joins the more peripheral events and so finds his way in the labyrinth of memory. But we mark the events by the dates, and thereby obtain an estimation of the course of time, which is the last acquisition of the human mind in this domain. The mode of reckoning in question penetrates deeply among the culture peoples.

The same method of distinguishing the years from one another was employed in ancient Babylonia, in the days of the Sumerian kingdom of Ur in the second half of the third millenium B. C., and also later under the first dynasty in Babylon, and was only replaced by the reckoning according to the years of the king’s reign under the dominion of the Kassites[457]. For our historical knowledge of the events these so-called ‘year-formulae’ are of extreme importance. They vary in each case according to the towns, and shew that these in some respects maintained an independent position. The adoption of the year-formulae of the main locality implies the complete subjugation of the town[458]. No trace of an era or any reckoning by the years of the reign is to be found. Only the king’s accession to the throne is utilised for distinguishing the years, the first complete year of his reign (not the year of accession, therefore,) being described as the year of King X. As marks of the other years the most important national events in the domain of the religious cult and of politics are almost universally employed. Only exceptionally is the year named after some violent natural catastrophe. Rather, it is a striking fact that in none of the 66 year-formulae hitherto discovered is there any mention of an eclipse of the sun, or a comet or meteor. If no important event has occurred, the year is described as the one following such and such a year, e. g. the year 49 of king Dungi is called ‘the year in which the temple of X. was built’; year 50 = ‘the year following that in which the temple of X. was built’; year 51 = ‘the year following that in which the temple of X. was built, the year after this’. We see the clumsy method used in order to avoid counting, instead of simply saying ‘the second year after etc.’: so firmly is the concrete description adhered to. These year-formulae were however used for the dating of documents, and not simply, as among the primitive peoples with whom we have hitherto been concerned, for the retaining of past events in the memory. Hence arises the difficulty that often an event of such importance that the year can be named after it does not occur until well on into the year, that is, the event from which the year is named does not take place until a greater or smaller part of the year has already passed by. Until the event takes place indications of the kind already mentioned, having reference to the preceding year, are employed, e. g. the year 17 of Dungi = ‘the year after that in which the ship of Belit (was launched)’; when a noteworthy event happens it gives its name to the year: thus the same year is ‘the year in which the god Nannar was brought from Kar-zi-da into his temple’. Hence arise twofold descriptions, and they are indeed necessary in this kind of designation when events of the current year are to be dated by the year. An example containing a political event is the year 36 of Dungi = ‘the year after that in which Simuru was destroyed’, or ‘the year in which Simuru was destroyed for the second time’. It is characteristic to count the destructions of a town but not the years[459]. During the reign of Rimsin of Larsa, a contemporary of Hammurabi, the years begin to be run together into an era: there are many datings from the capture of Isin, up to thirty years after that event,[460] and so under the second king of the first Babylonian dynasty five years were reckoned after the taking of Kazallu[461]. So also under the first dynasty of Babylon the years were described by occurrences, by events in the religious and political life, especially religious acts and buildings of the kings, by wars, and lastly by natural catastrophes, especially inundations of the country[462]. Dates given by events of a previous year are also found. At that period however the year-formula seems to have been given at the New Year’s Day and therefore to have been determined beforehand: when important historical events occurred, the year was given a new name from these[463].

In the older period of Egyptian history each year of the king’s reign is described by an official name borrowed from the festivals—e. g. those of the king’s accession, of the worship of Horus, of the sowing, of the birth of Anubis—from buildings, wars, and the censuses for purposes of taxation. Gradually the simple counting of the years of the reign appears alongside of these names, and from the end of the old empire completely supplants the former method even in official dates. The years however are not calendar years, but begin with the day of the king’s accession: they therefore offer the disadvantage of running from different dates according to this. At certain periods however the reigns, as in Babylon, were counted only from the first New Year’s Day. Of an era there is only a single example[464]. The Egyptians also began with the concrete descriptions, but passed over, at least within the separate reigns, to the counting of years which is so much more suitable for a survey of the course of time. The Assyrian designation of the year after eponyms, limmu, the Greek after archons, ephors, and other eponymous officials, the Roman after consuls etc. are no different. For a people with a fully developed political life and annually changing supreme officials the latter naturally offer a means of distinguishing the years; the life was too regular and too well-established for events of such a decisive nature that they could impress themselves upon the memory of everyone and become available for time-reckoning to be able to happen to the whole people in smaller intervals of time. Here however the system shews a weak point. It is very difficult to keep an arbitrary series of many names in its right order without confusing the names, and only very few persons can do it. The system therefore did not provide that survey over the whole course of time which the awakening historical sense rendered more and more necessary. So men were led to the only practical method, that of simply counting the years and marking them by figures, by which means everyone without more ado became quite clear as to the dates of earlier or later events, whether these were expressed in olympiads, in ab urbe condita etc., or in the countless local eras of antiquity. It was long before it was seen that the starting-point is a matter of indifference, and that the only essential is that all should use the same starting-point. In this respect the old reckoning in epochs long continued to influence the minds of men.

CHAPTER IV.
THE STARS.

The time-indications from the phases of the climate and of Nature are only approximate: they themselves, like the concrete phenomena to which they refer, are subject to fluctuation. Even in the tropics, where the regularity of the climatic changes is greater than in our latitudes, the beginning of the rains, the dry season, or monsoons may be to some extent advanced or retarded. In the temperate zones the fluctuations are very perceptible. In the year in which I write this (1916) the corn harvest has been delayed by nearly a month, not only on account of bad weather in harvest-time but also owing to the unusually low temperature of the past summer. Even the townsfolk notice that the days are shorter and the weather is colder than is usual at the time of harvest. Further, incidents of plant and animal life—e. g. the blossoming of certain trees and plants, the arrival of the migratory birds—vary somewhat in different years. In general primitive man takes no notice of these variations: the Banyankole, for instance, are indifferent as to whether the year is one or even three weeks longer or shorter, i. e. whether the rainy season opens so much earlier or later[465]. The days are not counted exactly, but the people are content with the concrete phenomenon. More accurate points of reference are however especially desirable for an agricultural people, since, although the right time for sowing can be discerned from the phenomena and general conditions of the climate, yet a more exact determination of time may be extremely useful. The possibility of such a determination exists—and that at a far more primitive stage than that of the agricultural peoples—in the observation of the stars, and especially in the observation of the so-called ‘apparent’ or, more properly, visible risings and settings of the fixed stars, the importance of which has already been explained (pp. [5 ff].) The observation of the morning rising and the evening setting is extraordinarily wide-spread, but other positions of the stars, e. g. at a certain distance from the horizon, are also sometimes observed[466]. The Kiwai Papuans also compute the time of invisibility of a star. When a certain star has sunk below the western horizon they wait for some nights during which the star is ‘inside’; then it has ‘made a leap’, and shews itself in the east in the morning before sunrise[467].

Any reader of the classics will be familiar with the risings and settings of the stars: Virgil, for example, mentions them often. With him however they are pre-eminently a traditional ornament of poetic style: the richest sources are the peasants’ rules of Hesiod, in which the stars are mentioned as time-indications along with phenomena of plant and animal life, and appear just as frequently as the latter, often in combination with them. But Homer not only knows several stars but is also acquainted with the rising and setting. A much quoted passage in the Iliad runs:—

“Him first king Priam saw with his old eyes,

As o’er the plain he lightened, dazzling bright,

Like to the star that doth in autumn rise,

Whose radiant beams, pre-eminent to sight,

Shine with their fellow stars at noon of night:

Orion’s Dog we mortals call its name:

Sign is it of much ill, thought clear its light,

And mighty fever brings to man’s poor frame:

So, as he ran, the brass upon his breast did flame”[468].

The lines refer to the morning rising of Sirius at the beginning of the fruit-harvest, which about 800 B. C. took place on the 28th of July (Julian). A modern reader, thinking only of the splendour of the star as it shines in the sky at night, entirely fails to understand the darker and more fateful side of the simile. Only when it is realised that the time of the morning rising of Sirius is the time of the greatest heat and sickness, a period believed to be induced by the rising of this star at the beginning of the fruit-harvest, is the right idea obtained. Like Sirius appearing in the sky in the morning twilight of later summer, Achilles stands out upon the battle-field, eclipsing all others and bringing destruction to the Trojans[469]. A difficulty has been found in the passage in that Sirius at his rising is only just visible and therefore does not shine in his brightest splendour. But Sirius is for the poet the typical brightest fixed star, just as he speaks of the heavens as ‘starry’ even when the sun is ascending in them[470]. On every day of the opōre Sirius rises higher and shines more brightly—one must not think only of the actual first rising, the first day of the star’s appearance. Hence the star becomes the symbol of the opōre, ὀπωρινὸς ἀστήρ[471]. Since it is a star of evil omen it is also called ‘the disastrous-shining star’[472]. A star-setting is implied in the words ‘the late-setting Arcturus’[473]. The ‘late’ refers to the fact that the circle which Arcturus describes in the heavens is great, since he stands so far north. Here belongs also the observation that the Great Bear alone of the (greater) stars does not dip down into the ocean[474]. The stars further serve as a guide to navigation[475]:—

“And treacherous sleep ne’er fell on the eyes that were watchful still,

For he kept the Pleiads in front, and the Herdman, who slowly doth gain

His rest, and the Bear,—they are wont to call it moreover the Wain:

Ever turning at bay, doth it glare on Orion’s falchion-gleam,

And alone it hath no share in the baths of the Ocean-stream:—

For Calypso, the Goddess divine, had bidden him still to keep

Over his left that sign as he fared on the face of the deep”.

The Pleiades, the Hyades, and Orion are also mentioned, but not in any special connexion with the indication of time[476]. The morning-star helps to determine time on a night journey[477].

Hesiod says that at the time when the thistle blooms and the cricket chirps Sirius burns heads and knees[478], and that when the late autumn rains come men feel relieved, since the star Sirius is not passing over their heads for so long a time but uses the night more[479]. Commentators of classical times have indeed here taken Sirius to mean the sun. But wrongly; for Sirius, whose rising introduces the time of greatest heat, is for the Greeks the cause of the heat, just as the Pleiades are for the Australians, and as all stars are held to be the causes of those climatic changes which are connected with any of their risings or settings[480]; when Sirius rises earlier, i. e. remains in the heavens for some hours during the night-time, the heat declines. The other passages are:—vv. 564 ff., evening rising of Arcturus (60 days after the winter solstice, Feb. 24, Julian), followed by the coming of the swallow, messenger of spring, before this time the vines should be pruned; vv. 597 ff., the winnowing of the harvested corn at the morning rising of Orion (July 9); vv. 609 ff., when Orion and Sirius are in the middle of the heavens and the dawn sees Arcturus (morning rising Sept. 18), it is the time of the vine-harvest; vv. 615 ff., at the (morning) setting of the Pleiades (Nov. 3), of the Hyades, and of Orion (Nov. 15) it is time to think about sowing; vv. 619 ff., when the Pleiades, fleeing from Orion, fall into the sea, storms rage, and the ship should be drawn up on land. Alcaeus says:—“Drink wine, for the star (viz. Sirius) revolves”[481].

The time-indications from the stars are therefore much older in Greece than the lunisolar calendar, and always existed alongside of the latter—which was of a religious and civil character—as the calendar of peasants and seamen, who must hold to the natural year and its seasons. The watchman who speaks the prologue of the Agamemnon of Aeschylus says:—

“ ... On elbow bent, watching, as ’twere a dog,

I mark the stars in nightly conclave meet.

And those bright constellations, without peer,

Lords paramount in heaven, that winter bring

And summer in their train for mortal men,

Right well I know them as they come and go”[482].

The discovery of star-observation and of its use in time-reckoning and navigation is ascribed to the heroes Prometheus and Palamedes. The latter is regarded by the tragic poets as the founder of all the elements of intellectual culture, and so also of the science of the stars[483]. And Prometheus, who glories in having brought to men every advance in civilisation, includes therein the knowledge of the risings and settings of the stars:—

“Of winter’s coming no sure sign had they,

Nor of the advent of the flowery spring,

Of fruitful summer none: so fared through each,

And took no thought, till that the hidden lore

Of rising stars and setting I unveiled”[484].

Later, the phases of the stars have become so familiar to everyone that Sophocles can say, ‘a time of six months from spring to Arcturus’, i. e. the morning rising of Arcturus on Sept. 18[485].

Whether the Romans made use of time-indications from the stars before they borrowed them from the Greeks is uncertain; in any case they had their own names for some constellations:—vesperugo, iubar = lucifer, the evening star, septentriones or iugulae, the Great Bear, vergiliae, the Pleiades. Suculae, the Hyades, and canicula, the Dog-star, are translations of the corresponding Greek names[486].

At a later period the risings and settings of the stars, together with the climatic phenomena accompanying them or believed to accompany them, were brought into a calendar, which was then arranged according to the signs of the zodiac, or later according to the months of the Julian or Egyptian solar year. The Greek lunisolar year was unsuitable for the purpose, since it varied in reference to the sun and the stars. How both were adjusted to practical needs is shewn by the remains of two stone calendars found at Milet. On the stone are inscribed the risings and settings of the stars, arranged according to the signs of the zodiac: by the side of these are holes into which little tablets containing the days of the lunisolar calendar could be fitted, these tablets being arranged according to the relation of every lunisolar year to the solar one[487].

The Arabians also carefully observed the stars, and many of their proverbs couple the risings of the stars with natural events[488]. Since these constellations are the so-called lunar stations their use here is not primitive, but must have been added on to a primitive usage. The Pleiades were observed throughout their course, and about most of the positions which they take up mnemonic verses were made. Mohammed swears by the setting Pleiades in the 53rd chapter of the Koran.

We return once more to the primitive peoples. It may be well first to show by a few examples how far they were acquainted with the stars and saw in them images of terrestrial things. The Chukchee give names to the most important constellations. Among divinities are reckoned ‘the Motionless Star’ or ‘the Nail-star’ or ‘the Pole-stuck Star’, the Pole-star, ‘the Front Head and the Rear Head’, Arcturus and Vega, and pchittin, a part of Aquilo. Orion is an archer with a crooked back, who has shot a copper arrow, Aldebaran, against a ‘group of women’, the Pleiades. His wife is Leo, ‘the Standing Woman’. Capella is a reindeer-buck which is tied behind the sledge of a man driving with two reindeer; a fox approaches from the side. Six of the stars of the Great Bear are men throwing with slings, the seventh is a fox gnawing at a pair of antlers. The Twins are two elks running from two hunters who are driving two reindeer-teams. Corona is the paw of the Polar Bear. Delphinus is a seal, Cassiopeia represents five reindeer-bucks standing in the middle of a river[489].

The Eskimos of Greenland have a good knowledge of the stars. The Great Bear is a reindeer, or the little stool on which they fasten their ropes and harpoons, Aldebaran is the eye of the bull, the twins are the breast-bone of the heavens, the belt of Orion is composed of three ‘scattered ones’—Greenlanders who were taken up into the sky and could not find their way back—Sirius has a man’s name, the Pleiades are to be regarded as baying hounds with a bear among them, Cygnus as three kayaks which have been out seal-hunting. Venus is the follower or man-at-arms of the sun. When one planet crosses the path of another it is a wife and a concubine who have one another by the hair, or else it is a visit of two stars[490]. By the Ammasalik names are given to Vega (‘the Foot of the Lamp’), which, like the moon, is the brother of the sun, to the Great Bear, the Pleiades (‘the Barkers’), the belt of Orion, and Aldebaran; Jupiter is the mother of the sun[491]. Among the Konyag of the island of Kodiak, off the south coast of Alaska, two months are named after the risings of the Pleiades and Orion respectively[492]. Of the Thlinkit it is said that few constellations or stars appear to have been named by them: those to which names are given are ‘the Great Dipper’, which by night used to serve as a guide, the Pleiades (sculpin), ‘Three-men-in-a-line’ (probably the belt of Orion), Venus as the morning star (‘Morning-round-thing’), and Jupiter (?) as the evening star (‘Marten-month’ or ‘Marten-moon’). If the morning star comes up above a mountain south-east of Sitka, it means bad weather, if well over in the east, good weather[493]. Otherwise the North American Indians have paid less attention to the stars: but it is exaggerated to say[494] that the sum-total of their astronomical knowledge was the ability to point to the Pole-star from which they took their way when they travelled at night, which however they did unwillingly. The tribes of Pennsylvania had names for a few stars, and observed their motions: the Pole-star shewed them by night the direction they must take in the morning[495]. The Omaha called the Pole-star ‘the Not-moving-star’, the Pleiades were called by an old name, ‘the Deer’s Head’; this name, which had a religious significance, was not commonly used, the popular name being ‘Little-duck’s-foot’. The Great Bear was ‘the Litter’, Venus ‘Big-Star’[496]. For the Klamath are mentioned only the three stars in the belt of Orion[497], for the Biloxi and Ofo ‘Stars-all-heads’ (?) (three large stars near the Pleiades), ‘Stars-in-circle’ (the Pleiades), and ‘Big Star’, the morning star[498]. The Luiseño of southern California name the most important stars. The associated stars form much larger groups than those common among us. The stars were chiefs among the first people. Those most frequently mentioned are Antares and Altair. Arcturus is the right hand of Antares, it rises before the latter and announces his coming, the other stars around Antares are his suite. Other chiefs are Spica, Fomalhaut, and the Pole-star. Orion and the Pleiades are always mentioned together; the latter were seven sisters, pursued by Aldebaran. The Diegueño constellations are altogether different from the Luiseño, and are based upon totally different ideas: it has not been possible however to obtain an accurate account of them[499]. Of the natives of Guadeloupe it was reported at their discovery:—In other places they merely reckon the day by the sun and the night by the moon; these women however reckoned by other stars, and said that when the Great Bear rose or a certain star stood in the north it was time to do this or that[500].

The Indians of South America have observed the stars in much greater detail. The descriptions of von den Steinen are well known, in particular for the Bakairi of Central Brazil. Orion is a large frame on which manioc is dried, the larger stars are the tops of posts, Sirius is the end of a great cross-beam supporting the frame from the side. The Pleiades are a heap of grains of meal that have fallen out at the side: a larger mass, ‘the father of the heap’, is Aldebaran. Capella is a little capsule such as the Bakairi wear in their ears, two other stars of Auriga are the ear-rings of the Kayabi, the feathers of which are stuck backwards. One star, probably Procyon, is an ear-piercer, or more properly the hole bored in the ear. Castor and Pollux are the holes of a great flute. Canopus has no name. The Southern Cross is a bird-snare on a twig, and the two large stars of the Centaur represent two canes belonging to it. In the snare a mutum cavallo (crax) was taken, and this could be seen in a dark patch of the Milky Way close beside. A Sokko heron with a little basket full of fish corresponds approximately to the stars of Pisces and Argo. The Scorpion is a drag-net for children, the Milky Way is a huge drum-stick, and the holes in it (the dark spots) are observed and explained by stories. The Paressi have a name for the Southern Cross, above which they see an ostrich whose figure is to be recognised in a dark spot of the Milky Way: other animals are also found in the sky. To the Bororo the Southern Cross represents the toes of a great ostrich, the Centaur a leg belonging to them, Orion is a Jabuti turtle and in the parts verging on to Sirius a cayman, the Pleiades are the bunches of blossom on the angico tree. The name of Venus was not translatable[501]. The Karaya of Central Brazil knew many constellations, and drew some of them in our informant’s sketch-book. The Southern Cross, for example, is a ray (the fish), the two stars of the Centaur above it represent an ostrich, upon which a jaguar, Scorpio, is leaping[502]. Of the natives of Brazil in general it is stated that there is hardly a single important constellation which does not explain to them some event, or represent some idea in connexion with things that happen upon the earth, though they certainly have no heroes to set in them. Myths of Orion, of the Pleiades, and of Canopus were related[503]. E. Nordenskiöld has repeatedly visited the border districts between the Argentine, Bolivia, and Brazil. Of the Chané and Chiriguano Indians he says that they do not give names to many constellations, but they know them very well. The part of the Milky Way lying nearest to the Southern Cross is called the Ostrich Way, the Southern Cross together with a few neighbouring stars is the head of the ostrich, and the two largest stars of the Centaur are its collar. Orion with his sword is called ‘Birds-meet-each-other’, another constellation is ‘the Roe-buck’s Horn’, still another ‘the Tapir’; the Pleiades are the most important constellation, they are called yehu, but the natives do not know the meaning of the name. Venus is called coemilla, ‘morning’. The Guarayu call Orion ‘the Black Vulture’; at his side lies a heap of snake’s bones (the sword). The Southern Cross with the stars around it is an ostrich, the two large stars of the Centaur are a roe-buck, the Great Bear is a road, a cluster of stars in the south is ‘the Eel’s Nest’. The Pleiades are called piangi, a word of unknown meaning; when, on their return after their period of invisibility, they are surrounded by a circle, it is a good omen: if the circle is missing, all men will die. Venus is called ‘the Big Star’[504]. The Karai tribes called α, β Centauri the ostrich’s feet, the body is the neighbouring ‘coal-pit’ (the dark spot of the Milky Way), the Southern Cross is a fresh-water ray, the Pleiades are a flock of parakeets, Orion is the burning roça, the tail of the Scorpion is called unze. The Ipurina of Rio Purus call Orion a beetle, the Pleiades a serpent, the Hyades a turtle, the Cross forest-folk[505]. In a Chilean word-list there are words for star, constellation, the Pleiades, Orion, planet, Venus[506].

In Africa the comparatively more civilised negro Tribes seem to have paid less attention to the stars than the more primitive tribes of the south. The Ho tribe considers the stars to be the children of the moon: it recognises and names the most important constellations, the morning star (‘the Clucking Hen’), and the stool-bearer of the moon, a star always situated in the vicinity of that planet. The Milky Way is composed of stars forming a cord[507]. Of the Ibo-speaking tribes we are told that they seem to be singularly incurious about heavenly bodies and occurrences; however names were got for the following constellations:—The Pleiades (‘Hen and Chickens’), the belt of Orion (‘Three and Three’), for the Great Bear two names not translated were given, Venus (‘the Wise-Man-who-can-talk’)[508]. In French Guinea η ursae is an ass, and the little star above it is a thief pursued by the six other stars, members of the tribe to which the stolen animal belongs. For other peoples the Great Bear is the star of the camel, Cassiopeia is that of the ass, the Pleiades have the name ‘murmur’, i. e. a confused thing. Jupiter (?), the companion and guardian of the moon, is held in particular veneration. The marabout in the morning awaits the rising of Venus, and announces by cries, or sometimes by blows on a gong, the hour of prayer. Everyone has his good and bad stars, which the magician takes carefully into account[509]. The intrusion of astrology is not striking, since the people are Mohammedans, while the names of the constellations must be of native origin. The Bakongo call the three stars in Orion’s belt ‘the Dog’, ‘the Palm-rat’, and ‘the Chief Hunter’; Venus is the wife of the moon. The people think that the rain comes from the Pleiades, who are regarded as the ‘Caretakers-who-guard-the-rain’, and if, at the beginning of the rainy season, this constellation is clearly seen, they expect a good rainy season, i. e. rain for their farms without superabundance[510]. The Bangala call the Pleiades a group of young women; five stars in Lepus, kole, are a man with head, hands, and feet; the belt of Orion represents three rowers; five stars in Orion are bundles of thunder and lightning; the evening star also has a name. From the appearance of the Milky Way they draw conclusions as to the lack or abundance of rain; when it is bright and clear there will be much rain[511]. Ten star-names of the Shilluk are given, but only two are translated: the Pleiades are ‘the Hen’, and ‘Three Stars’ is Uranus (sic!). Venus and a fore-runner of Venus are known[512]. The Wagogo know the Milky Way, the Pleiades, and the belt of Orion; the western star of the last-named is to them a boar, the middle star is the dog, and the eastern the hunter[513]. Of the Thonga it is further stated that the stars play a remarkably small part in their ideas. Venus is the best known, the Pleiades is the only constellation with a name; they have no notion whatever of constellations, their mind seems not to have tried to group the stars, or to have seen figures of animals or objects in the sky[514]. In Loango the following constellations are distinguished:—the false Southern Cross (‘the Turtle’), the Scorpion (‘the Serpent’), the Pleiades (‘Ants’), Orion (‘the Fish’), his belt (‘the Line of the Hunter’, who leads a dog), Sirius (‘the Rain-star’). The natives are aware that certain stars move; Jupiter is called ‘the Great Star’, Venus as the evening star is the wife of the moon, as a morning star she is the liar, spy of the moon, or false moon, illusory moon[515].

Far greater knowledge is possessed by the Hottentots, who know the planets accurately. Venus is ‘the Fore-runner of the sun’, or the star at whose rising men run away (i. e. from illicit intercourse), Mercury ‘the Dawn-star’, or the star that comes when the udders of the cows (which are milked morning and evening) are filled again: as an evening star he is not observed. Venus as an evening star is recognised to be the same celestial body as the morning star, and is called ‘the Evening Fugitive’, since it does not remain long in the sky. Jupiter is known, but is sometimes identified with Venus; when however he is seen in ‘the middle of the sky’ he is called ‘the Middle Star’. The six stars of the belt and sword of Orion are grouped together as ‘the Zebras’: δ, ε, ζ are three fugitive zebras against the middle one of which the hunter ι shoots his arrow θ and c. The Pleiades, on account of their thick cluster of stars, are called by a name derived from a verb meaning ‘assemble’, or are otherwise known as ‘the Rime-star’. The Milky Way is called ‘(glowing) Embers’, the Magellanic Clouds ‘Embers’ in the dual. Of single fixed stars our author heard only Sirius called by a name, ‘the Side-star’[516]. The Bushmen divide the stars into night-stars and dawn-stars: of the latter they relate very fine and complicated myths, such as that of the connexion between ‘the Dawn’s Heart’ (Jupiter) and a neighbouring star, his daughter (Regulus or α leonis). Achernar is ‘the Star-digging-stick’s-stone’, or ‘the Digging-stick’s-stone of Canopus’; the Pointers to the Southern Cross are three male lions; α, β, γ crucis are lionesses; Aldebaran is a male hartebeest, α Orion is a female hartebeest, Procyon a male eland, Castor and Pollux his wives, the Magellanic Clouds a steinbok, Orion’s sword three male tortoises hung upon a stick, his belt three female tortoises so hung[517].

The Toda of S. India know the Pleiades, Orion’s sword (‘the Porcupine-star’), the Great Bear, and Sirius, and relate about them myths which are probably borrowed from the neighbouring Badaga[518]. The pagans of the Malay Peninsula know the evening and the morning stars, and the stars of the astrological seasons (sic!), or the Pleiades[519]. In the Indian Archipelago the observation of the Pleiades as a sign of the arrival of the season for sowing is very common. Of the Kayan of Borneo it is stated that though they do not observe the stars or their movements for practical purposes, they are familiar with the principal constellations, and have fanciful names for them and relate mythical stories about the personages they are supposed to represent. The Klementan call Pegasus ‘the padi store-house’, the Pleiades are ‘a well’, the constellation to which Aldebaran belongs is ‘a pig’s jaw’, Orion is a man whose left arm is missing[520].

The natives of Australia have a rich stellar mythology[521]. The evening star has its name and its myths. The Pleiades are women who in the Alcheringa period lived at Intitakula: this is believed by all the tribes whom our authority studied. Orion they regard as an emu, and the stars in general as camp-fires of natives who live in heaven. As a general rule, however, the natives appear to pay very little attention to the stars in detail, probably because these enter very little into anything which is connected with their daily life, more especially with their food-supply. By the northern Arunta and the Kaitish the Magellanic Clouds are supposed to be full of evil magic, which sometimes comes down to earth and chokes men and women in their sleep[522]. According to another author acquainted with the Arunta the Pleiades are seven maidens who had danced at the circumcision ceremony and then ascended into heaven. Two stars in the neighbourhood of the Magellanic Clouds are called ‘the two Gland-poison Men’: the Clouds are the smoke of their fires; the dark patch in the Milky Way is an article of adornment (ngapatjinbi), the Southern Cross ‘an eagle’s foot’. The morning star is also known[523]. The tribes of S. E. Australia give names to many stars and group some of them together in constellations, among which are the sons of Bunjil. The Wiiambo thought that the stars were once great men. The Southern Cross is an emu, Mars an eagle, another star is a crow. The Pleiades, according to the Wotjo-baluh, are some women, corona australis is ‘the Laughing Jackass’, a small star in Argo is ‘the Shell Parakeet’[524].

A very high stage of development in stellar science and mythology is reached among the Euahlayi tribe of the north-west district of New South Wales; anyone interested in the catasterisms of ancient mythology should read the full account given for this tribe. Venus is called ‘the Laughing Star’—the reason for her laughter is a coarse jest—, the Milky Way is an overflow of water. The stars are fires which the spirits of the dead have lit in their journey across the sky, and the dusky haze—i. e. presumably the dark patches without stars, which interest primitive peoples as much as the stars themselves—is the smoke of the fires. A waving dark shadow which you will see along the Milky Way is a crocodile. Two dark spots in Scorpio are devils who try to catch the spirits of the dead; sometimes they come down to earth and make whirlwinds. The Pleiades are seven sisters, ice-maidens; two have been dulled because a man caught them and tried to melt the ice off them: they succeeded in escaping to heaven, but do not shine so brightly as their sisters. The sword and belt of Orion are boys who on earth loved and followed the Pleiades, but after death were turned into stars. In order to remind people of them the Pleiades drop down some ice in the winter, and it is they who make the winter thunderstorms. Castor and Pollux are two hunters of long ago. Canopus is ‘the Mad Star’: he went mad on losing his loves. The Magellanic Clouds are ‘the Native Companions’, mother and daughter, pursued by Wurrawilberoo. ‘The Featherless Emu’ is a devil of water-holes, who goes every night to his sky-camp, ‘the Coal-pit’, i. e. the dark spot beside the Southern Cross. Corvus is a kangaroo, the Southern Crown an eagle-hawk, the Cross the first spirit-tree, a huge yaraon which was the medium for the translation to the sky of the first man who died on earth. The white cockatoos which used to roost in the branches of this tree followed it and became the Pointers[525].

Ridley has obtained from the former chief of the Gingi tribe a long series of star-names. Especially noteworthy for the observation of the risings is the following. The Northern Crown is called mullion wollai, ‘the Eagle’s Nest’, when it stands exactly north on the meridian. Altair rises, and is called mullion-ga, ‘Eagle-in-action’, the eagle springs up to guard his nest. Later Vega rises, and is also called mullion-ga. The ‘holes’ are also well known. The dark spot at the foot of the Cross (the zuu tree) is called an emu, the bird sits under the tree[526]. Elsewhere the star at the head of the Cross is an opossum fleeing from a pursuer—the ‘hole’ between the fore-feet of Centaurus and the Cross[527].

As to the stellar science of the Melanesians we are very variously informed. The tribes of the Torres Straits have a richly developed mythology and observation of the stars[528]. They distinguish the planets from the fixed stars, at least they notice that Venus does not twinkle[529]. The Banks Islanders never travel by night, and consequently do not use the stars in navigation; in consequence of this, says our authority, no definite information about the names of stars or constellations could be obtained. A native gave a few names, but could not point out the stars which they were said to denote[530]. The Moanu of the Admiralty Islands understand the moon and the stars, but the Matankor know neither stars nor moon[531]. A statement such as this must be received with great reserve, especially when it comes from a native of another tribe. In any case it would constitute an exception, since extremely primitive tribes know the stars quite well, the natives of New Britain and of the Solomon Islands even very well. The Pleiades and corona borealis play an important part (cp. [below, p. 141]). The former are called in Lambutjo kiasa, on the Gazelle Peninsula ‘the People-at-the-feast’, and on Bambatana and Alu the year is reckoned according to them: the Crown is called in Lambutjo ‘the Fisher’, in Buin ‘Taro-leaf-greens’, on the Gazelle Peninsula ‘the Thornback’. Further star-names are:—for the Hyades in Buin ‘Earth-rat’, in Lambutjo kapet, a large net for deep water, on the Gazelle Peninsula kakapepe, a kind of small fish, the star in the middle of the constellation is called ‘Hog-fish’. Cygnus is called in Buin ‘Hog-bearer’, in Lambutjo ‘the Three Men’. ‘The Dog’ or ‘Shark’ is a large star ‘that pursues the Fishes’. Many myths are told of the stars[532]. Another authority remarks that the natives of the Solomon Islands are more concerned about the stars than the eastern Polynesians, perhaps because of their longer sea-voyages. The possibility of influence from the astronomically learned Polynesians must also probably be entertained. The people of Santa Cruz and the Reef Islands excel all others in their practical astronomy. The natives of Banks Island and the northern New Hebrides content themselves with distinguishing only the Pleiades, by which the approach of the yam-harvest is marked, and with calling the planets masoi from their roundness, as distinct from vitu, ‘star’. In Florida the early morning star is called ‘the Quartz-pebble-for-setting-off-to-sea’: when it rises later, however, it is ‘the Shining-stone-of-light’. The Pleiades are ‘the Company of Maidens’, Orion’s belt is ‘the War-canoe’, the evening star ‘Listen-for-the-oven’ because the daily meal is taken as evening draws on. All stars are called dead men’s eyes. At Saa the Southern Cross is a net with four men letting it down to catch palolo, and the Pointers are two men cooking what has been caught—because the palolo appears when one of the Pointers rises above the horizon. The Pleiades are called ‘the Tangle’, the Southern Triangle is ‘Three-men-in-a-canoe’, Mars is ‘the Red Pig’[533].

The Polynesians are very learned in astronomy, and their bold and wide sea voyages have helped to make them so, since in these the stars are their principal guide. The Tahitian, Tupaya, who accompanied Cook on his first voyage, could always point out to him the direction in which Tahiti lay[534]. When the Society Islanders put to sea in the evening, as was most commonly the case in their voyages, one constellation, preferably the Pleiades, was chosen as a point to steer by[535]. A detailed report is given for the Marshall Islands:—In the journey from atoll to atoll the course of the boat is commonly directed from a certain passage, island, or promontory to a passage or promontory of the atoll to be reached. Above this spot stands the star that gives the direction. It is the sailor’s business to know for how many hours a star can serve him as compass, so that immediately after the apparent turning of the star from east to west he may choose another. Of great interest also is the idea of the connexion between the atmospheric and other phenomena and the stars. Certain periods of bad weather recur every year with tolerable regularity, so that the sailors attribute them to the immediate influence of the stars. When, for instance, at 4 o’clock in the morning—at which time the signs of the weather are observed—the stars stand just above the eastern horizon, they stop up the east, so to speak, and prevent the free passage of the wind. But if the pernicious star in question is at the given time 20° or 30° above the horizon, there is enough space between star and horizon for the wind to be released. This strong wind will last until another influential star arises under the first. This lower star acts like a wind-chute placed against an open hut. The strength of the wind is therefore reduced. This explains why every storm is followed by a wind favourable for sailing. For example when Spica is 20° above the horizon a violent storm is developed, but this only lasts until Arcturus some time later becomes visible on the eastern horizon. The most important of the stars that bring bad weather are Spica, Arcturus, Antares, the claw of the Scorpion, Altair, Delphinus, β, μ, λ and γ, ξ, π Pegasi. With the rising of Cassiopeia the time of calms begins. Jedada (γ, ζ, π aquilae) ‘disembowels the heavens’. Altair is regarded as a bad fellow. When he rises in the east before dawn it is commonly a time when food supplies have run low, so that quarrels arise: only when he rises higher and the hot season (June-August) brings plenty of food, do reconciliation and goodwill return. Of ‘King Jäbro’, the Pleiades, long myths are related: when they emerge from the horizon joy prevails, but tears are shed when they vanish again into the west[536]. The knowledge of the stars was often a carefully guarded secret, but through prevailing European influence it has now fallen entirely into decay. In Samoa it is now an exception for a native to know the name of this or that constellation, since an islander engaged in the fishing trade can only indicate and name this or that star if it marks the beginning of some important native occupation[537].

The Polynesian material for star-names is exceedingly abundant, and can here only be represented in outline, so as to give some idea how far astronomy may advance at this stage of civilisation[538]. The Marquesas Islanders know and name a great number of constellations and separate stars, e. g. ‘the Little Eyes’ (the Pleiades), ‘the Rudder’ (Orion’s belt)[539]. Constellations mentioned as being known to the Society Islanders are:—the Pleiades, Orion’s belt, Sirius (‘Big Star’), the Magellanic Clouds (the upper and lower ‘Haze’), the Milky Way (‘the Long-blue-cloud-eating-shark’), Venus, called sometimes ‘Day-star’ or ‘Herald-of-the-morning’, and sometimes ‘Taurna-who-rises-at-dusk’, Mars (‘the Red Star’), Jupiter, and Saturn[540]. The people of Nauru, west of the Gilbert Islands, observe the stars, chiefly the Pleiades, Orion, Sirius, and the morning and evening stars[541]. For the Marshall Islands see [above, p. 125]. For Tahiti names are given for Venus, Jupiter, Saturn, the Pleiades (‘Star-of-the-nest’), Sirius (‘Big Star’), and the belt of Orion, and it is further stated that many other stars are known by separate names[542]. The Hawaiians had names for many constellations, and they also knew the five planets[543]. An apparently distinguished native astronomer, named Hoapili, stated that he had heard from others (Europeans?) that there was one more travelling-star, but he had never observed it, and was acquainted only with the five[544]. The Maoris had names for all the principal stars and for a great number of constellations. The most important of the latter is ‘the Canoe of Tamarereti’, which consists of the following parts:—the three stars of Orion’s belt form the stern, matariki (the Pleiades) is the prow, te toke o te waka is the mast, the Southern Cross is the anchor, and the two Pointers are the cables. Further, Orion’s belt is called ‘the Elbow of Maui’; the Scorpion is ‘the House-of-Te-Whiu-and-his-slaves’; Waka mauruiho and Waka mauruake are the husbands of Hurike and Angake, and their daughters are Tioreore and Tikatakata, the two Magellanic Clouds, whose husbands are Taikeha and Ninikuru. By the position of the Magellanic Clouds the natives think they can tell from what quarter the wind will blow. One constellation is called ‘the Garment of Maru’, which he let fall as he ascended into heaven. Unfortunately the names corresponding to our star-map are not given, and I have omitted many which are not translated[545]. Some stars are mentioned below in the account of the Maori calendar of months[546].

The Micronesians know the stars well; long lists of star-names come from the Carolines. 18 names are given for Ponape, among them names for the Pleiades, the Southern Cross, and the Magellanic Clouds; from Lamotrek come 24, e. g. ‘the Leather-jacket-fish’ (the Southern Cross), ‘the Broom’ (Ursa Minor), ‘the Virile Member’ (Aldebaran), ‘the Body-of-the-animal’ (Sirius), ‘the Centre-of-the-house’ (Arietes), ‘the Two Eyes’ (Scorpio), ‘the Fowling-net’ (Corona), ‘the Tail-of-the-fish’ (Cassiopeia), etc.; from Mortlock 23, e. g. (Ursa Minor) fusa-makit, ‘the Seven Mice’, or it may mean ‘the Star-that-changes-its-position’ (sic!), Leo, ‘the Rat’, the Southern Cross (perhaps), ‘the Shark’, Delphinus and Cygnus, ‘the Bowl-in-the-midst-of-Sota’, Sirius, ‘the Animal’, Orion and Aldebaran, ‘The Branch-of-the-tree’, not identified, ‘the Fish-net’; from Yap 25, unidentified[547]. The Fijians on the other hand knew little about the stars. They had no names even for the most important constellations. The evening and morning stars were known, under the names of ‘Marking-day’ and ‘Marking-night’, but the natives did not distinguish between the planets and the fixed stars. Their ignorance is ascribed to the fact that they never undertake voyages beyond the limits of their groups, and are bad navigators in the technical sense, although good sailors[548].

Stellar science and mythology are therefore wide-spread among the primitive and extremely primitive peoples, and attain a considerable development among certain barbaric peoples. Although this must be conceded, some people are apt to think that the determination of time from the stars belongs to a much more advanced stage: it is frequently regarded as a learned and very late mode of time-reckoning. Modern man is almost entirely without knowledge of the stars; for him they are the ornaments of the night-sky, which at most call forth a vague emotion or are the objects of a science which is considered to be very difficult and highly specialised, and is left to the experts. It is true that the accurate determination of the risings and settings of the stars does demand scientific work, but not so the observation of the visible risings and settings. Primitive man rises and goes to bed with the sun. When he gets up at dawn and steps out of his hut, he directs his gaze to the brightening east, and notices the stars that are shining just there and are soon to vanish before the light of the sun. In the same way he observes at evening before he goes to rest what stars appear in the west at dusk and soon afterwards set there. Experience teaches him that these stars vary throughout the year and that this variation keeps pace with the phases of Nature, or, more concretely expressed, he learns that the risings and settings of certain stars coincide with certain natural phenomena. Here, therefore, there lies ready to hand a means of determining the time of the year, and one which is indeed much more accurate than a method depending on a reference to the phases of Nature. However it would seem as if this mode of indicating time would require a greater knowledge of the stars, such as only few peoples possess,—as if it would constantly be necessary to observe a fresh star for each of the smaller divisions of time. This is not the case, since, as appears from statements already made, for the purpose of determining the seasons a star may be observed when it is stationed at other positions in the sky than on the horizon, e. g., very conveniently, at its upper culmination, but other positions, expressed by us in so many degrees above the horizon, may also serve. Just as the advance of the day is discerned from the position of the sun, so the advance of the year is recognised by the position of certain stars at sunrise and sunset. Stars and sun alike are the indicators of the dial of the heavens. A determination of this kind, however, is not so accurate as that from the heliacal risings and settings. Hence the latter pass almost exclusively or at least pre-eminently under consideration wherever, as in Greece, a calendar of the natural year is based upon the stars: sometimes however the upper culmination (μεσουράνημα) is also given. Finally the stars can also be observed at other times of night than just before sunrise or after sunset[549]: the Marshall Islanders, for instance, were accustomed to observe the signs of the weather at 4 a. m. With the lack of a means of accurately telling the time such an observation is very uncertain and unpractical, and is therefore seldom found.

In order to determine the time of certain important natural phenomena it is therefore sufficient to know and observe a few stars or constellations with accuracy and certainty. The Pleiades are the most important[550]. It has been asked why this particular constellation, consisting as it does of comparatively small and unimportant stars, should have played so great a part, and the answer given is chiefly that its appearance coincides (though this is true of other stars also) with important phases of the vegetation. This is correct, but something else must be added. To create constellations in which terrestrial objects, animals, and men are arbitrarily seen requires no inconsiderable degree of imaginative power. The Pleiades however form themselves into a group without any aid from the imagination, and can without difficulty be recognised as such. It is because they are easy to recognise immediately that the observation of these stars plays so important a part. A similar case is that of the Magellanic Clouds, which, where they are visible, belong to the best known phenomena of the heavens, and we may also compare the dark starless patches which so largely occupy the attention of primitive peoples, although neither of these two phenomena is used in determining time, since neither can be observed at the favourable moment, viz. the twilight.

An account of the Bushmen shews how extremely primitive peoples may also observe the risings of the stars, may connect them with the seasons, and—which is indeed somewhat rare—may even worship them. The Bushmen perceive Canopus; they say to a child:—“Give me yonder piece of wood that I may put (the end of) it (in the fire), that I may point it burning towards grandmother, for grandmother carries Bushman rice; grandmother shall make a little warmth for us; for she coldly comes out; the sun shall warm grandmother’s eye for us”. About the same time as Canopus, Sirius appears, and a similar ceremony takes place. Sirius comes out: the people call to one another:—“Ye must burn (a stick) for us (toward) Sirius.” They say to one another: “Who was it that saw Sirius?” One man says to the other: “One brother saw Sirius.” The other man says to him: “I saw Sirius.” The other man says to him: “I wish thee to burn a stick for us towards Sirius, that the sun may shining come out for us, that Sirius may not coldly come out.” The other man says to his son: “Bring me the piece of wood yonder, that I may put it in the fire, that I may burn it towards grandmother, that grandmother may ascend the sky, like the other one”, i. e. Canopus. The child brings him the piece of wood, he holds it in the fire. He points it burning towards Sirius, he says that Sirius shall twinkle like Canopus. He sings; he points to them with fire that they may twinkle like each other. He throws fire at them[551]. Canopus and Sirius appear in winter, hence the cold is connected with them. The ceremony just described is obviously a warming-incantation. It is said also that it will make the stars rise higher, for the higher they stand above the eastern horizon at sunrise and the more brightly they twinkle, the more nearly winter draws towards an end. The Hottentots connect the Pleiades with winter. These stars become visible in the middle of June, that is in the first half of the cold season, and are therefore called ‘Rime-stars’, since at the time of their becoming visible the nights may be already so cold that there is hoar-frost in the early morning. The appearance of the Pleiades also gives to the Bushmen of the Auob district the signal for departure to the tsama field[552].

The Euahlayi tribe also connect the Pleiades with the cold: they call the stars ‘the Ice-maidens’, imagine them to be covered with ice, and say that in winter they let ice drop on the earth and also cause the winter thunderstorms[553]. Another tribe danced in order to win the favour of the Pleiades; the constellation is worshipped by one body as the giver of rain, but should the rain be deferred, instead of blessings curses are apt to be bestowed on it[554]. The Arunta say that the Pleiades are seven maidens who ascended into heaven, but after many wanderings came back to Okaralyi, where they again gathered ugokuta fruit and danced in the women’s dance. During this period the Pleiades are not to be seen in the sky, i. e. it is the time between the evening setting and the morning rising. Here therefore the constellation is connected with a phase of Nature, and the whole is mythologically explained. According to another Arunta myth the Pleiades are maidens who had danced at a circumcision ceremony. After they had taken part in all the ceremonies in which to-day the assistance of women is still requisite at this festival, they went back to their native district, whence they ascended to heaven and are now to be seen as the Pleiades. Not without reason did the circumcision most frequently take place at the season when the Pleiades rise at evening in the east and remain in the sky all night long (this is the case in the summer months), so that this prominent constellation was regarded as a spectator of the festivities connected with the rite[555]. The Pleiades therefore serve to determine the time of the feast, and this circumstance is again invested with a myth. A tribe of Western Victoria connected certain constellations with the seasons. The Pleiades are young maidens playing to a corroboree-party of young men, represented by the belt and sword of Orion. Aldebaran, ‘the Rose-crested Cockatoo’, is an old man keeping time for the dancers. This group corresponds with the months of November and December. As the year advances Castor and Pollux appear: they are two hunters who pursue and kill a kangaroo, Capella. The Mirage is the smoke of the fire at which the kangaroo is cooked by the successful hunters. Those two groups set forth the period of the summer. The breaking up of a prolonged drought is thus explained:—Berenice’s Hair, which culminates in March, is a tree with three big branches. When a shower of rain has come, every drop is nevertheless sucked up by the dusty earth. A small cavity formed at the junction of the three branches has however retained a little water, and here it is imagined some birds drink. The winter stars are Arcturus—who is held in great respect since he has taught the natives to find the pupae of the wood-ants, which are an important article of food in August and September—and Vega, who has taught them to find the eggs of the mallee-hen, which are also an important article of food in October. The natives also know and tell stories of many other stars[556]. Another authority states that they can tell from the position of Arcturus or Vega above the horizon in August and October respectively when it is time to collect these pupae and these eggs[557]. An old chief of the Spring Creek tribe in Victoria taught the young people the names of the favourite constellations as indications of the seasons. For example when Canopus at dawn is only a very little way above the eastern horizon, it is time to collect eggs; when the Pleiades are visible in the east a little before sunrise, the time has come to visit friends and neighbouring tribes[558].

The Chukchee form out of the stars Altair and Tarared in Aquila a constellation named pchittin, which is believed to be a forefather of the tribe who, after death, ascended into heaven. Since this constellation begins to appear above the horizon at the time of the winter solstice, it is said to usher in the light of the new year, and most families belonging to the tribes living by the sea bring their sacrifices at its first appearing[559].

Among the N. American Indians the determination of time from constellations is rare. The Blackfeet Indians regulate their most important feasts by the Pleiades, a feast is held about the first and the last day of the occultation of these stars. It includes two sacred vigils and the solemn blessing and planting of the seed, and is the opening of the agricultural year[560]. According to another legend of the same tribe, the Pleiades are seven children who ascended into heaven because they had no yellow hides of the buffalo calves. Therefore the Pleiades are invisible during the time when the buffalo calves are yellow (the spring). But when these turn brown, in autumn, the lost children can be seen in the sky every night[561]. Among the Tusayan Indians of Arizona the culmination of the Pleiades is often used to determine the proper time for beginning a sacred nocturnal rite[562].

The S. American Indians have much greater knowledge of the stars, and in consequence frequently connect stellar phenomena, especially those of the Pleiades, with phases of Nature. In north-west Brazil the Indians determine the time of planting from the position of certain constellations, in particular the Pleiades. If these have disappeared below the horizon, the regular heavy rains will begin. The Siusi gave an accurate account of the progress of the constellations, by which they calculate the seasons, and in explanation drew three diagrams in the sand. No. 1 had 3 constellations:—‘a Second Crab’, which obviously consists of the three bright stars west of Leo, ‘the Crab’, composed of the principal stars of Leo, and ‘the Youths’, i. e. the Pleiades. When these set, continuous rain falls, the river begins to rise, beginning of the rainy season, planting of manioc. No. 2 had 2 constellations:—‘the Fishing-basket’, in Orion, and kakudzuta, the northern part of Eridanus, in which other tribes see a dancing-implement. When these set, much rain falls, the water in the river is at its highest. No. 3 was ‘the Great Serpent’, i. e. Scorpio. When this sets there is little or no rain, the water is at its lowest[563]. The natives of Brazil are acquainted with the course of the constellations, with their height and the period and time of their appearance in and disappearance from the sky, and according to them they divide up their seasons. In the valley of the Amazon it is said that during the first few days of the appearance of the Pleiades, while they are still low, birds, and especially fowls, roost on low branches or beams, and that the higher the constellation rises the higher the birds roost also. These stars bring cold and rain: when they disappear the snakes lose their poison. The canes used for arrows must be cut before their appearance, or else the arrows will be worm-eaten. The Pleiades disappear, and appear again in June. Their appearance coincides with the renewal of the vegetation and of animal life. Hence the legend says that everything that has appeared before the constellation will be renewed, i. e. its appearance marks the beginning of spring[564]. The Bakairi reckoned by natural phases, but were also well acquainted with astronomical signs, and spoke of certain constellations which reappeared at the beginning of the dry season: they referred to stars in the vicinity of Orion, ‘the Manioc-pole’[565]. The Tamanaco of the Orinoco called the Pleiades ‘the Mat’. They recognised the approach of winter from the signs of Nature[566], but also from the fact that the Pleiades at sunset were not too far distant from the western horizon: the evening setting falls at the beginning of May[567]. The Lengua Indians of Paraguay connect the beginning of spring with the rising of the Pleiades, and at this time celebrate feasts which are generally of a markedly immoral nature[568]. The Guarani of the same country recognised the time of sowing by the observation of the Pleiades[569]. The Guarayu call the Pleiades piangi; when they disappear the dry season begins, and when Orion is no longer visible a period of cold dew begins. The Chacobo of north-eastern Bolivia regulate the time of sowing by the position of the Pleiades in relation to the spot where the sun rises[570]. The Chané and Chiriguano do the same. When the Pleiades rise above the horizon very early in the morning, the time for sowing has come: it is important for this to be finished before the rainy season sets in[571]. Still further tribes, for which I refer to Frazer, relate myths about the Pleiades, worship them, and celebrate feasts at their appearance. So did the inhabitants of ancient Peru, who called the Pleiades ‘the Maize-heap’[572]. It might probably be thought that the observation of the Pleiades has spread from this ancient civilised people among the inhabitants of S. America, but it is of so primitive a character that it rather appears to have been one of the rudiments of the astronomical knowledge of the people of the Incas.

In Africa also the observation of the stars, and above all of the Pleiades, is wide-spread. In view of the dissemination of this knowledge all over the world it is making a quite unnecessary exception to state that it came into Africa from Egypt. Moreover this assertion does not correspond with the facts, since among the Egyptians Sirius, and not the Pleiades, occupied the chief place. The observation of the appearance of Canopus and Sirius we have already found highly developed among the Bushmen, that of the Pleiades among the Hottentots. The Bechuana of Central S. Africa are directed by the positions of certain stars in the heavens that the time has arrived in the revolving year when particular roots can be dug up for use, or when they may commence their labours of the field. This is their likhakologo (‘turnings’ or ‘revolvings’), at what we should call the spring-time of the year. The Pleiades they call selemela, which may be translated ‘cultivator’ or ‘the precursor of agriculture’ (from lemela, ‘to cultivate for’, and se, a pronominal prefix, distinguishing these stars as the actors). When the Pleiades assume a certain position in the heavens it is the signal to commence cultivating their fields and gardens[573]. The Caffres determine the time of sowing by observing the Pleiades[574]; the Bantu tribes of S. Africa regard their rising shortly after sunset as indicating the planting-season[575]. The Amazulu call the Pleiades isilimela, which has the same meaning as the Bechuana name, since they begin to dig up the soil when the Pleiades appear. The people say: ‘isilimela dies and is not seen’, and at last, when winter is coming to an end, it begins to appear, one of its stars first and then three, until, continuing to increase, it becomes a cluster of stars and is perfectly clearly seen when the sun is about to rise. Then they say: ‘isilimela is renewed’, ‘the year is renewed’, and they begin to dig[576]. Among the Thonga the Pleiades are the only constellation which bears a name—shirimelo; it rises in July and August, when tilling is resumed[577]. At the southern corner of Lake Nyassa the rising of the Pleiades early in the evening gives the sign to begin the hoeing of the ground[578]. The Kikuyu of British East Africa say that this constellation is the mark in the heavens to shew the people when to plant their crops: they plant when it is in a certain position early in the night. A dancing-song begins:—“When the Pleiades meet the moon, the people assemble etc.”[579] The Masai know whether it will rain or not according to the appearance or non-appearance of the Pleiades, and the last month of the period of the great rains, in which their evening setting falls, is named after them. When they are no longer visible the people know that the great rains are over, and they are not seen again until the following season—the season of showers—has come to an end. The Masai call the sword of Orion ‘the Old Men’, and his belt ‘the Widows’ who follow them[580].

To the Isubu in Kamerun the constellations, which they combine in certain groups, shew the course of the seasons; such constellations are e. g. tole a nyou, the tole of the elephants, in contradistinction to tole a moto, the tole of men; another is ‘the Orphans’. These are summer signs, they are all found in the eastern part of the sky[581]. In Sierra Leone the proper time for planting is shewn by the position in which the Pleiades are to be seen at sunset: the Bullom do not observe or name any other stars[582]. The Bakongo associate these stars with the rainy season: the rain comes from them, they are called ‘the Caretakers-who-guard-the-rain’[583]. When the constellation kole[584] reaches the meridian, the Bangala plant more than at any other time, because the rains, though not infrequent, are then fairly certain[585]. In Loango Sirius is called ‘the Rain-star’, since as long as he is visible the rains persist. Alongside of him Orion is regarded as a sign of the rainy season[586]. In French Guinea the people know that when the winter constellations appear above the horizon, indicating that the end of the rains has come, it is the time of harvest[587].

In the Indian Archipelago the observation of the Pleiades is the most general and frequent means of determining the time for tillage. Hence these stars are mythologically regarded as the originators of the rice-culture. The Dyaks of Sarawak say that Si Jura on a sea-voyage once found a fruit-tree with its roots in the sky and the branches hanging downwards. He climbed up into it, and since his comrades sailed away, he was obliged to climb on and on until he reached the roots and found himself in a strange land—the country of the Pleiades. There Si Kira received him kindly, and invited him to eat. “Those little maggots?” replied Si Jura. Si Kira answered:—“They are not maggots, but boiled rice”, and he explained to his guest how the rice was cultivated and reaped, and then let him down by a long rope near to his father’s house. Si Jura taught the Dyaks how to cultivate rice, and the Pleiades themselves tell them when to farm; according to the position of these stars in the heavens, morning and evening, they cut down the forest, burn, plant, and reap[588]. In another legend the Pleiades are six chickens which the hen follows, invisible; formerly there were seven, and at that time men did not know of rice, but lived on the products of the forest. One of the chickens had come down to earth, where men gave it to eat: it would not eat, however, but brought them a fruit with three husks, in which there were contained three kinds of rice, that would ripen in four, six, and eight months respectively. The hen was angry, and wished to destroy both men and the chicken: the former were saved by Orion, but only six chickens were left. During the time in which the Pleiades are invisible, the hen is brooding, but the cuckoo calls as long as they are visible[589]. The Sea-Dyaks determine the time of sowing by observing the Pleiades. Some tribes determine the approach of the time of rice-sowing from the observation of the stars. The Kayan of Borneo know the most important constellations, although they do not observe them and their motions with a practical end in view[590]. However one of the joint authors just quoted says in another place that although the Kayan more usually determine the time of sowing by the observation of the sun, yet both they and many other races in Borneo sow the rice when the Pleiades at daybreak appear just above the horizon[591]. When the time to clear fresh land in the forest draws near, a wise man is appointed to go out before dawn and watch for the Pleiades. As soon as they are seen to rise while it is still dark, the people know that the time has come to begin work, but not until they are at the zenith before dawn is it considered desirable to burn the fallen timber and sow rice. The Dyaks begin the rice-planting when the Pleiades reach the same position at about 3 or 4 o’clock in the morning as the sun reaches at 8 o’clock. Old and experienced men are on the watch to determine the spot exactly. Then a feast begins[592]. The natives of Nias, an island to the south of Sumatra, assemble to till their fields when the Pleiades appear, and regard it as useless to do so before that time[593]. In Sumatra also the time for sowing was determined in this way. The Batak of the middle of the island regulate their various agricultural operations by the position of Orion and the Pleiades. The Achenese of the north know that the sowing-time has come when the Pleiades rise before the sun, at the beginning of July[594]. In northern Celebes the rice-fields are prepared for cultivation when the Pleiades are seen at a certain height above the horizon[595]. The Kai of German New Guinea say that the time for labour in the fields has come when the Pleiades are visible above the horizon at night: the Bukaua of the same country also follow the Pleiades[596]. When the natives of the Torres Straits Islands see the Pleiades on the horizon after sunset, they say that the new yam-time has come[597]. The western tribes of these straits have names for many stars, which are largely grouped into constellations. The seasonal appearances of certain stars or constellations were noted, and their rising regulated particular dances, and also, as our authority thinks, the planting of yams and sweet potatoes[598].

Accurate information for these tribes is given by Rivers in the Reports of the Expedition to the Torres Straits. The most important constellations are ‘the Shark’ (= the Great Bear together with Arcturus) and corona borealis. Still larger is Tagai. This constellation represents a man, Tagai (= Centaurus, Lupus), standing in the prow of a canoe (Scorpio); in the stern sits Kareg (Antares). Tagai holds in his left hand (the Southern Cross) a fishing-spear, in his right (Corvus) some kupa-fruit. Below the canoe is a sucker-fish, consisting of a part of Scorpio. Naurwer are ‘the Brothers’—Vega the elder, and Altair the younger—who in their outstretched arms are holding sticks (β, γ lyrae, β, γ aquilae). In Mabuiag this constellation is called Dogai. Our Delphinus is called ‘the Trumpet-shell’, kek is probably Achernar. Others I omit. The most important star was kek, whose rising indicated not only the beginning of many ceremonies but also the planting-season. The risings and settings of the stars were observed, and certain rites and agricultural occupations regulated thereby. In Badu it was said that when only the tail of the Shark is above the horizon, the north-west wind begins to blow ‘a little bit’: when the tail has gone down altogether, the people begin to plant yams, and when the Shark comes up again, yams, sweet potatoes, and bananas are ripe. The stars also help to determine the seasons. A native of Mabuiag gave the following list of the stars relating to the season called aibaud:—kek comes up, he is the sign for everything to be done: ‘start meeting’, i. e. at the feasts the holding of which is dependent upon plentiful supplies of food; gil, usal (the Pleiades): at this time the ovaries of the turtles enlarge; pagas and dede (Betelgeuze); utimal; wapil. Towards the end of the season the Shark becomes visible, and then the pigeon migrates from New Guinea to Australia, as does the birubiru-bird when gitulai (the Crab) appears. It is expressly noted that when the people speak of the rising or setting of a constellation or star at a certain season, they have in mind the time of the year when the star or constellation in question first appears or disappears on the horizon at daybreak. Of Tagai a catasterism is related which at the same time has reference to the phenomena of the seasons at the appearance of the stars in question. On a fishing expedition the crew stole the water from him and Koang. They therefore killed them and said:—“Usal (the Pleiades), you go to New Guinea side, when you come up there will be plenty of rain. Utimal, you go to New Guinea side, you have to bring rain. Kwoior, when you come up over Mangrove Island just before the south-east monsoon sets in, there will be rain in the morning. Then the wind will shift and it will rain in the afternoon, and you, Kek, will come up in the south between Badu and Moa and it will be cold weather. When you go round this way and when you come up, then the yams and sweet potatoes will ripen. You all have work to do”[599]. A similar story is told of the Kiwai Papuans, who have for the most part the same star-names and call most of their months after stars: the Shark is also implicated in this story. When the fin sets, there is more wind and high-water; when the tail sets, more high-water; when the head rises, the copulating-season of the turtles commences. Another myth tells how Javagi got angry and threw Karongo up into heaven, where he and his three-pronged spear became the constellation Antares[600].

The Melanesians of Banks Island and the northern New Hebrides are also acquainted with the Pleiades as a sign of the approach of the yam-harvest[601]. The inhabitants of New Britain (Bismarck Archipelago) are guided in ascertaining the time of planting by the position of certain stars[602]. The Moanu of the Admiralty Islands use the stars as a guide both on land and at sea, and recognise the season of the monsoons by them. When the Pleiades (tjasa) appear at night-fall on the horizon, this is the signal for the north-west wind to begin. But when the Thornback (Scorpio) and the Shark (Altair) emerge as twilight begins, this shews that the south-east wind is at hand. When ‘the Fishers’ Canoe’ (Orion, three fishermen in a canoe) disappears from the horizon at evening, the south-east wind sets in strongly: so also when the constellation is visible at morning on the horizon. When it comes up at evening, the rainy season and the north-west wind are not far off. When ‘the Bird’ (canis major) is in such a position that one wing points to the north but the other is still invisible, the time has come in which the turtles lay eggs, and many natives then go to the Los-Reys group in order to collect them. The Crown is called ‘the Mosquito-star’, since the mosquitoes swarm into the houses when this constellation sets. The two largest stars of the Circle are called pitui an papai: when this constellation becomes visible in the early morning, the time is favourable for catching the fish papai[603]. The natives of the Bougainville Straits are acquainted with certain stars, especially the Pleiades; the rising of this constellation is a sign that the kai-nut is ripe: a ceremony takes place at this season[604]. On Treasury Island a grand festival is held towards the end of October, in order—so far as could be ascertained—to celebrate the approaching appearance of the Pleiades above the eastern horizon after sunset. In Ugi, where of all the stars the Pleiades alone have a name, the times for planting and taking up yams are determined by this constellation[605]. In Lambutjo the year is reckoned according to the position of the Pleiades. When they are in the east, it is said that ‘they are waiting’, when at the zenith, ‘they stand in the middle’, when in the west, they are ‘bowed down’. When they stand low, the turtles come up on land: the people say that they ‘go to play’, i. e. it is the pairing season. When the Pleiades are high overhead, the white men celebrate Christmas. When they ‘come up anew’, the people go to look for fish. At that time ‘the Fishes’ are in the water. ‘The Fishes’ (corona borealis) dip down when the Pleiades come up. When ‘the Fishes’ are in the sky, there are no fish in the water. In both Alu and Lambutjo one division of the year is reckoned by the return of the Pleiades, another by the almond-ripening. On the Gazelle Peninsula the time for good fishing is the time of the appearance of the Pleiades: at this time the fishing-nets are spread out. It is said that ‘the Thornback’ (Pisces) and ‘the People-at-the-feast’ (the Pleiades) must not see each other; the former constellation is called galial (‘fishes’), which at this time are not to be eaten[606]. On the island of Saa, one of the Solomon Islands, the Southern Cross is the net with four men letting it down to catch palolo, and the Pointers are two men cooking what is caught, since the palolo first comes when one of the Pointers appears above the horizon[607]. In the list of star-names given for the Carolines there are also references to the seasons. In Ponape le-poniong is seen at the time of the variable winds. In Lamotrek Corvus is called ‘the Viewer-of-the-taro-patches’, since he is visible during the taro season; the name of Arcturus is formed from ara, ‘to conclude’, and moi, ‘to come’, and the star is so called because his rising indicates the end of the north-east winds, which bring visiting parties to the island; the appearance of Capella means heavy gales and bad weather[608].

Among the astronomically learned Polynesians time-estimations according to stars play an important part: most of these however belong to the chapters on the months and the year. In Samoa it is at present an exception if an old fisherman can indicate and name this or that star which at its entrance into this or that constellation (sic!) announces the beginning of an abundant bonino-catch, the immediate return of the South Sea herring, the atuli, to its accustomed spawning-grounds, or some other similar event of importance in the life of the natives[609].

When the stars indicate this or that event, the primitive mind, as so often happens, is unable to distinguish between accompanying phenomena and causal connexion; it follows that the stars are regarded as authors of the events accompanying their appearance, when these take place without the interference of men. So in ancient Greece the expressions (a certain star) ‘indicates’ (σημαίνει) or ‘makes’ (ποιεῖ) certain weather were not kept apart, and the stars were regarded as causes of the atmospheric phenomena[610]. A similar process of reasoning is not seldom found among primitive peoples, and a few instances have already been given, such as the warming-incantation of the Bushmen against Canopus and Sirius, the name given to the Pleiades among the Bakongo (‘the Caretakers-who-guard-the-rain’), and the belief that the rain comes from them, the myth of the Euahlayi tribe that the Pleiades let ice fall down on to the earth in winter and cause thunderstorms, in other words send the rain, and the belief of the Marshall Islanders that the various positions of certain stars cause storms or good winds[611]. The same idea is very clearly seen in the account of the Hottentots given by a missionary of the 17th century[612]. At the return of the Pleiades the natives celebrate an anniversary: as soon as the stars appear above the eastern horizon the mothers lift their little ones in their arms, run up to some eminence, and shew to them these friendly stars, and teach them to stretch out their hands towards them. The people of the kraal assemble to dance and sing according to the old custom of their ancestors. The chorus is always: “O Tiqua, our father above our heads, give rain to us that the fruits (bulbs etc.), uientjes, may ripen and that we may have plenty of food: send us a good year!”

The natives of Australia (perhaps of Victoria), according to an old account, worship the heavenly bodies and think that natural causes are governed by certain constellations. They have names for these, and sing and dance to win the favour of the Pleiades, which are worshipped by one group as the giver of rain; should the rain be deferred, curses instead of blessings are bestowed on them[613]. The Euahlayi tribe thinks that the Pleiades bring frost and winter thunderstorms, and that the Milky Way by its change of position brings rain[614]. An old native, chief of the Gingi tribe, when the rain would not stop, turned to the souls of his dead friends in the Milky Way with certain charms, until they made the rain cease. The Milky Way is regarded as a stream with fertile banks[615].

These facts being so, there is nothing strange in an account which unfortunately comes from a writer whose evidence in other respects is open to grave doubt. We are told that Andy, a native of New South Wales, found the statement that the sun is the source of heat ridiculous, and said:—“If the sun makes the warm weather come in summer-time, why does he not make the winter warm, for he is seen every day?” The influence which produces heat, in the belief of the natives, accompanies the Pleiades. When these are visible at a certain altitude above the horizon, it is spring, begagewog; when they rise to their highest altitude, it is summer, winuga; when in autumn they sink down again towards the horizon, it is domda (‘autumn’); in winter they are barely visible or are lost to view altogether; it is then winter (magur), and cold. The ordinary stars have no kind of influence on the seasons, but simply the Pleiades[616]. The account agrees very well with what is otherwise known of the stellar science of the Australians, and is perfectly credible. A precisely similar story comes from the other side of the globe. At the beginning of the 18th century, when the Lapps were still heathens, one of the questions which a missionary among these people put to them about their gods was:—“Have you prayed the Pleiades to warm the weather?” In accordance with this a Lapp myth relates that a servant driven out on a very cold night by a cruel master was saved by the Pleiades. One of the Lapp names for these stars, which evidently points to this idea, is ‘the Sheep-skins’[617]. The Greeks had the same belief in Sirius as the cause of the summer heat.[618]

From this belief in the stars as causes of the natural phenomena it is but a short step to attempt to draw from the manner of their appearance conclusions as to the kind of phenomenon caused by them. To the Bakongo the Pleiades are the guardians of the rain, and when they are clearly to be seen at the beginning of the rainy season the people expect a good season, i. e. sufficient but not too much rain[619]. The Nandi of British East Africa know by the appearance or non-appearance of the Pleiades whether they may expect a good or a bad harvest[620]. The Guarayu of S. America believe that when the Pleiades at their reappearance are surrounded by a circle, it is a good omen: but if this circle is wanting, all must die[621]. In Macedonia the Pleiades are called ‘the Clucking or Brooding Hen’ (ἡ κλωσσαριά); their setting announces the advent of winter, and from the accompanying conditions omens are drawn as to the quantity of the forthcoming crop and the fertility of the cattle. If the constellation sets in a cloudy sky, this portends a rich harvest[622]. Similar weather-rules and prognostications are found in abundance in modern European folk-lore and in the so-called peasants’ calendars. The origin in the popular astrological beliefs of antiquity is usually taken for granted. It is true that astrology, especially under Mohammedan influence, has penetrated very deeply even among little civilised peoples such as the negroes of Central Africa and the Malays of the Indian Archipelago; but I see no cogent reason for finding in the above-mentioned world-wide examples of a belief in the influence of the stars upon natural phenomena any influence of that astrology which derives from ancient Babylon. Rather do these myths and traditions seem to afford an analogy to the initial stages of the Babylonian astrology, and to shew that the whole vast system of astrology had its root in primitive thinking. And the Babylonian prognostications from stars and sky remained, until a very late period, quite primitive. These observations cannot be followed up further: astrology and its origins lie outside the limits of the present study.

It has been shewn, then, that even among the most primitive peoples of the globe the stars are known, observed, considered, and used for the determination of time—the Pleiades, indeed, first and foremost, but other constellations as well; of the not nearly so frequent determination of the advance of night from the motions of the stars we have already spoken in chapter I. There is however a difference that should not be neglected between this method of determining time and the time-indications from natural phases. So far as I have been able to discover, the stars are never used in a narrative, i. e. where the date of any familiar event is to be given, but only where practical rules for the constantly recurring occupations and labours are concerned, and also for the festivals. The method therefore does not apply to the historical event in the wider sense, but only to the reiterated event the recurrence of which is empirically known. The consciousness of a fixed and constant order is therefore impressed upon the mind of primitive man much more powerfully by the eternal revolution of the constellations than by the variation of the seasons.

CHAPTER V.
THE MONTH.

The course of the sun determines the variation between day and night, and causes the natural phases of the year. From the position of the sun the times of the day can be given with ease and certainty, but not so the seasons of the year,—to the exceptions I shall recur in chapter XII. From the fixed stars the hours of the night can be determined, and still more frequently are the seasons regulated by them. But this kind of time-determination necessarily refers to points of time, and not to periods. Only for one or two days has the star the position which serves for the determination of time. No division of the year into parts can be carried out by this method, the most that can be done is to regulate the already existing divisions by it.

As well as the sun and the fixed stars the moon appears in the heavens. It does not entirely vanish before the sunlight like the fixed stars, in the night-time its light eclipses that of the smaller stars. Its shape, the strength of its light, and the time of its appearance vary quite perceptibly from day to day. As long as the human race has existed, man’s attention must have been drawn to the moon. The course of the moon, thanks to the rapid revolution of the planet round the earth, forms a shorter unit, which steps in between day and year. The shorter interval of time defined by it, unlike the too lengthy period of the year, is easily kept in mind and taken in at a glance. This unit has further its peculiar characteristics. In the first place it has nothing to do with the natural phases conditioned by the course of the sun: it is in fact incommensurable with the seasons. In the second place it immediately obtrudes itself into notice as a unit. The time-reckoning according to the moon is in its nature continuous. One moon follows another with a short interruption, to which at first little attention is paid: for compared with the 27–28 days in which the moon can be seen in the sky the 1–2 days in which it is invisible are little noticed. The phases of the moon represent a gradual waxing and waning, a continuous development. The principle of continuous time-reckoning is therefore suggested by the moon, in opposition to the time-indications from natural phases and from the stars.

The observation of the moon is often said to be the oldest form of time-reckoning. This statement involves a certain danger, viz. the overlooking of the fact that the time-indications from natural phases and from the stars—as I hope has been shewn above—are just as primitive and must be just as old. But if by time-reckoning the continuous principle and measure of time are implied the statement is in that sense true. The moon is indeed the first chronometer, and this fact is due to the nature of its concrete appearance, which draws attention to the duration, and not to the point, of time. And this, as always, is the starting-point: practically everywhere the month as a unit of enumeration or a measure is denoted by the same word as the moon. The linguistic distinction between ‘moon’ and ‘month’ only follows at a stage which primitive peoples still living have not yet reached. All peoples know the moon and use it for time-reckoning. Of the S. American Indians, who observe the stars so well, it is stated that the month is everywhere the natural division of time[623].

While the human mind therefore arrives only gradually at the conception of the year, the month is already given by the natural phenomenon. Consequently it is only to be expected that it should be expressly stated that the revolution of the moon determines the greatest measure of time[624], and that we should find peoples who can count reckoning by months and not by years. Thus, for example, it was often said in southern Nigeria: “I sold this canoe to him eight moons ago”[625]. As in the counting of the years a well-known event is used as a starting-point, so it is also with the months. In the New Hebrides they said:—“Two moons have gone since this or that event took place”[626]. But this principle has not prevailed in the counting of the months, since it gives too many months in the course of one human life, and since the months are drawn into another connexion, to which the following chapter is devoted. Only in one case is a reckoning of this nature common, viz. in pregnancy. Examples are superfluous, but I give at least one:—The Samoan woman looks at the moon and expects the beginning of menstruation at a quite definite position of that planet, each woman naturally having a different position of the moon in view. If menstruation does not take place then, she perceives that she is pregnant, and expects her confinement after ten moon-months[627].

No attention is paid at first to the number of days in the month: many primitive peoples cannot even count so far as thirty. A significant passage in a Ho text originating from a native runs:—“The months are reckoned from the moon (the same word is used for both), which stands in the sky. When the moon appears, remains long in the heavens, and then again for a short time is invisible, we say that a month has just gone. We know nothing about the number of days constituting a month. When we see the moon and then it is lost again a month has gone”[628]. A native Basuto says that little regard is paid as to counting the number of days in any month, since the bulky moon itself fills up the deficiency[629]. When men begin to count the days great uncertainty at first prevails: in Buin, for example, the statements vary between 15 and 31 days[630]; the Caffre month is said to have 25 days. Apparently only the time during which the moon is visible is at first counted. So it is said of the Caffres that they count the month from the phases of the moon during its visibility, and that the days of its invisibility are not counted: the moon has gone to sleep[631]. For the Basuto on the other hand only expressions for the two days of the moon’s invisibility are mentioned: the first, ‘the moon has gone into the dark’, the second, ‘the moon is greeted by the apes’, since this animal can see the moon sooner than man[632]. The Ibo-speaking peoples also reckon only 28 days to the month[633], and so do the Dakota[634]. It is only natural that the days of the darkness should soon be included, so that the following month follows directly upon the preceding; many peoples say, like the Banyankole, that the month lasts 29 days: for 28 days the moon is visible, and for one day hidden[635]. As always, therefore, the concrete phenomenon is the starting point. Here, however, not only the varying shape of the moon, not only its phases, are taken into account, but also, as in the case of the sun and the stars, its position in the sky. On the analogy of the rising and setting of the stars the new moon can be described as the evening setting, the full moon as the evening rising or morning setting, and the disappearing of the moon as the morning rising of that planet. A description of this nature, of course without the above scientific terminology, does occur, but in isolated instances. In the above-mentioned Ho text a further passage runs:—“When the moon appears and comes nearer, we say ‘it stands overhead’. After this it stands in the middle (of the sky). When the moon does not rise until after night-fall we say that it ‘stands on the edge (of the sky)’. When it does not rise until very long after night-fall we say ‘it shines unto day-break’. When the moon is once more on the wane, it will not be long before another appears.” Other expressions are:—‘the moon falls upon the forest’, i. e. stands low on the horizon, ‘it sleeps in the open air’, when it is in the sky at day-break[636]. At the south of Lake Nyassa the day of the month is denoted by indicating the position of the moon in the sky at day-break[637]. Of the Seminole of Florida it is reported that the months seem to be divided simply into days, and that the latter are, at least in part, described by reference to the successive positions of the moon in the sky at sunset. When our informant asked a native how long he would remain at his present camp, he answered by pointing to the new moon in the west, and sweeping his hand from west to east to the spot where the moon would be when he should go home. He meant to answer, “About ten days hence”[638].

To indicate the day by the position of the moon in the sky is however exceptional, and it is just as exceptional for descriptions of the day according to the position of the moon to be consistently carried out. The Ewe tribes also have expressions which refer to the shapes of the moon. These different shapes have in general attracted most attention, and serve for time-reckoning. At first the phases of the moon are distinguished only roughly, but greater and greater refinement of observation is ever being attained, until every day of the moon’s revolution is described by a name, and the names not only refer to the phases of the moon but also indicate its position in the sky.

Among the different phases of the moon’s light two stand out with especial prominence—the first appearance of the crescent of the new moon in the evening twilight, and the full moon. Both events are joyfully greeted and celebrated among many peoples, in particular the appearance of the new moon, the full moon also, but not so often. The explanation of this fact must partly lie in the circumstance that the full moon does not suddenly appear like the new moon, but fills its disc gradually, so that the days of full moon are more numerous, instead of being one exactly determined day like the day of the new moon. Hence there may be a counting of the months in new moons instead of a continuous reckoning in moons, as when the natives of the Solomon Islands count the months which must elapse before the funeral feast by making a notch in a stick or a knot in a rope at the appearance of the new moon[639].

The hailing of the new moon with joy is wide-spread[640]. The Dieri of Australia relate that there was once no moon, so that the old men held a council and a Mura-mura gave them the moon; in order that they might know when to hold their ceremonies, he gave them a new moon at certain intervals[641]. Heathen Eskimos in West Greenland celebrate at every new moon a feast with a performance of the sorceror, an extinguishing of lamps, and the barter of women[642]. The Patagonians welcome the new moon by patting their heads and murmuring an incantation[643]. Certain tribes of North America at the eagerly expected appearance of the new moon uttered loud cries and stretched out their hands towards it[644]. The Natchez of Louisiana at every new moon celebrated a feast which took its name from the principal fruits reaped in the preceding moon, or from the animals that were usually hunted then[645]. In the villages of Port Moresby (British New Guinea) the people at the first sight of the new moon give a prolonged somewhat shrill cry which is taken up by all and repeated in chorus: there is no mention of any time-reckoning[646]. On the southern side of Dutch New Guinea we learn that the first sight of the new moon was signalised by a short sharp bark rather than a shout. Several times on the day following the first sight of the new moon our authority noticed that a spear decorated with white feathers was exposed in a conspicuous place in the village. The author states that he is unable to say whether this custom had any connection with the calendar[647]. In Buin at the appearance of the quarter (sic!) of the new moon the people immediately utter the ‘war-cry’, ‘so that the new moon may not break the cocoa-nuts’. When the new moon comes up, the people of Buin trill with their under-lip, plucking at it with the forefinger and at the same time sending out a high note (‘a’). In Lambutjo the people howl and strike themselves on the mouth with their hands, at the same time uttering ‘a’, so that a kind of quacking is heard. On the Gazelle Peninsula the natives put their forefingers in their mouths and trill a high ‘u’, the result being a gurgling noise[648].

The same custom recurs in Africa. When the Bushmen catch sight of the new moon they pray:—“Young Moon! Hail, Young Moon, hail, hail, Young Moon! Young Moon, speak to me, hail, hail, Young Moon! Tell me of something! Hail, hail! When the sun rises, Thou must speak to me, that I may eat something. Thou must speak to me about a little thing, that I may eat. Hail, hail, Young Moon!”[649]. The Bechuana watch most eagerly for the first glimpse of the new moon, and when they perceive the faint outline after the sun has set deep in the west, they utter a loud shout of kua! and vociferate prayers to it, e. g. “Let our journey with the white man be prosperous!”[650]. The Ba-Ronga always greet the apparition of the new moon with cheers. The first person who sees it shouts kengelekezee (kenge = ‘half-moon shaped’), and this exclamation is repeated from one village to another. According to a Nkuma informant the day of the new moon is shimusi, a day of rest. The appearance of the crescent was carefully examined. If the horns were turned towards the earth, this shewed that there was nothing to fear, the dangers of the month had been poured out. If the opposite was the case, it shewed that the moon was full of weapons and misfortunes[651]. As soon as the new moon is seen, the Banyankole of Uganda come out of their huts and clap their hands. Everyone lights a fire in front of his hut and lets it burn for four days continuously. A number of royal drums are brought out and beaten without cessation for four days[652]. The Wadschagga climb a hill in order to see the crescent properly, and pray at its appearance:—“One, two, three, four (the day of the new moon is reckoned as the fourth day of the month), give me peace, give me food, send me blessing, and drive want far away. O my moon, break him (my enemy) neck and throat!” Since in the evening so many curses are uttered, this day is also termed an evil day. Its peculiarities decide the character of the whole month. For this reason no one should go to rest on this evening hungry or only half-satisfied, or else he will be hungry the whole month long. The master of the house admonishes his wife:—“Day of the moon! Honour the moon, and go in quest of food for the children, that they may not go to sleep hungry every day.” On this day no legal business is done and no debts are paid. But whoever can manage to get his debt paid on that day will have luck and his possessions will increase[653]. This custom is of a highly developed order and exactly resembles the well-known ancient Roman and modern New Year superstition, in which moreover the new moon also plays a prominent part; one can hardly avoid suspecting foreign influence. At Nibo when the new moon comes out they salute it with:—“u-u, don’t let disease catch me, or a bad moon!”; the Ibo celebrate a children’s festival at the time of the new moon[654].

The full moon also gives rise to special feasts: half Africa dances in the light of the nights of full moon. The Bushmen, for example, never neglected the dance at the time of the new and full moon. Dancing began with the new moon and was continued at the full moon[655]. In Dahomey the festivals take place at full moon, the days being fixed by the native government[656]. This is also the case elsewhere. The people of Timor on the night of the full moon dance from night-fall till sunrise: the dancing songs are principally of an erotic character[657]. On the Nicobars at new and full moon feasts were celebrated in which great quantities of an intoxicating beverage prepared from the juice of the cocoa-palm were drunk[658]. The Celtic Iberians of ancient Spain assembled outside their gates on the nights of full moon and celebrated a feast and danced in honour of an unknown god[659]. Who can help thinking here of the well-known words of Tacitus about the Germans?—“Their meetings are, except in case of chance emergencies, on fixed days, either at new moon or full moon: such seasons they believe to be the most auspicious for beginning business”[660]. A fact is here mentioned to which we shall recur below, viz. that the feasts and religious festivals are often celebrated during the time of full moon. This is due not only to the full light of the moon but also to the world-wide idea that everything which is to prosper belongs to the time of the waxing moon, and above all to the days when it has reached its complete phase[661].

New moon and full moon, therefore, by the religious significance attached to them, prove themselves to have been the two phases which were first observed. It is certainly no mere accident that in a word-list of an Australian tribe, the Kakadu of North Territory, only terms for new moon and full moon exist (malpa nigeri and mirrawarra malpa respectively)[662]. Starting from these two phases, the whole period of the moon can be divided into two halves, formed by the waxing and the waning moon. The phases are the same in both halves, but follow one another in the inverse order. Hence they can be described by the same word, with an additional word for the half of the month: but this is only vouched for in one instance, viz. for the Mendalam Kayan of Borneo[663]. On the other hand this division is extremely common, especially among more highly developed peoples, in the counting of the days of the month, to which I return below. Quite primitive peoples cannot count so far as 15, or do so only with difficulty: instead of this they distinguish still further phases of the moon.

In the next place the crescent of the wasting moon is added, so that three phases are given: waxing, culmination, and waning. Thus the Andamanese call the new moon ogur-lo-latika, the full moon ogur-dah, and the waning moon ogur-boi-kal[664]. Another writer gives different names, no doubt for another tribe:—New moon = ‘moon-baby-small’, first quarter = ‘moon-big’, full moon = ‘moon-body’, last quarter = ‘moon-thin’[665]. The literal translation shews however that this author wrongly makes these phases equivalent to our quarters; the full moon and the third quarter are not identical. In reality, besides the full moon, two phases are distinguished during the time of the waxing moon, and only one when the moon is on the wane. The Indians of Pennsylvania distinguish by special names the new, the round (i. e. the full), and the waning moon: the last-named they call the half-round moon[666]. The Negritos of Zambales have periods corresponding to the phases of the moon: the new moon they call bay’-un bu’-an, the full moon da-a’-na bu’-an, the waning moon may-a’-mo-a bu’-an[667]. In Wuwulu and Aua there were words for the full moon, the waxing and the waning moon, and for the time of the moon’s invisibility[668]. This last is not a phase in the proper sense: as soon as it was recognised, however, it was natural that it should be introduced as equivalent to the phases and should thus complete the circle of the month.

In regard to the further development of the phases it is to be noted that this does not as a rule take place with any regularity, but the phases are more specialised during the period of the waxing than in that of the waning moon. The Karaya of Central Brazil were overjoyed to note the first appearance of the crescent. Apparently five phases of the moon are distinguished, for which our authority obtained the following names from an Indian:—First crescent, ahandu loita; not yet quite full moon, ahandu laläli; full moon, djulum läaläli; last crescent, ahandu aluläna; new moon, ikona. Of these ahandu laläli denotes a phase between half and full moon: ‘there are two moons’. Probably the bright and the dark moon are meant. This was confirmed for other Indians, but without its being possible to obtain any accurate account, says our authority. The theory however fits badly, since the earth-light disappears in the second quarter, but is very prominent in the first. The people however were themselves not clear as to the succession of the phases, they gave different orders and often corrected themselves[669].

The Hottentots call the just emerging, hardly yet perceptible crescent by a name which means ‘unripe’ and is also used to denote a premature fruit. The slender shining crescent, in which the moon as it were ‘revives’, is called by a name with that significance. The first two quarters have two names common to both of them, ‘the moon which becomes great or old’, and ‘the moon which becomes wise’. In the last quarter only the slender crescent is distinguished: it is called ‘the dying moon’[670]. In exceptional cases no name for the full moon is given, but we can hardly conclude that such a name was wanting. An Australian tribe of the North Territory calls the full moon igul, the half-moon idadad, and the crescent of the new moon wurdu[671]. The terminology in Central Australia is far richer:—atninja quirka utnamma = new moon, a. q. iwuminta = half-moon, a. urterurtera = three-quarter moon, a. aluquirta = full moon[672]. No terms whatever are given for the waning moon, but that they were entirely lacking is doubtful, though it is also to be doubted whether terms for the half and three-quarter moon cannot also be applied to the waning moon. It should be noted that in Central Australia, as the words shew, the new and the full moon are the original phases.

The observation and naming of the phases of the moon long remain quite unsystematic. The names are mingled with terms arising from other circumstances. Of the Thonga of S. E. Africa it is reported:—When the first quarter appears, the moon is said to thwasa, a Zulu word which corresponds to tjhama in Thonga, and is very much used in the terminology of possessions. Eight days later it is said to basa, to be white or brilliant; full moon is said to sima or lata batjongwana, to put the little children to bed, because when it rises it finds them already sleeping on their mats. The wane is called kushwela dambo, the moon is then found by the rising sun to be still in the sky, not having yet dipped below the horizon. When at last it disappears, it is munyama, the obscurity, the moon is said to fa, to have died[673]. The position of the moon in the sky is also taken into consideration, but not to such an extent as among the Ewe tribes[674]; the latter however are also acquainted with another terminology. Full moon is called ‘the moon fits’, i. e. nothing of it is wanting, new moon ‘the moon is dead’. In the first quarter and at the half-moon they say: ‘the moon is half round’ or ‘falls upon the wood’, i. e. stands low on the horizon; shortly before full moon ‘the moon is about to become complete’, ‘is on the increase’; after the full moon ‘the moon is about to wane’; three days after full moon ‘the moon has cheated some people’, since it leaves in the lurch those who wish to play in the evening; in the last quarter ‘the moon is like the tail of the cock’ or ‘sleeps in the open’, since it stands in the sky at day-break[675]. For the pagan races of the Malay Peninsula words are given for the new moon, the crescent of the moon, the half-moon, the end of the waning moon, no moon[676]. The Bontoc Igorot of Luzon describe three phases between full moon and the waning moon, and three between new moon and full moon, eight altogether therefore, and have special names for them, but rarely make use of them in time-reckoning[677]. The Nabaloi have other words for the same phases, and also one for the moon showing a rim of light[678]. The natives of New Britain (Bismarck Archipelago) observed the phases of the moon (kalang), and had separate terms for them, e. g. ‘moon not visible’, ‘first quarter of the moon (sic!)’, ‘nearly full moon’ (in which they hunted for the land-crabs), full moon, ‘beginning to wane’, ‘moon when seen in the morning’, etc. They also measured time between sunset and moon-rise by the ‘smouldering of a torch’, the time occupied in cooking yams, taro, and wild taro[679]. In Buin the crescent as it becomes visible is first called rubui, ‘the pupil (of the eye) is dead’, since the whole moon is often to be seen as a dark disc when the crescent is first formed. Later they say motoguba, ‘a hook is made’. Still later, nobele, ‘a piece’, ‘a bit’. When the moon’s disc is full, mairen, ‘it is ripe’ or ‘old’, and roukeu, ‘it is equal’, i. e. full. When the moon begins to wane, it is called ingom, ‘puffed out’. The ‘puffing out’ becomes weaker, and now the moon will die, ekio buagi. Throughout the period of the waning moon the expression used is buan-gubio-eiraubi, ‘it is on the point of passing away to die’. During the period of the waxing moon they say (ekio) duabegubi-eiraubi, ‘(the moon) is about to pass away to the sun(light)-making’. During the time of new moon they say mamarabui, ‘the great kobold is dead’, or ekio buaguro, ‘the moon is dead’. When it appears again they say ekio rukui, ‘the moon again makes pupils’, i. e. is in the sky. From the appearance of the moon until the time of new moon they reckon 25 days. The number however is not always the same, but is variously given as 30–31 days or sometimes as only 15. It must be supposed that thick clouds often hinder the observation. The natives count from the rising of the moon[680]. Of the tribes of the Torres Straits we are told:—In Mabuiag the following descriptions of the phases of the moon are used:—dang mulpal, ‘tooth-moon’, since the crescent at its first appearance is described as unmarried: a little later the moon is called kisai, and termed young. The half-moon is ipi laig, ‘married person’; the moon in the third quarter is described as kazi laig, ‘person with child’, and is regarded as having one child, i. e. presumably as being pregnant; the full moon is badi, which is said to mean ‘big one married’. In Mer the crescent of the moon when first observed was called aketi meb, the moon in the first quarter was meb digemli, in the third meb zizimi, almost full eip meb, and full moon giz meb[681].

Among the tribes of Central Brazil (the Bakairi), as also elsewhere, the phases of the moon have found mythological expression. The moon is represented as a shuttle-cock; the phases start from the full moon. First a lizard comes and takes hold of it, on the second day an armadillo, and then a Giant armadillo, whose thick body soon quite covers the yellow feathers[682]. The phases are similarly explained among the Paressi[683].

In regard to the more accurate determination of the days of the moon-month up to the point when each day has its separate name, it is possible to proceed in two ways, either to develop more and more elaborately the concrete descriptions from the phases and positions of the moon, until every day thus takes its name from the shape or the position of the moon, or else simply to number the days. The simple counting and numbering of all the days of the month from the new moon up to 29 or 30 is the most abstract method, and it is only found among the most highly developed peoples. Commonly a mixed system obtains, such, for instance, as that of the Romans, so that within the month, from the starting-points offered by the phases, the days of a certain smaller division are counted, or a short phase is distinguished by means of adjectives in the first, the second, and even the third day of the phase.

The following may serve as an example of a purely concrete system. Among the Mendalam Kayan of Borneo the different days of the period of the moon’s visibility have the following names in the Busang language (the common commercial tongue of the Bukau):—njina (see) dang (pretty well); matau (eye) dang; lekurdang; butit (belly) halab (tetrodon, a trunk-fish) ok (little); butit halab aja (big); keleong (body) paja ok; keleong paja aja; beleling (edge) dija; and kamat (full moon). The days following have the same names, but in the inverse order, and with the addition of uli, i. e. to go home. The days of the moon’s invisibility are not reckoned[684]. The days mentioned amount to only 2 × 8; others must therefore be lacking, or do the names given apply to moon-phases of more than one day’s duration? The author’s wording seems to contradict this. The Batak of Sumatra describe the days by the names of the planets (borrowed from the Sanskrit), repeated four times. To distinguish one from another they make use of additions some of which may probably be referred to original Batak terms[685]. A complete system exists among the Toradja of the Dutch East Indies, in connexion with a fully developed day-superstition such as so often accompanies the moon-month. On certain days, here distinguished by an asterisk, it is forbidden to work in the fields: other work is however permitted. *1, eo mboeja, ‘day of the moon’, from the evening on which the crescent of the moon was first seen. 2 to 9 have no special names: they are called altogether oeajoeeo, ‘the eight days’; the people count ka’isanja oeajoe, ‘the first of the eight’, or oejoeënja, ‘the beginner’, then the second, the third, etc., and so on up to kapoesanja oeajoe, ‘the end of the eight’. 10, woeja mbawoe kodi, ‘the little pig moon’. *11, woeja mbawoe bangke, ‘the great pig moon’; there is a danger that the pigs may break into the fields. *12, taoe koi, 13, taoe bangke, ‘the little’ and ‘the great man moon’; 14, kakoenia, from koeni, ‘yellow’ (among the To Pebato sompe, ‘lying’, i. e. on the horizon). *15, togin enggeri, from gengge, ‘to run to and fro’ (of animals seeking food), i. e. one is annoyed by those who run to and fro. *16, pombarani, ‘the burner’, since the moon in the morning shines on the house-door; or more rarely pombontje. 17 to 20, wani, ‘dark’. 21, merontjo, among the To Pebato wani of kapoesa mbani, the last dark day. *22, kawe, ‘to wink’, 23–25, the second, third, and last kawe. *26, toe’a marate, ‘the long tree-trunk’ (trunk of a felled tree). 27, toe’a rede, ‘the short stump’, in the east ojonja saeo, ‘with a day in between’, i. e. until the vanishing of the moon. 28, polioenja, ‘passing’, i. e. the moon goes past the sun. 29, soea, ‘going inside’, ‘inside’, because the moon is then completely inside. Every second month has 30 days; the *30th is called soea ma’i, the soea ‘on this side’, the second soea. The days are named from the position of the moon at sunrise, since only the agricultural day is of any importance[686].

In Micro- and Polynesia this kind of terminology is best developed. In Samoa the period of the new moon has few names; the new moon is called masina pupula, the nights after this—when a little of the moon is once more visible—mu’a mu’a. On the other hand the days up to and after the full moon have separate names, and are of importance on account of the palolo, which is then eagerly sought after. Full moon, masina ’atoa, ‘full’; 1, night after full moon, masina le’ale’a; 2, masina fe’etelele; 3, masina atatai, the sea sparkles at the rising; 4, fana’ele’ele, according to Stair ‘paling tide’; 5, sulutele, the mali’o-crab is caught with torches (sulu), according to Stair poolesa, night of the lesa; 6, masina mauna, according to Stair popololoa, ‘long nights’; 7, masina mauna; 8 (the first palolo-day), usunoa, ‘wandering about aimlessly’, also called salefu, since foam (lefu) appears as the first sign of the palolo; 9, masina motusaga (second palolo-day), motu ‘fragile’, saga ‘continuing’; 10, tatelego, great palolo-day, which may also begin on the 9th, ta = to fish; 11 (new moon), masina punifaga, ‘only a little covered’; 12, masina tafaleu, ‘little cut away’; 13, masina tafaleu. The crescent shortly before new moon is called masina fa’atoaoina[687].

In Hawaii the system was very elaborately developed. The month had thirty days; 17 of these had compound names (inoa huhui), and 13 had simple names (inoa pakahi). These names were given to the different nights to correspond with the phases of the moon. There were three phases—ano—, marking the moon’s increase and decrease of size, (1) the first appearance of the new moon in the west at evening, (2) the time of full moon when it stood directly overhead (lit. over the island) at midnight, (3) the period when the moon was waning, when it shewed itself in the east late at night. It was with reference to these three phases of the moon that names were given to the nights that made up the month[688]. In former times there is said to have been a division of the month into periods of ten days, corresponding to the increase, the full, and the decline of the moon[689]. The names of the nights were:—1, hilo, ‘to twist’, because the part then seen was a mere thread; 2, hoaka, ‘crescent’; 3, kukahi; 4, kulua; 5, kukolu; 6, kupua; 7, olekukahi; 8, olekulua; 9, olekukolu; 10, olekupau. When the sharp points were lost in the moon’s first quarter, the name of that night was 11, huna, ‘to conceal’; the next, on its becoming gibbous, was 12, mohalu; 13, hua, ‘egg’; and when its roundness was quite obvious, 14, akua, ‘God’. The nights in which the moon was full or nearly so were:—15, hoku; 16, marealaui; 17, kolu. The night in which the moon’s decrease became perceptible was called 18, laaukukahi. As it continued to diminish the nights were called:—19, olaaukulua; 20, laaupau; 21, olekukahi; 22, olekulua; 23, olepau; 24, kaloakukahi; 25, kaloakulua; 26, kaloapau; when the moon was very small, 27, mauli; the night in which it disappeared, 28, muku. This is Dibble’s list (pp. 24 ff.). Fornander (p. 126) counts in the same way up to 26, kaloapau, and then continues, 27, kaue; 28, lono; 29, mauli; 30, muku. Malo gives the same names as Dibble, with the following additions:—The 15th night had two names. If the moon set before daylight it was called hoku palemo, ‘sinking star’, but if, when daylight came, it was still above the horizon, it was called hoku ili, ‘stranded star’. The second of the nights in which the moon did not set until after sunrise (the 16th) was called mahealaui. When the moon’s rising was delayed until after the darkness had set in, it was called 17, kulua, and the second of the nights in which the moon made its appearance after dark was 18, laau-ku-kahi; the moon had now waned so much as again to shew sharp horns. The night when the moon rose at dawn of day was kane (the 27th), and the following night, in which the moon rose only as the day was breaking, lono (the 28th). When the moon delayed its rising until daylight had come, it was called mauli (the 29th), ‘fainting’, and when its rising was so late that it could no longer be seen for the light of the sun, it was called muku (the 30th), ‘cut off’. Thus were accomplished the thirty days and nights of the month. A bare list of the thirty names of the days is given for the Marquesas[690]. Alongside of these a bipartite division of the month is mentioned—the moon arriving, and the moon about to be extinguished[691]. In New Zealand there are various lists of the nights of the moon. The month is also sometimes divided into halves according to the waxing and waning moon[692].

I give the Tahitian names in order to point out that here, as also in Hawaii, some days in the middle of both halves of the month have the same names, which are distinguished from the next following by additions the sense of which is unfortunately not always given. Thus:—1, tirreo; 2, tirrohiddi; 3, o-hatta; 4, ammi-amma; 5, ammi-amma-hoi; 6, orre-orre; 7, orre-orre-hoi; 8, tamatea; 9, huna; 10, orabu; 11, maharru; 12, ohua; 13, mahiddu; 14, ohoddu; 15, marai; 16, oturu; 17, ra-au; 18, ra-au-hoi; 19, ra-au-haddi; 20, ororo-tai; 21, ororo-rotto; 22, ororo-haddi; 23, tarroa-tahai; 24, tarroa-rotto; 25, tarroa-haddi; 26, tane; 27, oro-mua; 28, oro-muri; 29, omuddu (28 and 29 together matte-marama, on the Society Islands they say during these days that the moon is dead)[693]. In the islands just mentioned the names of three successive days are often formed from mua, ‘fore’, roto, ‘in the middle’, and muri, ‘hinder’[694], and in the Carolines names of the days are similarly combined in groups. From these lists it becomes plain how the names of the separate days have been first worked out from the phases of the moon. When only 29 names are given, the thirtieth day occurring only in every other month has evidently been left out. This must be the case, because the month always begins with the new moon. We further possess lists of the days of the month for the Mortlock Islands, and some for the Carolines, Ponape, Yap, Uleai, Lamotrek[695]; the lists for Lamotrek, Uleai, and the Mortlock Islands differ only in the dialect. It is to be noted that in some cases the month falls into smaller subdivisions, as in Ponape, where it begins after the full moon and consists of three periods:—1, rot, ‘darkness’, i. e. nights when there is no moon, 13 days; 2, mach, new moon, 9 days, which are numbered consecutively; 3, pul, the time of full moon, 5 days. Three days are therefore lacking (the time of invisibility?). In Yap 1, pul, new moon, 13 days; 2, botrau, full moon, 9 days; 3, lumor, ‘darkness’, 8 days.

The very fully developed system of the Nandi is curious in that not the phase but the time of the moon’s rising chiefly gives the name of the day. 1, ‘the tanners have seen the moon’; 2, ‘the moon is white’ or ‘new’; 3 and 4, ‘the moon has cast a light’; 5 and 6, ‘the moon has become warm’; 7 and 8, ‘the moon has leisure’; 9 and 10, ‘the herdsmen play in the moonlight’; 11 and 12, ‘the moon is high in the evening’; 13, ‘the moon turns’; 14, ‘the moon has accompanied the goats to the kraal’[696]; 16 (full moon), ‘the moon has passed along (the heavens)’; 17, (morning) ‘the birds have driven away the moon’, (evening) ‘the moon has disappeared for a short while’; 18, ‘the moon has commenced to rise late’; 19 to 21, ‘the moon is late’; 22, ‘the moon has climbed up’ (i. e. stands high in the heavens in the morning); 23 to 25, ‘the moon is late up above’; 26 and 27, ‘the moon has turned’ (i. e. goes towards the west); 28, ‘the moon is nearing death’; 29, ‘the people discuss the moon’ (discuss whether it is dead), or ‘the sun has murdered the moon’; 30, ‘the moon is dead’, or ‘the moon’s darkness’[697].

An example of the naming of smaller groups of days after the phases of the moon is afforded by the old Arabian names for the nights of the month[698]. The nights are grouped in threes, and are called:—1–3, ghurar, ‘the bright ones’; 4–6, nufal, ‘the overlapping nights’ (?); 7–9, tusa’, ‘the nine’; 10–12, ‘ushar, ‘the ten’; 13–15, ‘the white nights’, lit. ‘ajjam al-lajālī l-bidi, ‘the days of the white nights’, the time of full moon; 16–18, dura’, ‘the white nights with black heads’, since the moon does not rise until the night; 19–21, zulam, ‘the dark nights’; 22–24, hanadis or duhm, ‘the very dark nights’; 25–27, da’ādī’, perhaps after mihaq; 28–30, mihaq, from mhq, ‘to extinguish’. The time of the moon’s invisibility, mihaq, consists of the following days:—1, ad-da’dja, ‘the black one’; 2, as-sirār, from srr, ‘to be hidden’; 3, al-falta, ‘sudden event’, ‘attack’. According to some this last name is used only on the night before, according to others after, a holy month. This looks like an attempt to regulate the insertion of the 30th day.

Hitherto we have observed the division of the month into small and the smallest phases of the moon, in which three or at most four days have the same name, and are numbered in order that they may be distinguished. Other peoples count the days beginning at the principal moon-phases. The Central Eskimos can determine the days of the month very accurately from the age of the moon[699], the terms are unfortunately not given. So also for the Kaigan of N. W. America names of the nights reckoned from the phases of the moon are quoted; unfortunately only very confused and inaccurate information could be obtained, and only 14 names are given:—1, new moon; 2, ‘second sleep’, etc., up to 9, full moon or ‘great moon’, the third night after which is ‘the first night after the full moon’[700]. For the inhabitants of southern Formosa the bare and therefore almost useless statement is made that they reckon according to the age of the moon[701]. Of the Wagogo of what was formerly German East Africa we are told that the phases of the moon and the numbers of the nights serve as more accurate determinations of time. For instance, the third night after the next appearance of the moon will be the day following the third night after the moon’s appearance, and therefore the fourth of a month, since the crescent is visible exactly on the first day of a month[702]. Unfortunately we are not told what phases, other than the new moon, serve as starting-points for the reckoning. The same remark applies to an account for Sumatra. The Central Sumatran Expedition has proved that names for days of the week and for months are unknown among the Rawa and the Djambi Kubu of Djipati Mando. The people count by the phases of the moon, and say e. g. the 1st, 2nd, 3rd day of the moon[703].

These accounts are unfortunately of little use, since they say too little about the method of the counting. Even when a complete list of the days or nights of the month does seem to be forthcoming (the Wagogo, the Kubu), it generally happens that the counting proceeds from several starting-points, so that the month is divided up into smaller divisions. This is natural, since primitive peoples not only possess small capacity for counting but also prefer to keep the concrete phenomenon in view. It has already been pointed out that the counting frequently begins at the two most prominent phases, the new and the full moon; by this means the month is divided into the two corresponding halves of the waxing and the waning moon, or in respect of the appearance or non-appearance of the moon in the evening and early night into the light and the dark halves. The difference between these halves follows from direct observation of nature, and they are therefore known even to peoples which do not count the days, e. g. the inhabitants of Buin[704], the Germanic tribes, and others. In Swedish the distinction between ny and nedan, i. e. the time of the waxing and of the waning moon, is still known. The Masai, besides a full list of the days of the month, have a second reckoning according to the light and the dark halves of the month[705]. The Hindus and the civilised peoples of S. E. Asia reckon in the same way: of these systems of time-reckoning the Hindu has exercised a powerful influence. Avesta shews the same reckoning. In the old Gallic calendar of Coligny each month is divided into two sharply distinguished halves. The Romans indeed, in the form of their calendar known to us, reckoned so many days before the Kalends (the first day of the month), the Nones (the 5th or 7th), and the Ides (the 13th or 15th), but before their calendar settled into its curious and quite irrational historic form the Kalendae must have been the day of the new moon, which was publicly proclaimed, and the Idus the day of full moon. The Nonae are secondary: the word simply means the ninth (day), i. e. before the Ides, which position the day occupies in the inclusive reckoning employed. The Greek reckoning in decades is well-known, but in earlier times a bipartite division of the month appears. Homer divides the month into ἱστάμενος and φθίνων (‘rising’ and ‘fading’), Hesiod once mentions a ‘thirteenth day of the rising moon’[706].

We have seen above how to the phases of the new and the full moon that of the waning moon is added as a third. When the gradual development of the moon is regarded—as is done when numbers are used—and not the particular shape of it appearing on a certain day, we also get three periods, since between the waxing and the waning occurs the full moon, and this, although not in the strictest sense, lasts longer than a day, and unlike the waxing and the waning moon remains in the sky the whole night long. The time of full moon therefore appears as a third independent period between the waxing and the waning. The impulse to a tripartite division hereby given clashed with the decimal system of enumeration of most peoples; as a rule the counting was suspended at the basal series of numbers. In this manner we may account for the not uncommon phenomenon that only ten months are numbered, the two others being called by special names[707]. Thus arises the division of the month into three decades, in which however the last decade may vary between 9 and 10 days.

The division into decades is not so common as the halving of the month. The Zuñi of Arizona divide the month into three decades, each of which is called a ‘ten’[708]. The Ahanta of the western Gold Coast divide the moon-month into three periods, two of ten days each, the third—which lasts until the new moon appears—of about 9½ days (more correctly, no doubt, varying between 9 and 10 days). The Sofalese of East Africa must have done the same, since de Faria says that they divided the month into 3 decades and that the first day of the first decade was the feast of the new moon[709]. The Masai, who number either the days of the whole month consecutively or the days of its two halves, nevertheless give special prominence to the initial days of the decades (alongside of other notable days), and call them negera[710].

Among the Greeks the division into decades displaced the older bisection. Of the names of the decades the first and third refer to the concrete form of the moon: μὴν ἱστάμενος, older ἀεξόμενος[711], literally ‘the appearing, waxing moon’, and μὴν φθίνων, ‘the waning moon’. For originally μήν must here have had the sense of ‘moon’ which the etymology suggests. The second decade was called μὴν μεσῶν, ‘the month at the middle’: the epithet shews that μήν here means ‘month’, and not ‘moon’. This name is therefore younger than the two others, which must once have been used to describe the two halves of the month, and do so still in Homer[712].

The custom of reckoning on the fingers or on a notched stick has doubtless lent assistance to the counting of the days of the month. The Wa-Sania make a notch in a stick for every day, and when the month is ended they put this stick aside and begin a new one[713]. At the southern corner of Lake Nyassa the days are counted by means of pieces of wood threaded on a string[714]. A complete enumeration of the days however only exists among highly developed peoples who have discarded a more concrete time-reckoning in favour of an abstract system, just as the civilised peoples of modern Europe abandoned the Roman system of time-reckoning, which was still often used in the Middle Ages (though indeed it had long since departed from its concrete basis), in favour of a simple enumeration of the days of the month.

Finally a couple of curious East African reckonings of the days of the month are to be mentioned, although they are not primitive but have a lengthy development behind them. A common feature of both is that the day of the new moon is already the fourth day, so that the counting of the days begins with the moon’s invisibility, which can hardly have been the original practice. The Wadschagga divide the month into four parts the days of which are numbered, the first and third parts consisting of ten days each, and the second and fourth of five days each. Accordingly they begin to count the new moon at ‘the fourth day, which brings the moon’, the day on which the slender delicate crescent of the moon first reappears after sunset: for the rites of this day see [above, p. 153]. On the fourth day of the second division (the eleventh after new moon) they say that ‘the moon turns to the back of the house’: when twilight falls it is already seen beyond the culmination-point. The fourth day of the third division (the 16th after new moon) is called ‘the day that brings the moon up from below’ (i. e. from the eastern horizon), where ‘it appears like a pot’; the fourth day of the last division is called ‘the four, which dismisses the moon’, and the first of the first division, when the moon vanishes, ‘the one, which floats away the moon so that it is no longer visible’: it ‘tramples into pieces the days of the God’[715]. The natural phases of the moon therefore make themselves felt in spite of the counting. With this, as is so often the case, is connected a fully developed superstition concerning the days of the month. The Masai in ordinary life reckon their moon-months as consisting of 30 days, and number the days from 1 to 30 or 29. Besides this there is a second way of counting which begins at the 16th and reckons the days of darkness (en aimen). Further, special prominence is given to certain days and groups of days, e. g. to the 4th, the new-moon day, hence called also ertaduage duo olaba, ‘the moon is to be seen’, to the 15th, ol gadet, i. e. the rising moon ‘looks over’ to the sun which has not yet set, and to the concluding day, the eng ebor olaba, ‘the brightness of the moon’, but especially to the days of the dark half of the month, en aimen. The 16th is called ol onjori, ‘the greenish day’, the 17th, ol onjugi, ‘the red’, 18 to 20, es sobiaïn, 21 to 23, nigeïn, 27 etc., en aimen nerok, ‘the black darkness’. The people also emphasise the concluding days of the decades[716]. The natural foundation afforded by the phases of the moon therefore appears very clearly: the only noteworthy feature is that the days of the moon’s invisibility are included in the division which is called ‘the brightness of the moon’. An outside influence must no doubt be assumed. Among the Masai also the selection of lucky and unlucky days is common.

The starting-points in the counting of the days of the month also afford evidence for the question as to which phases of the moon are the oldest, and were already utilised for this purpose. Both the methods of counting and the phases themselves are based upon a bisection or trisection of the month: to this were then added other phases, originally quite unsystematically. Among us the quarters of the moon are common; but of their use among primitive peoples I have found only a single instance. Of the Papuans of the Indian Archipelago it is stated that they divide the month into four parts according to the phases of the moon: paik baleo, the new moon, paik jouwar, the first quarter, paik plejif, the waning of the moon, and paik imar, the old moon[717]. It must not, of course, be taken for granted that these phases are of equal length, as ours are.

That the quadripartite division of the month should be practically non-existent among primitive peoples is easily to be understood in view of the considerations already mentioned. Unlike the halving it is not based upon any very clearly distinguishable phases, nor is there in the phases any such suggestion of a quadripartite division as is offered for a tripartite. The shape of the moon on the 8th or the 22nd day differs very little from that of the previous and the following days, and does not constitute a turning-point like the full moon. From the phases of the moon no quadripartite division can arise: the brightest phase of all, the full moon, has an unnatural position in such a division. It can only be understood as a halving of the halves of the month, and this presupposes that the moon’s variation in light is regarded as a unity and divided into parts. The primitive peoples however start not with the abstract unity but with the concrete phases, proceeding at first quite unsystematically, and only subsequently combining them into a system. The quadripartite division therefore is in its very nature a numerical system. That it has penetrated so profoundly into our natures that even ethnological scholars and travellers are not always able to get away from it, is due to the connexion with the seven-day week, which is regarded as a division of the month, and also to the fact that we so seldom take any notice of the concrete phenomena of the heavens.

The quadripartite division must therefore be described as not original (the case is different when the time of the moon’s invisibility is added as a fourth phase to the three already mentioned). To the best of my knowledge it appears first in Babylonia[718], and gains ground together with the sabattu, i. e. the appointing of every seventh day of the month as tabooed: it has become common among us on account of the seven-day week, which was conceived as a division of the month. In reality the tripartite division is also the natural one, since it arises from the concrete phenomenon of the moon, and not from any division of the month into parts consisting of a certain number of days. Here the full moon takes its proper place, which it misses in the quadripartite division. The limitation of the divisions to a definite number of days is secondary throughout.