A Manual of Buddhist Philosophy

BY WILLIAM MONTGOMERY McGOVEKN, PH.D. (Oxox.),
Lecturer in Japanese and Chinese at the School of Oriental Studies,
University of London; Priest of the Nishi Honganji, Kyoto, Japan.
Author of Introduction to Mahdydna Buddhism, Modern Japan,
Colloquial Japanese, etc.

VOL. I
COSMOLOGY

LONDON
KEGAN PAUL, TRENCH, TRUBNER & CO., LTD.
NEW YOrK: E. P. DUTTON & CO.
1923

CONTENTS
Introduction
1. THE STUDY OF BUDDHIST PHILOSOPHY . 1
2. THE DIVISIONS OF BUDDHIST COSMOLOGY . 8
3. THE THREE COSMOLOGICAL SCHOOLS . . 12
4. SOURCES OF REFERENCE . . . .22 PARTICULAR AUTHORITIES SELECTED FOR THE

Part I
1. COSMIC SYNTHESIS 39
2. COSMIC GEOGRAPHY 48
3. THE DIVISIONS OF THE THREE DHATUS . 60
4. THE WORLD OF SENTIENT BEINGS . . 73
Part II : Cosmic Analysis . . . .81
Part III : Cosmic Dynamics
1. THE NATURE AND SCOPE OF CAUSALITY . 163 /
2. THE LAW OF CAUSALITY APPLIED TO THE
A
INDIVIDUAL 169
ADDITIONAL REMARKS ON KARMA . .180
THE LAW OF CAUSALITY APPLIED TO THE
ELEMENTS OF EXISTENCE . . . . 183

A

INTRODUCTION
1. THE STUDY OF BUDDHIST PHILOSOPHY LL forms of Buddhism, however divergent, claim to have but three objects of worship, viz. the Buddha, the Dharma, and the Sangha. The first is the founder of the faith, the second the teaching which he gave, and the third the order which he founded. Regarding each of the Ratnas or jewels, as they are called, an enormous amount of speculation has grown up, with many different opinions concerning the proper method of interpretation.
Questions concerning the Sarigha are largely dealt with in the various versions of the Vinaya Pitaka, or books of Canon law, and their later commentaries. These are con cerned with the proper organization of the monasteries, the rites which should accompany the reception of men and women into the order, and the food, clothing, and furniture which should be used by them afterwards. All these questions lie entirely outside the scope of philosophy, and hence outside the scope of our present undertaking.
Speculations concerning the Buddha, or, rather, the Buddhas, together with less elevated beings such as Pratyeka Buddhas and Arhats, have played an even greater part in Buddhist history, for fierce controversies have been waged concerning the nature and powers of each type of Being, and the means by which such rank might be achieved. But as all such problems are more closely associated with religion than with philosophy, they are equally foreign to our present work.
The dharma or law taught by the Buddha to his disciples and thence transmitted to later generations who gradually modified and reinterpreted the older doctrines, was the basis upon which Buddhist philosophy, properly so called, was later erected. Primitive Buddhism was much more a simple

2 MANUAL OF BUDDHIST PHILOSOPHY
religious and ethical code than a metaphysical attempt to solve the problems of the universe, but as time elapsed and later commentators, delighting in subtlety, strove to further truth and enhance their own reputations by applying the old formulae to a rational explanation of the whole universe, the old Dharma was supplemented by a new Abhidharma. Abhidharma is usually called metaphysic, but the first Abhidharma works were but dogmatic treatises, giving a would-be systematic classification of the older doctrinal categories, gradually fitting in others as these came to be formulated.
As yet there was little formal logic or even rational argument from universally accepted data. The basis of truth was considered to be the body of doctrines laid down by an omniscient teacher, and the most that a commentator could do was to give a new and better arrangement to the old dharma-paryayas, and to bring out the ideas which were considered to be inherent in the older statements. Consequently the old or Canonical Abhidharma is more to be ranked with what we should call theology (save that Buddhism theologized without a thesis) than with philosophy.
Controversy between the various Buddhist sects necessitated the introduction of more abstract reasoning, but appeal could still be made to the sutras, or, at least, to those sutras which were held to be sacred by all the schools, so that the most important developments of Buddhist philosophy were made when the Buddhist speculators came into violent conflict with fully developed alien philosophies. This took place, as far as India is concerned, between the first and the fifth centuries A.D., and in China between the fifth and ninth centuries A.D.
Beginning with Buddhaghosa, the Pali school of Buddhism in Ceylon and later in Burma made many important doctrinal additions to the Buddhist lore, but for the most part the Pali school remained sufficiently sheltered to make it pre dominantly theological rather than philosophical in tone.

INTRODUCTION 3
Those who would study Buddhist philosophy must turn their attention to the Abhidharma works, including in that term not only the original Canonical books of such schools as the Sarvastivadins and Sthaviravadins, but also and more especially the later commentaries, with particular reference to those works in which the Buddhist monks set out to defend their doctrines from outside attack. For the Pali theories we have, of course, to turn to the works of Buddhaghosa and Anuruddha and their followers. For the Hinayana 1 tradition of India proper, we must look to the Sarvastivadin school, including therein the Mahavibhasa, and the works of Vasubandhu and Sanghabhadra. For the Mahay anists of India we have the great representatives of the two principal Mahayana schools, the Madhyamika and the Yogacarin systems, including Nagarjuna, Arya Deva, and Candragomin f or the former, and Asanga, Vasubandhu, Dignaga, Dharmapala, and Dharmakirtti for the latter. The mainsprings of the Chinese Buddhist philosophy are to be found in the writings of the patriarchs of the T ien T ai and Hua Yen schools.
Here, however, we meet with great initial difficulty. The Pali texts have indeed come down to us, and the Pali Text Society has rendered, and is rendering, very valuable service in making them accessible to the Western public by means of new editions in the original tongue, and by means of English translation. But the labour of those who would study the more philosophic works of Sanskrit Buddhism is rendered difficult by the fact that most of the important philosophic texts are no longer extant in their original form, for with the destruction of Buddhism in India, much of the literature of Buddhism likewise perished, and though a number of Sutras, chiefly Mahayana, and Avadanas or pious tales have remained, the more serious works have disappeared. From among the numerous Abhidharma works of the
1 For the definition of Hinayana and Mahayana and for the chief points of difference between the two branches of Buddhism see the introduction to my Introduction to Mahayana Buddhism.

Sarvastivadins, only a single commentary on the Abhidharma Kosa remains. Though we have in the Madhyamika Vrtti a fair exposition of the Madhyamika school, most of the philosophic texts of the Yogacarins have been lost, and what remain are like the Sutra Alamkara, devotional handbooks for the aspirant after Buddhahood, rather than textbooks of metaphysics.
Fortunately a certain number of texts were translated into Tibetan, and as years go by, these will probably be revealed to us. But, alas, the Tibetan texts are by no means complete. The Tibetan people, prior to the introduction of Buddhism, were possessed of practically no culture of their own. They were ignorant even of the art of writing, so it is little wonder that when they adopted Buddhism there was a greater demand for Sutras, Avadanas, Jatakas, and Dharams 1 than for abstruse works dealing with the minutiae of ontology and phenomenology. Kather must we be thankful for what has been preserved.
On many points where the Sanskrit remnants are silent, however, and where even the Tibetan Canon can give no help, we find a full explanation in the Chinese translations of the Buddhist works, which are generally though somewhat incorrectly known as the Chinese Buddhist Tripitaka. Here are to be found all the important works of the Sarvastivadins, Madhyamikas, and Yogacarins, in addition to the original works of the philosophers of the T ien T ai and Hua-yen schools, to whom reference has already been made.
Consequently, he who would make a study of Buddhist philosophy, either of its final systematic form or of the stages of its development, must apply himself to learned treatises on the ultimate nature of the noumenon and phenomena in a language of monosyllables, most of whose characters may be a noun, or a verb, or a preposition, according to the context or the whim of the author or translator.
1 For these and other branches of Buddhist literature see introduction to Burnouf s Bouddhisme Indien.

INTRODUCTION 5
Owing to the linguistic difficulties which beset the path, it is no small wonder that in spite of the enormous strides which Oriental scholarship has made in the last few years, much of Buddhist philosophy remains unknown. We may, indeed, claim to know a good deal of the religious side of Buddhism, particularly of the Pali school, for most of its five Nikayas have found their way into one or other of the European languages. Some five or six well-known Mahay ana sutras have also been translated, while of the Vinaya we have at least the more than adequate rendering of the Pali version by Oldenburg and Rhys Davids. Our knowledge of Buddhist philosophy, however, or even of Buddhist theology is confined to three or four translations from the Pali Abhidharma lore, precious as a foretaste of what is to come, but insufficient to satisfy those who seek for a more substantial philosophical diet. Almost nothing has as yet been done to render the Northern philosophical works accessible to the Western reader, and it may be safely said that even the Pali works cannot be fully understood until their tenets are compared with those of the other streams of Buddhist tradition.
The present work is a not altogether successful attempt to further, in an infinitesimal degree, our knowledge of the history of Buddhist philosophy. At the best we can but nibble at the vast storehouse of material before us, and then scurry away into print before the immensity of the subject shall have dawned upon us to such an extent that we are awed into silence.
But if we are to nibble at all, let us at least attempt to do so systematically, and in such a way as will assist further exploration in the future, gnawing indeed but a tiny hole in a giant structure, but a hole which may serve for re-entry both for ourselves and for others. Consequently, instead of trying to cover the whole field of Buddhist philosophy within the limits which, perforce, must be ours, let us single out some one line of inquiry, and attempt to institute a somewhat detailed survey thereof.
 
A suitable subject is not difficult to find. Buddhism has all along recognized a distinction between relative and transcendental philosophy, between theories concerning the analysis of phenomena and theories concerning the ultimate nature of things. The first covers much of the ground which in Europe has become the special field of the particular sciences, such as astronomy, physics, chemistry, biology, and psychology, while the latter is more in accord with the Occidental delimitations of philosophy, or metaphysics proper. It is here that we find the kernel of Buddhist philosophy, its would-be solutions of problems touching upon the reality or non-reality of external phenomena, its evaluation of the universe from the standpoint of optimism and pessimism, and all the other eternal insoluble riddles which make philosophy so interesting.
Of these two divisions of the subject matter of Buddhist philosophy, our impulse, naturally, is to choose the latter for our special consideration, and this for several reasons. In the first place there is greater chance for treasure trove, for the Buddhists in common with other Orientals had a better chance to vie with Occidental metaphysicians than with Occidental scientists. The measured and weighty dialectic of the Buddhist sage on the ultimate nature of things is sometimes almost as impressive as the polysyllabic proofs of his Western cousin, but Buddhist astronomy and geology is not even quaint enough to prevent its falsity jarring upon us. It is faintly possible that one out of the many contradictory solutions of noumenal problems offered by the Buddhists may be right if for no other reason than that some philosopher among them seems to have given every possible answer to the problem, but even the wiliest of Orientals would find it difficult to allegorize away the improbabilities of Buddhist physics.
In the second place we should prefer to deal with the transcendental side of Buddhist philosophy, because it is here that Buddhism was most original and escaped furthest


from the bonds of tradition, Buddhist or Tirthaka. The reason for this is to be found in the fact that the earlier works gave minute details concerning relative philosophy, and very little about noumenal matters. Consequently, when the great thinkers of Buddhism, such as Vasubandhu, Dignaga, and Dharmapala, arose, they dared not alter the older myths concerning the size or composition of the sun or moon, but were able to formulate highly interesting views concerning the ultimate nature of any and every form of matter and mind.
For these and many other reasons, then, we should prefer to deal with the transcendental side of Buddhist philosophy. Practical necessity, however, keeps us to the other and less interesting aspect of the subject, in this initial survey, if for no other reason than that Buddhist transcendental philosophy is completely unintelligible without a knowledge of the relative philosophy upon which it is based. Modern Western philosophy is far more comprehensible without a knowledge of Western science than is Buddhist metaphysics without a knowledge of Buddhist pseudo-science. A Buddhist can only discuss the reality of the universe in terms of dharmas, paramanus, caitasikas, citta-viprayukta dharmas, dhatus, ay a tanas, and skandhas and, unless we are acquainted with the implications of these terms, none of which have exact Western equivalents, we shall have studied Buddhist philosophy in vain. The atheism, polytheism, pantheism, and monotheism of the various aspects of Buddhism can only be properly understood when one knows of the devas of the Kama dhatu, and of the Brahmas of the Rupa dhatu.
Consequently, in order that we may lay the foundations for a later, more serious study of Buddhist philosophy, it is necessary that we flounder among mythical continents and impossible seas. We must seek to solve the problem hotly disputed between the Sarvastivadins and the Mahasanghikas, as to whether the lictors of Hell are really sentient beings, or are specially created automata. We must wander through long, dull, and badly arranged lists of states of consciousness,


and when all this has been accomplished we may claim indeed, not to have become acquainted with Buddhist philosophy, but to have become capable of discussing its problems.
2. THE DIVISIONS OF BUDDHIST COSMOLOGY
This relative philosophy of Buddhism we have chosen to call Buddhist cosmology, not that the term is particularly appropriate, but because a better is lacking. Thus denned, the subject has three main divisions : (1) Cosmic synthesis ; (2) Cosmic analysis ; (3) Cosmic dynamics.
(1) The first covers the myths concerning cosmography, the nature, size, and shape of the phenomenal universe taken as an existing entity and unanalysed. This covers the field of what in the West would be called astronomy and geology, save that the Buddhist astronomer would aspire to Dante s role, and explain not only the stars, but also the heavens and the various inhabitants thereof.
This is the point on which Buddhism is least original. Most of its myths were not only borrowed from contemporary Indian thought, but were even part of that primitive cosmography which Warren, in his book on the subject (Paradise Found) sought to explore. Nevertheless, once incorporated in the Buddhist system, it had an important part to play in the moulding of later speculation, as when the commentators drew up their list of states of consciousness upon the threefold division of the universe into Kama, Rupa, and Arupa dhatus.
During the course of our present work, it will be seen that it is upon this mythology that all systems of Buddhism are most agreed, which shows that the main structure of Buddhist cosmography was an integral part of Buddhism at a date prior to its division into variant schools. This gives us a very early date. The embellishments of the later commentators (those of Ceylon may be seen in Spence Hardy s Manual of Buddhism) never succeeded in altering the essential structure.
Here, however, we have to distinguish between (a) cosmography proper or the map of the present material world, with its central mountain rocky rings, oceans, and continents, which are in accord with the myths outside of India ; (6) and the development of the Heaven and Hell idea into the various divisions of the three dhatus. This is almost purely Indian in its origin, and as regards the order of the upper layers of the heavens, represents the order of development of thought in India. The fact, for example, that the Devas are placed above the Asuras goes back to the time when the Persians and the Indians were united. The thirty- three gods who inhabit the summit of Mount Meru are a remnant of the theological development of the Vedas. The heavens of Mahabrahma represent the later Brahmana and Upanisad struggle after some higher ideal than the Vedic deities. The four Arupa dhatus are Buddhist attempts to show the inferiority of contemporary ideas concerning the highest bliss as compared with their own Nirvana.
Details concerning the inhabitants of these abodes are largely the work of the ever-ready Buddhist commentators working upon the material already provided. The great age of most of Buddhist cosmology is further shown by the fact that all versions of the Dirgha Agama contains an account of the cycle of creation and destruction. As far as we know, this doctrine of the never-ending cycle of creation and destruction was only evolved by the Aryan mind after the period of the Vedas, and is therefore comparatively modern. It is certainly pre-Buddhist, however, though the Buddhist acceptance of the doctrine did much to further the central idea.
(2) Cosmic Analysis. In addition, however, to accepting contemporary myths regarding the geography of the phenomenal universe, the Buddhists made a great forward step in trying to split up this variegated whole into a number of fundamental units, of which all phenomena are but com pounds or combinations. This striving to reduce the complex to a small number of simple and uniform ultimates played a very important part in Buddhist philosophy. Though some centuries were to elapse before the full and final list of elements could be given, yet the tendency towards cosmic analysis began with the first days of Buddhism, as may be seen from such primitive categories as the five skandhas, twelve ayatanas, and eighteen dhatus. The marked disagreement between the various schools on these factors of existence, apart from these primitive categories, however, shows that cosmic analysis was developed later than cosmic synthesis, and was the result of much careful and reasoned thought. No longer are non-Indian and pre-Brahmanic elements found, and very little indeed seems to have been borrowed from rival systems of thought, except perhaps the atomic theory, so that the Buddhist classification of the elements of existence is an only original even though valueless contribution to human thought. Of even more importance from the philosophical standpoint are the theories which the Buddhists advanced concerning the nature of their own or any other list of the factors of life.

The development of this ultimate-factors-of-life theory was as important to Buddhism as was the discovery of the elements by the modern scientists to the recent philosophers of the West, save that the eighty odd elements of our own science are all material, while the Buddhist elements are mostly mental, since the Buddhists regarded the mind as no less a compound than the body or any inanimate object. Unfortunately, the Buddhists were too busy with religio- ethical considerations in their enumeration of these elements for their lists to be of much value for the modern psychologist or philosopher, but with some justice it may be claimed that their analysis is quite as acute as that of any other school of Oriental philosophy.

(3) Cosmic Dynamics. This refers to the causal forces which bring into combination the factors of existence. The development of this theory was an answer to the question as to how the analysed universe became the synthesized universe, as to why a certain number of elements should not

remain disassociated but combine to form the human personality. According to some thinkers, this combination of the elements to form concrete phenomena might be due to chance, the spontaneous will of the elements themselves, fate or destiny, or the decree of God. To the Buddhists, however, none of these explanations were acceptable, and they strove to show that the formation and dissolution of compounds was due to an endless cycle of fixed causes. In the earlier stages this consideration of causes was largely made from the personal point of view, and had particular reference to the doctrines of Karma and reincarnation taken over by primitive Buddhism from earlier thought. At this point the doctrine of causality was largely of a mythical or religious nature, as may be seen from the numerous Jatakas and Avadanas, stories showing how the performance of a good deed in a past life brings about a happy rebirth in this. Associated with this side of causal theory is the old rune known as the Pratitya Samutpada.
But in addition to this more popular conception of causality even the oldest Sutras show that consideration was given to the philosophical problem of how one thing is caused or conditioned by another, and when the personal aspect of causality came to be overshadowed by the cosmic or universal co-ordination of causes, Buddhism made an important step along the path of philosophic progress. As might be expected, this was one of the later developments of Buddhist thought, and one which was entirely posterior to the time of the Buddha and the early Church ; for whereas all branches of Buddhism agree as to the main points of cosmography (cosmic synthesis), and there are certain categories such as skandhas, ayatanas, and dhatus, which are also common to all forms of Buddhism, however much they may disagree upon other points of cosmic analysis, yet the twenty- four pratyayas or types of causal relationship given in the Pali works are different not only in name, but also in philosophic significance from the types of causes and conditions

enumerated by the North Indian schools. Among all the later philosophers, however, problems concerning the number of causal influences were of such great importance that cosmic dynamics deserves to rank on an equal plane with the other divisions of cosmology.
3. THE THREE COSMOLOGICAL SCHOOLS
We have thus succeeded in laying down the limits and internal divisions of our present undertaking, but it is also necessary to ascertain what aspects of Buddhism are to be included and what are to be excluded from our present survey, for the word " Buddhism " covers every conceivable type of creed, religious and philosophical. Within the ample bosom of the Buddhist order are to be found schools teaching realism, idealism, nihilism, pantheism, and pragmatism, so that before we can begin using the glib formula " Buddhism says ", or " Buddhism teaches ", it is necessary to state explicitly just what branch of that vast congeries of associated faiths we intend to be implied, by the use of the general title.
For certain branches of Buddhist research it would be necessary to survey almost every sect and sub-sect before we could say that our investigation was complete. But, having limited ourselves to matters touching upon cosmology, we are more fortunate, for only three schools of Buddhism have paid sufficient attention to cosmological matters, or made sufficiently original contributions to the subject, to warrant our careful examination.
The three schools in question are : (1) The Ceylonese school, founded on the Pali tradition, which claims to be the direct descendant of the earliest Buddhist school known as the Sthaviravadins or Theravadins. We are not yet in a position to know how close was the filiation between the true Sthaviravadins and the Ceylonese school, but for the present we have given the Singhalese scholastics the benefit of the doubt (a rather serious doubt) and dub all those who follow the Pali Canon genuine Sthaviravadins.
 
(2) The second school is that of the Sarvastivadins, the learned and philosophical school of Hinayana as it developed in India proper, whose canon was probably in some Prakrit originally, but was later transformed into Sanskrit, which henceforth became its canonical language.
(3) The third school is that of the Yogacarins, who were also frequently known as the Vidyamatrins, or Vijnanavadins. This differs from the preceding by being Mahayana rather than Hinayana. Consequently it had an idealistic rather than a realistic interpretation of cosmological problems. Never theless, historically, it was closely filiated with the Sarvastivadin school, and its philosophical and other works were likewise composed in Sanskrit.

The schools which are not of importance for the study of Buddhist cosmology may be summed up in the following way :
(1) Indian Hinayana schools other than those representing the Sthaviravadin and Sarvastivadin tradition. We are told 1 that there were eighteen principal Hinayana sects in India, though no two accounts give quite the same list of names or specific points of difference. Nevertheless, some eight or nine of these were of such considerable influence that remnants of their philosophy have come down to us. From these we may see that practically all of the Hinayana sects were concerned to a greater or less extent with cosmological matters, but that only two of them (the Sthaviravadin and Sarvastivadin) composed an Abhidharma Pitaka or wrote elaborate commentaries on philosophical matters, so that not only are these two the only schools concerning which we have exact information, but they alone seem to have been possessed of a definite and systematic cosmological philosophy as opposed to the general discussion and counter-discussion of the other sects.

(2) Indian Mahayana schools other than that of the Yogacarins. Tradition 2 tells us that there were but two main
1 Cf. Points of Controversy for Pali account, Nanjio s 1284, 1285, 1286 for Chinese account, and RockhilTs Life of the Buddha for Tibetan account.

2 Cf. introduction to Takakusu s translation of I-tsing. philosophic systems of Mahayana in India, one the early Madhyamika school, represented by Nagarjuna and Arya Deva, and the other the Yogacarin school, represented by Asanga and Vasubandhu. For the present purpose the Madhyamika school may be completely excluded, for it had nothing whatever to add to cosmological speculation. It was absorbed in disputes concerning the nature of absolute truth and the ultimate nature of reality. From the absolute standpoint the universe was completely non-existent why, therefore, should one bother to classify the fictitious. From the relative standpoint the Madhyamika philosophers accepted the Sarvastivadin categories without altering them in any respect.

An examination of the various Mahayana Sutras and Sastras rendered into Chinese shows that there were a number of important points of view prevalent among Indian Mahayanists which found presentation in neither of the two above-mentioned schools, but these may also be neglected, first, because they were never developed in a systematic form, and secondly (and even more important), because such works are more concerned with metaphysical or ontological matters than with problems of cosmology.
(3) Chinese and Japanese schools of Buddhism. For the most part, Buddhist communities founded outside of India were content to derive their fundamental philosophic con ceptions from the original Indian schools of thought, though, for this purpose, it may be necessary to count Ceylon, the home of the later Sthaviravadin movement, as part of India. Thus Burma and Siam have accepted their cosmology as well as their ontology from the Pali Abhidharma works.

The Tibetan and Mongolian monks when not executing devil dances, or composing incantations and charms, study the philosophical systems of India, adding thereto very little of their own. Chinese and Japanese Buddhism, however, have developed schools of thought which are quite original and important, and any general history of Buddhist philosophy must or should give considerable attention to these innovations, more particularly the development of the Indian Buddha Ksetra cult into philosophical monotheism, and of the Madhyamika nihilism into realistic pantheism ; but these developments had little reference to cosmology, and the new schools were content to accept the cosmic geography evolved by either the Sarvastivadin or Yogacarin schools. 1

We have seen, then, that a study of the three schools, Sthaviravadin, Sarvastivadin, and Yogacarin, is sufficient to make us familiar with the whole field of Buddhist cosmology. In order, however, that the history of the development of cosmological ideas may be understood, it is necessary to say a word concerning the chronology and inter relation of these three schools.

(1) The Sthaviravadins, 2 or, at least, the school which flourishes in Ceylon, Burma, and Siam, and which lays claim to that title, occupies a curiously isolated position among the various Buddhist schools. Its Sutra and Vinaya Pitakas are essentially the same as those of the other Hinayana schools, as may be seen by comparing the Pali versions with the various Hinayana Sutra and Vinaya works contained in the Chinese Buddhist Canon, but its chief claim to philosophical distinction is its Abhidharma Pitaka, which seems to have been brought from Ceylon shortly after the conversion to Buddhism of that island by Mahinda (Mahendra), the son or nephew of Asoka, in the third century B.C. It is doubtful if the whole of the Pali Pitaka was brought over at this time, or even that the work of composition had been completed at so early a date. In any case, the transmission was oral and subject to interpolation. But the seven Abhidharma works must have existed very much in their
1 Cf. the cosmological sections of the jgc jj| for T ien-t ai and 35. IJC SL for Hua-yen.

2 The general history of Buddhism in Ceylon is too well known to make it necessary to quote authorities ; compare, however, Geiger s translation of Mahavamsa, and the section on Ceylon (vol. iii, p. 11 ff.) in Sir Charles Eliot s new book Hinduism and Buddhism. present form at the beginning of the Christian era, about which time they were reduced to writing. At this period, although the Canon was kept in Pali, a number of com mentaries were written in, or translated into Singhalese, but these have now been lost. Thereafter the most important period of Pali Buddhism, during which this school was almost, though never quite, transformed from a theology into a philosophy, is represented by Buddhaghosa, about A.D. 400. He wrote a number of new commentaries (if history can be trusted), destroyed the older commentaries, and came to be considered the official interpreter of the Pali school. After the time of Buddhaghosa, though from a material point of view the Buddhist order experienced many vicissitudes, its philosophical development was very slow and gradual. As the Ceylonese type of Buddhism eventually dominated Burma and Siam, where it seems to have supplanted other forms of Buddhism, the Buddhism of Buddhaghosa is now the predominant form of Hinayana, and therefore the system of thought best known to the West.

One point, however, deserves attention, and that is the complete absence of all North Indian mention of the Abhidharma books of the Pali Canon. The Pali school makes very sweeping claims for itself. It claims that Pali was the original language of the Buddha, that the seven Abhidharma works are part of his gospel, and that they were recited at the first council. It is also stated that Buddhaghosa, the great commentator, came from somewhere in North India, and was a scholar of some repute before his arrival in Ceylon. Both of these statements imply a close relationship between the Ceylonese Buddhist school and that of India. It is there fore important to point out the following facts : The only Hinayana Abhidharma Pitaka which we can prove to be known to the Buddhists of North India was that of the Sarvastivadins. For a long time it was thought x that these
1 See e.g. Kern s Manual of Indian Buddhism, p. 5, n., and Vidyabhusana s Indian Logic; Mediaeval School, pp. 63-4.

works were but different versions of the Pali Abhidharma Canon, or that if different, the Sarvastivadin works were probably half commentaries or rewritings of the works preserved for us in Ceylon. We now know, however, that there is no connexion between the two sets of works, that the Sarvastivadin writings were composed by persons whom it is scarcely possible to conceive could have seen the PaH works, or even to have heard of their categories. Nor do we find any scholar either inside or outside the Sarvastivadin school who accepted, quoted, or even attacked the Pali Abhidharma works. They were completely ignored as far as we have any record, and though the Sthaviravadins were cited from time to time, there is no place where we can identify their quoted statements in such a way as to prove the possession of a definite Abhidharma Canon. Furthermore, there are several places where the Sautrantikas agree with the Sthaviravadins as opposed to the Sarvastivadins, but in their arguments with the latter they merely say : " We do not accept the Abhidharma Pitaka, but hold only to the Sutras " ; but as in these passages the seven works of the Sarvastivadins are expressly referred to, it is curious that the Sautrantikas do not mention any rival Abhidharma Pitaka, particularly as the existence of such a rival would have been an argument against accepting any Abhidharma Pitaka.
In the same way whenever the Madhyamika philosophers refer to the Hinayana Abhidharma works, the Sarvastivadins are the only ones quoted. In fact, among the Madhyamikas the term Abhidharmika is used as a synonym for Sarvastivadin.

It is difficult to argue from silence, but in any case it can be seen that the Pali Abhidharma can never have been con sidered the fountain-head of wisdom among the North Indian Buddhists. It was probably composed in South India, where Buddhist philosophy developed on lines of its own. 1
1 On the geographical distribution of the Hinayana sects see introduction to Takakusu s translation of the travels of I-tsing (I ching sj/J j^).

Secondly, as regards Buddhaghosa. He can hardly have been a North Indian, because we know that in North India the Sarvastivadin influence was particularly strong, and yet in his Attha Salmi he mentions various opinions concerning the Abhidharma Pitaka, but makes no reference to the Sarvastivadin Abhidharma. Problems concerning the lineage of the Ceylonese school, however, present a special field of research, so that at present we may pause no longer over the matter.

(2) The Sarvdstivddins. Every year the researches of scholars into the recesses of Buddhist history tend to bring out the importance of the part played by the Sarvastivadins. Tradition states that they became a definite sect about three hundred years after the death of the Buddha, but the Sarvastivadins themselves claim that three of their works were composed at the time of the Buddha, one about a hundred years thereafter, and the remaining three about three hundred years after his death. One of these last is the famous Jnana Prasthana, which was regarded as the most important, in fact the keystone of them all, the other six words being known as the six pada or feet.
The headquarters of this sect seem to have been in Gandhara and Kasmira, which were then, even more than now, integral parts of India. Originally the sacred language of the Sarvastivadins must have been some form of Prakrit, and the phonetic rendering of certain terms in the earlier Chinese translations would tend to show that the seven Abhidharma works themselves were composed in this style. Later, however, Sanskrit was adopted, and the influence exerted by this school did much to make this language the canonical tongue of Buddhism, though, of course, the Ceylonese re tained a language (Pali) which though composite, literary, and artificial, was nearer the original colloquial than Sanskrit. 1
1 Incidentally Pali and Samskrta are more correct forms, but the customary spelling has of course been retained.

No doubt the conversion of Kaniska to this school, in the first century A.D., was an influential factor in making the Sarvastivadins the predominant school over the whole of India, from whence it spread to Central Asia, Java, and China, etc. It seems to have been regarded in India as the Hmayana school par excellence, and therefore regarded as the object of refutation on the part of those persons who were not Buddhists, or who, though Buddhists, yet belonged to one of the Mahayana schools. Many of the founders of the other Buddhists schools were in fact originally Sarvastivadins, and bore traces of Sarvastivadin influence to the end of their days, however much they might differ on points of detail.
The culminating point of the Sarvastivadin philosophy appears to have been between the first and fifth centuries A.D. At the beginning of this period a number of commentaries appeared bearing the name of Vibhasa. Probably about the beginning of the second century A.D. all other such works were cast in the shadow by one such Vibhasa known as the Mahavibhasa. This is an encyclopaedic exposition of Buddhist philosophy in the form of a commentary on the Jnana Prasthana. Tradition says that it was compiled by a council of five hundred Arhats convoked by Kaniska. The time and the " Arhatity " of the compilers may be called into question, but one must pay a word of admiration to the stupendous nature of the work itself. It covers every point of philosophy as that word was understood in Ancient India.

The next and final stage in the development of Sarvastivadin philosophy is represented by Vasubandhu and his great opponent Sanghabhadra, whose dates are uncertain, though it is best to assign them to the latter part of the fourth century A.D. 1 Though less synthetic they were more critical and systematic, and they may be said to represent the high-water mark of Sarvastivadin philosophy.
1 The date assigned to Vasubandhu by Takakusu in his article in JRAS., 1905, has several times been called into question. In the present instance the matter is of little importance.

After them the influence of the Sarvastivadins and of all other branches of Hinayana declined, partly due, perhaps, to the conversion of Vasubandhu to the Mahayana fold. In any case no new developments took place, and in due course the Sarvastivadin school entirely disappeared from India along with all other forms of Buddhism. It survives in those Far Eastern countries where Buddhism still prevails, but chiefly as an object of study. In these countries all monks, irrespective of sectarian difference, commence their scholastic career by undertaking a course of Sarvastivadin philosophy, chiefly by means of the Vasubandhu s Abhidharma Kosa.

(3) The Yogacarin Sect. 1 The Yogacarin school presents a very interesting problem to the student of the development of Buddhist philosophy, for not only does it represent the highest point of Mahayana philosophy, but it is also an attempt to synthesize the best elements of both Mahayana and Hinayana ideas. One might be led to suppose that the earlier phases of Mahayana would be the most like Hinayana, but that once having departed from tradition the tendency towards differentiation would become more marked. The facts, however, seem far otherwise, for the Yogacarin school, which was comparatively late, is something of a compromise. Many points on which this assertion is based lie rather within the realm of transcendental than in that of relative philosophy, but the following point deserves attention. The Yogacarins, alone among the Mahayanists, revert to the cosmological position of all the Hinayana sects, and hence, as we have already seen, it is the only Mahayana school to possess a definite cosmological system. Hinayana was largely con cerned with points bearing upon phenomenology, the Madhyamikas and other representatives of early Mahayana idealism were concerned with ontology alone. The Yogacarins, after expressing their opinions on matters ontological, reverted to discussion concerning phenomenology. The most
1 The best summary of the traditional account of Yog. history is to be found in the Pa-tsung, vol. ii, p. 9 ff.

important difference between the Hinayanists, the Madhyamikas, and the Yogacarins is that the first believed in the existence of the external world and its constituent parts, the dharmas ; the second completely denied the existence of the world, and the dharmas ; while the third believed that the world, though an eject of the mind, has yet a relative existence, and that, in fact, the dharma are but stages of the mind s unfolding.
There is some doubt as to the exact date of the foundation of the Yogacarin philosophy, but its first patriarchs, Asanga and Vasubandhu, 1 cannot have lived before A.D. 359 nor after A.D. 450. Both Asanga and Vasubandhu, who were brothers, commenced their careers as Hinayanists, which may partially explain the synthetic nature of the Yogacarin school. Vasubandhu, the younger brother, in fact rose to great distinction among the Hinayanists, his Abhidharma Kosa and other Hinayana works giving him great renown all over the Buddhist world.

His cautious, sane, and philosophical mind probably prejudiced him against the existing forms of Mahayana, which were either wildly superstitious and exaggerated, or else concerned with points of more sophistic nihilism (Madhyamika). Asanga, however, as we may see from his books, though more than a passable philosopher, was possessed of a more religious frame of mind, and consequently went over at an early age to the more devotional and mystic Mahayana, into which, however, he introduced a great number of modifications, and developed the crude idealism found in many of the Sutras composed after the time of Nagarjuna into a definite and systematic school of thought. As yet, however, his system was more theological than philosophical, as may be seen from the introduction of the Maitreya myth.
1 For details concerning the lives of Vasubandhu and Asanga see Takakusu s translation of Paramartha s " Life of Vasubandhu " : T oung Pao, 1904, and also introduction to Levi s translation of Asanga s Sutralankara.

The system evolved by Asanga eventually won over Vasubandhu, who at once turned his greater philosophical powers to the development of the new school of thought, and though he died not many years after his conversion, it was not before he had laid the foundations of a philosophical movement which represented the highest phase of Indian thought, and which for consistency, acuteness, logical reasoning, and rational procedure, compares favourably with any philosophical system not based upon data provided by experimental science.
Many names figure in the galaxy of thinkers produced by the Yogacarins, between the death of Vasubandhu and the downfall of Buddhism in India, but of these Dignaga, Dharmapala, and Dharmakirtti are the most important. Dignaga revolutionized Buddhist logic, and in some ways may be said to be the real founder of Buddhist philosophy as opposed to Buddhist theology, since it was he who first laid down the principle that every doctrine must be proved either by sense-experience or reason without reference to tradition. Dharmapala carried on the work of Dignaga with more especial reference to metaphysics, and his com mentary on Vasubandhu s Vidyamatra Siddhi became, for the Chinese at least, the standard manual of the Yogacarin sect. Dharmakirtti, slightly later, made many notable additions and modifications to the Yogacarin philosophy. Unfortunately, the period of his activity was too late for the Chinese to take much note of him, but the Tibetan Tanjur contains many of his works.

We have already made reference to many of the important writings of each of the three sects in defining their chronological relationship. It would be well, however, to append a more complete list of the works which serve as authoritative sources of information concerning the cosmological (and other) doctrines of each school.
 
1. The Sthaviravddins . Here the scriptural authorities consist of the Pali versions of the Vinaya, Sutra, and Abhidharma Pitakas. Of these the first may, in the present instance, be entirely neglected. The Sutra Pitaka consists, as is well known, of the following five Nikayas :
(1) Dlglia Nikdya, a collection of 34 long dialogues with
no particular order. 1

(2) Majjhima Nikdya, or collection of 150 of middle length,
likewise in no systematic order.
These two are probably older than the next two Nikayas, and embody the older and less formally categorized doctrines.
(3) Samyutta Nikdya, 2 a collection of a large number of
small sutras, in which an attempt has been made to arrange them according to subject, much of the psychological material of early Buddhism, for example, being arranged under the group on Ayatanas.

(4) Anguttara Nikdya, likewise a collection of a great
number of small sutras, arranged according to numerical categories, a favourite method of classification.
These represent the second stage of Buddhist philosophy in which there is no longer a long dialogue leading up to the enunciation of a new truth, but the discussion of the meaning of certain terms arranged in categories. The fifth or Khuddaka Nikaya we can ignore in the present instance. In none of the Pali Nikayas is there a systematic exposition of any aspect of Buddhist cosmology, but many technical terms are employed here and there which serve as the basis for the later developments. As might be expected, the terms relating to cosmic synthesis are nearly all present, as are also many terms relating to cosmic analysis, but almost none relating to cosmic dynamics.
1 Translated into English by Professor and Mrs. R. Davids, Dialogues of the Buddha, 3 vols., quoted hereafter as D.B.
2 The first portion translated by Mrs. K. Davids, Kindred Sayings. Individual English translations of separate sutras and French and German translations, partial or complete, need no mention.
 
The next stage is represented by the seven works of the Pali Abhidharma Pitaka. These are : (1) Dhamma sangani ; (2) Vibhanga; (3) Dhatu Katha ; (4) Puggala-pafmatti ; (5) Kathavatthu; (6) Yamaka ; (7) Patthana. Of these, especial attention should be paid to :
(1) Kathavatthu (Points of Controversy is the title of the English translation, by Mrs. Khys Davids and Aung), as containing an account of the doctrines concerning which the Pali school differed from the other members of the Hinayana community. In certain cases these refer to cosmological matters.
(2) The most systematic treatment of cosmic synthesis is to be found in the Vibhanga (last chapter), though the bulk of the categories in this book, as in the others, refer to cosmic analysis.

(3) For us the most important treatment of cosmic analysis, however, is to be found in the Dhamma Sangani, particularly as we have a translation by Mrs. Rhys Davids. 1
(4) Pali theories concerning cosmic dynamics are to be found in that vast literary jungle known as the Patthana. Though this is regarded as the most important of the seven works, no one in the West has yet fought his way through the thicket. In spite of the enormous amount of repetition which it contains, a translation of this work will throw a great deal of new light on the evolution of Buddhist ideas.
Posterior to the Abhidharma Pitaka proper, but still Indian, or, at least, partially Indian, are such semi-canonical works as the Milinda Panha, Netti, Petakopadesa, etc. Of these the first is well known owing to Professor Rhys Davids translation, known as The Questions of King Milinda.
The next phase of Pali Buddhism, which is in some ways the most important, since it represents the final and complete system of cosmology, is to be found in the works of Buddhaghosa. Out of the vast number of works written by,
1 A Buddhist Manual of Psychological Ethics.


or at least ascribed to, him, we need mention only the following :
(1) Visuddhi Magga, the path of purity, an original work, which is, in fact, a very able encyclopaedia of Buddhist doctrine, corresponding to, but inferior to, the Sarvastivadin Abhidharma Kosa by Vasubandhu. The Visuddhi Magga is the standard authority for Neo-Sthaviravadin philosophy.
(2) Sumangala Vildsirii, a commentary on the Digha Nikaya, with an introductory survey of all Sutra literature.
(3) Attha Sdlim, a commentary on the Dhamma Sangani,
with an introductory survey of all Abhidharma literature. Now translated into English by Ting and Mrs. Rhys Davids (The Expositor).
(4) Panca-ppakarana Atthakathd, a commentary on the last
five books of the Abhidharma Pitaka.

Out of the large number of works produced after the time of Buddhaghosa only one calls for special comment. This is Anuruddha s Abhidhammattha-sangaha, a small summary of Buddhist doctrine which has become the handbook of every monkish student of Southern Buddhism, to a large extent replacing the Visuddhi Magga owing to its greater brevity and conciseness. The Aung-Mrs. Rhys Davids translation of this work * is invaluable not only because of its carefully considered rendering of those terms which are most common in all Buddhist literature, but also because of the introductory essay by Aung, who gives there a most valuable account of the later interpretations of the general field of Sthaviravadin philosophy, which, as we know, was largely concerned with cosmological matters.
2. The Sarvdstivddins. As in the case of the Pali school, we may neglect the Sarvastivadin Vinaya Pitaka, and notice only the other two. We now know quite definitely that the
1 Compendium of Philosophy, quoted hereafter as C.P. or Comp. of Phil.

Sarvastivadins possessed a Sutra Pitaka of their own, which was generally uniform with that of the Sthaviravadins, though differing from it slightly as regards details. Unfortunately, of this we have but fragments, and it would seem that many of the Chinese translations of the four Nikayas represent sects other than the Sarvastivadin. Nevertheless, by tracing quotations made by Vasubandhu, Sanghabhadra, and other Sarvastivadin writers, we see that with one probable exception the present Chinese version was in close agreement with that of the Sarvastivadins. It may, therefore, be of interest to give their names (restored into Sanskrit) :
(1) Dlrgha Agama x (Dlgha N.), which is generally con
sidered to represent the Dharmagupta school.

(2) Madhyama Agama 2 (Majjhima N.), which is likely a
translation of a version belonging to some branch of the Sarvastivadin school.
(3) Samyukta Agama 3 (Samyutta N.). There are three
translations of this work, of which only one is anything like complete. Some give this as belonging to the Sarvastivadins, others as belonging to the Kasyapiyas. In any case the arrangement is sadly at fault.

(4) Ekottara Agama 4 (Anguttara N.). This is undoubtedly
a rendering of some Mahasanghika version, as in some
parts it is almost Mahayana in tone.

There are also a large number of separate translations of individual sutras within the four Agamas. These are of great value for the purposes of higher criticism. It may be added here that the order of the sutras is very different from that of the Pali version, that some sutras are to be found only in the Chinese version, and some only in the Pali, and that some sutras are placed in different Agamas or Nikayas,

1 Nanjio s No. 545 (see note on next page). The word Nanjio followed by a number is the number given in Nanjio s Catalogue of the Chinese Buddhist Tripitaka.
2 N. 542.
3 N. 546, 547, 544, the last alone being complete. * N. 543.

but on the whole there is remarkable similarity between the Northern and Southern version.
One of the very few additions to the Northern Canon is a long sutra appended to the Dirgha Agama, giving a very long and detailed account of the Buddhist cosmorama, or cosmic synthesis. 1 Doctrinally it differs little from the items scattered through the other sutras of the Chinese translation, and is therefore in agreement with the Pali canon, but none of the sutras in the Southern version give at any one place so systematic an account. This tendency towards methodology is always more visible in the North than in the South.
The Abhidharma Pitaka of the Sarvastivadins contains seven works which constitute a definite authoritative canon. Six of these are pada or props of the seventh or Jnana Prasthana, which thus has even a more predominant place among the Sarvastivadin works than has the Patthana in the Sthaviravadin Abhidharma Pitaka. Like the Patthana, moreover, it was probably the last work to be compiled. As regards the Sarvastivadin Canon, there is evidence to show that in the earlier days, prior to the compilation of the Jnana Prasthana, one of the six pada, called the Dharma skandha, was considered the leading Abhidharma work a position which it did not readily lose even after the appearance of the Jnana Prasthana. The seven works of the Sarvastivadin Abhidharma Pitaka, together with their traditional dates and authors, are :

(1) Jnana Prasthana. 2 300 A.B., Katyayamputra.
(2) Dharma-skandha pada. 3 B. s lifetime,
Mahamaudgaly ayana .

1 There are three separate translations of this sutra, N. 549, 550, 551, but I have quoted from 545, since all four versions agree on essentials. This sutra and other portions of the Chinese version of the long collection is quoted as D.A., while D.N. stands for the Pali version, and D.B. for the English translation of the Pah .
2 N. 1273-5, quoted as Jn. Pr. N. 1296, quoted as Dh. Sk.

(3) Sangiti-parydya pdda. 1 B. s lifetime, Sariputra.
(4) Prajnapti pada? B. s lifetime, Katyayana.
(5) Vijndna-Jcdya pdda. 3 100 A.B., Devasarman.
(6) Prakarana pada.* 300 A.B., Vasumitra.
(7) Dhdtu-kdya pada. 5 300 A.B., Vasumitra.

Owing to Takakusu s brief summary of these works in the J.P.T.S., 1905, it is unnecessary to go further into detail concerning their contents.
From the philosophical point of view, however, the seven canonical works are greatly overshadowed by the Mahavibhasa. 6 Here, for the first time, Hmayana Buddhists discussed problems in a rational way, and though the older categories survived, an attempt was made to give them a rational foundation. It will undoubtedly prove the most fruitful source for any future work which attempts to give in detail the history of the development of Indian Hmayana Buddhist thought, since the views of every section of the Sarvastivadin community are represented. Among the number of different interpretations of the earlier theories as found in the Mahavibhasa we find especial emphasis laid upon the opinions of Vasumitra, Ghosa, Bodhideva, and Dharmatrata. Each of these four persons wrote a number of independent books, many of which have come down to us in a Chinese translation. A number of other works were composed during this period. These, however, are of less value.

1 N. 1276, quoted (infrequently) as San-par.
2 N. 1317, not quoted, as the translation is questionable.
3 N. 1281, not quoted, as the book is badly arranged and of little value.
4 N. 1292, 1277, quoted as Pr. pa. The first part is interesting and valuable.
5 N. 1282, quoted as Dh. K., interesting only because of the singularity of its enumeration and definition of the Caitasikas.
6 N. 1264, 1263. The latter is the only complete translation, and the only one cited here. Quoted hereafter as M.V. or Maha. Vibh. For methods of citation see next page.
 
Rather more than a hundred and fifty years after the composition of the Mahavibhasa we come to the time of Vasubandhu and Sanghabhadra. This is usually called the Neo-Vibhasa period. Owing to the great complexity of the Mahavibhasa on the one hand, which makes it of service only as a book of reference, and the lack of completeness on the part of the other works, the documents which must be regarded as the especial manuals for students of the Sarvastivadin philosophy are the three most important works of this Neo- Vibhasa period. These are (1) the Abhidharma Kosa l of Vasubandhu. (There are two Chinese translations, one by Paramartha and the other by Hsiian Tsang. The latter is considered the more authoritative.) (2) Nyayanusara by Sanghabhadra. 2 This is a commentary on the Abhidharma Kosa, and is at the same time a refutation of those points on which Vasubandhu has departed from the orthodox Sarvastivadin tradition (Trans. Hsiian Tsang). (3) Abhidharma Prakarana 3 (the title is questionable). This is a resume of the preceding work, also composed by the same author. In this case Sanghabhadra has omitted much of his polemic, and contented himself with the mere elucidation of the Sarvastivadin philosophy.
3. The Yogdcdrins. The scriptural authorities for the Yogacarin school consist of six Sutras and twelve Sastras. The six sutras constitute a kind of special Sutra pitaka, since, in addition to the six sutras themselves, the Yogacarins vaguely admit the authenticity of all other sutras, Hmayana as well as Mahayana. In the same way the twelve Sastras constitute a sort of special Abhidharma Pitaka.
 
1 N. 1270 (verses only), N. 1269, 1267. Frequently quoted hereafter as A.K., Hsiian Tsang s translation alone being used. In such numbers as (2-7a), 2 refers to the number of the fasc. ( ^), the same in all editions ; 7 refers to the number of the sheet, " a " means right side, " b " left side. The pagination of each edition, of course, differs. In this case, as in that of the M.V., N.A., and A.P., I have given that of the edition in my possession brought back from China.
3 N. 1265, quoted as N.A. 3 N. 1266, quoted as A.P.

As regards the six sutras, we know that the Yogacarin school founded by Asanga was a new and in many ways original contribution to Buddhist thought. Consequently, it was in exact accord with none of the preceding sutras even of the Mahayana school. Nevertheless, it was found necessary to give the new school added prestige and authority by the citation of certain sutras which were already known and accepted. Only a small number of sutras, however, could be used for this purpose. The four Agamas taught many things contrary to the doctrines of the new school. The same was true of the Prajna Paramita Sutras and others of a similar type which expounded the theories of the Madhyamika philosophers. The purely Buddhological sutras, such as the Sukhavati vyuha, were equally inconsistent with the philosophical tendency of the new system. Consequently, recourse could only be had to those sutras which taught the doctrine of transcendental idealism, a doctrine which seems to have developed subsequent to Nagarjuna, and in contrast to the latter s absolute nihilism. As finally constituted the new sutra canon consisted of :

(1) Abhidharma Sutra, and TatJidgata-dvirbhdva-guna-
alamkdra Sutra (titles questionable). These were never translated into Chinese, and seem to have been completely lost.
(2) Avatamsaka Sutra, of which we have two complete
Chinese translations, 1 in addition to renderings of
separate parts.

Like all the other Sutras of the Yogacarin Canon, this cannot be earlier than Nagarjuna, but it must be one of the oldest of the six, as the Tirthakas (heretics) and Hinayanists claim that it was compiled by Nagarjuna himself, and even the orthodox Mahayana tradition has it that he found this sutra in the dragon s cave. As a whole, however, it is probably later than Nagarjuna, as, in addition to the doctrine of
1 N. 87-8.

!unya or Nihilism, which we know was the special doctrine of Nagarjuna, it teaches in a rather vague way the doctrine that the universe is the product of the mind a later develop ment. Apart from Buddhological ideas, however, it is principally concerned with expounding the stages (fifty-two in all) of a Bodhisattva on his path to perfection or Buddhahood. This sutra is the basis of a special school of Chinese Buddhism 1 vaguely related to the Yogacarin school, but with many important differences. The sutra itself nowhere specifically teaches the Yogacarin doctrine, but it is probably posterior to the Saddharma Pundarika, which, in its present form, is itself posterior to Nagarjuna.
4. Lankdvatdra 2 Sutra. This is a highly important sutra from the philosophical standpoint, teaching in an almost systematic way a definitely idealistic system, differing, however, on various important points from the later orthodox Yogacarin system, chiefly because the sutra emphasizes the noumenal aspect of things with a tendency towards monism and transcendentalism, as opposed to the more pluralistic and phenomenalistic idealism of the doctrine of Asanga. Its doctrines are in general accord with the Mahayana Sraddhotpada, of which we have an excellent English translation by Suzuki. 3

5. Ghana Vyuha* This sutra was not translated into Chinese until after the time of Hsiian Tsang, and it is probable that the version which has come down to us has received several new interpolations and additions. In the main, however, its doctrine is in general agreement with that of the preceding sutra, save that the later esotericism is here given greater emphasis.

6. Sandhi Nirmocana Sutra. Of this we have several Chinese translations, 5 of which again that of Hsiian Tsang
1 The Hua-yen school mentioned above. The T ien-t ai school has for its basic sutra the Saddharma Pundarika.
2 N. 175, 176, 177. 3 Awakening of Faith in Mahayana Buddhism.

• N. 444. 5 N. 154, 155, 156, 246, 247.
  

is probably the best. The great number of the translations shows how much the work was appreciated in China. This sutra differs from all the others by being a work which really teaches the Yogacarin philosophy in its orthodox form. For this reason it is considered by the Yogacarins as being the most important of the six sutras. This identity of doctrine early aroused the suspicions even of the uncritical Orientals, and it has been frequently suggested that it was composed by Asanga himself. This view is partially supported by the fact that the sutra in question is embodied as a whole in the latter s Yogdcara-bhumi, and even more because of the fact that the style is much more like that of a Sastra than an ordinary Mahayana Sutra (a fact, we may add, that makes it of infinitely greater literary merit). The twelve Sastras are :
1. Yogacdra-bkumi. 1 This is the gigantic work in 100 fasc., which is supposed to have been dictated by the Bodhisattva Maitreya, and transcribed by Asanga. Buddhists who were not of the Yogacarin school have not scrupled to credit the whole work to Asanga. Just as the Sandhi Nirmocana is the basic sutra of this school, so is the Yogacara-bhumi the basic sastra. Later works are largely commentaries or epitomes of this sastra. In a general way it corresponds in scope to the Sarvastivadin Mahavibhasa, save that it is written from a single standpoint and its doctrines are given authoritatively, as opposed to the assembled arguments of the diverse thinkers of the Hinayana school which constitute the bulk of the Mahavibhasa. Consequently, while no Yogacarin work is lacking in philosophic thought, the Yogacara-bhumi is more theological than philosophical in tone. Translated into Chinese by Hsiian Tsang. The original lost.

2. Sutra-alamkdra. 2 This is a literary epitome of much of the practical side of the Yogacara-bhumi, i.e. the stages on the path of the aspirant, and the actions, thoughts, and studies which should accompany each stage, together with the
1 N. 1170, quoted as Y.B. or Yog. Bhu. 2 N. 1190.

different resultant meritorious qualities associated with the various degrees of sanctity. It has no connexion with the book of the same title by AsVaghosa. 1 We are fortunate in possessing French translations of both Asanga s and Asvaghosa s Sutra Alamkaras, the former by the veteran scholar, Sylvain Levi, and the latter by E. Huber, whose death was a great loss to Buddhist scholarship.

3. Arya-vdcd Prakarana? or the exposition of truth, an early and famous epitome of the metaphysical side of the Yogacara-bhumi. There is almost no doctrinal divergence from the original sastra, and together with the Yogacara- bhumi itself, it is considered to represent the earlier phase of Asanga s thought.
4. Abhidharma Sangiti, z or compendium of philosophy. The verses which serve as the text for each section are ascribed to Asanga, while the prose portion was by Bodhisirnha, who must have been an early disciple of the school. Being thus a joint work, a slight development of doctrine is noticeable, as e.g. in the form of the syllogism. Of great importance as showing the intermediate stage in the Yogacarin philosophical development is the commentary on the Abhidharma Sangiti, known as the Abhidharma Samyukta Sangiti, 4 by Sthitamati, not to be confused with the Sthiramati of the Madhyamika school, whose works are also well known.

5. Mahdydna Samparigraha, or an inclusive treatise on Mahayana, of which the verses are by Asanga and the commentary by Vasubandhu. Of this we have three Chinese
f translations, 5 of which that by Hsiian Tsang is again con sidered the most authoritative. Another translation, however, that of Paramartha, is noteworthy, since a special Chinese jo sect was founded upon it. 6 This highly important treatise 5 differs from the preceding as being a commentary on the Abhidharma Sutra mentioned above, and has no immediate

•  ! N. 1199. 2 N. 1177. * 3 N.82.
  

• * N. 1178, frequently quoted hereafter as Ab. Sam. San. M A
  

5 N. 1183, 1184, 1247. 6 The Jg ^, long since extinct. ho 44 ^
D Jci^ oi

as t* ft.*/

connexion with the Yogacara-bhumi. It is therefore cast on quite different lines.
6. Yoga-vibhdga, ascribed also to Maitreya. Of this we have no Chinese translation. It was probably a summary of the larger Yogacara-bhumi.
7. Dasa-bhumika Sdstra. 1 This is a work by Vasubandhu dealing with the ten stages of a Bodhisattva s career, and incidentally including a discussion of much metaphysical matter. Translated by Bodhiruci. This work, like the Mahayana Samparigraha, is of interest, because at one time there existed a special sect in China, 2 which was based thereon.
8. Madhydnta Vibhdga, the distinction of the mean, the verses ascribed to Maitreya, and the prose to Asanga. This presents the Yogacarin interpretation of the Madhyamika theory of fSunya. There are two Chinese translations. 3

9. Alambana Pratyaya Sdstra. This is a work dealing with the process and cause of sense impressions. For the meaning of the word Alambana Pratyaya (which may be roughly rendered Occasional Cause in the Cartesian sense), see the discussion in that part of the present work termed Cosmic Dynamics. The original work is ascribed to Dignaga, the great Yogacarin logician. Of this there are two Chinese translations. 4 There is also a Chinese translation of a com mentary by Dharmapala, which is even more famous than the original work itself. 5

10. Pramdna Samuccaya. This is a compendium of Buddhist philosophy, with especial reference to logic. Composed by Dignaga, it was translated by Paramartha, but it seems to have been lost.
11. Vidydmdtra Siddhi, or exposition of idealism in twenty verses by Vasubandhu, based on the Lanka vatara Sutra. This is a simpler and more general work than the following. There are three Chinese translations. 6

1 N. 1194. 2 j& f, likewise long ago extinct. 3 N. 1244, 1245,1248. 4 N. 1172-3. 5 N. 1174. 6 N. 1238--40.

I 12. Last and most important is the Vidydmdtra Siddhi in thirty verses, also by Vasubandhu. 1 This is a more systematic exposition of the whole Yogacarin philosophy in thirty mnemonic verses. Its vast influence in the Buddhist world is due to the fact that it was made the text for numerous commentaries composed by that galaxy of intellects that followed Vasubandhu. Perhaps the most notable commentary was that written by Dharmapala. Hsiian Tsang translated this into Chinese and fused with it selections from the commentaries of nine other great scholars, principally Sthitamati and Dignaga. 2 In their work the theology of the earlier treatises is completely transformed into philosophy, and into philosophy of a very high order.

The importance of this compendium was early recognized, and it became the standard manual for all students of the Yogacarin system. It is still, moreover, considered necessary for the priests of all other schools to have read through it. However one may disagree with its doctrines, it is difficult to overpraise its clear, concise, and logical form of exposition, differing so much from the slovenliness characteristic of a great deal of Buddhist thought. Unfortunately, the scope of the present work enables us to present only its weakest point, viz. the cosmology inherited from earlier tradition. But it is of interest to note that consistent with its principle of accepting only those things which could be proved by reason or experience, the myths which constitute cosmic synthesis found little or no place in it. 3

1 N. 1215. 2 N. 1197, frequently quoted hereafter as V.M.S.
3 Three other famous and important works of the Yogacarins, which are not included in the list of twelve Sastras are :
1. Panca-skandhaka-sastra, N. 1176, by Vasubandhu, quoted as Pan. Sk.
2. Panca-skandha-vaipulya-sdstra, N. 1175, a commentary on the preceding by Sthitamati, quoted as Par. Sk. Fat.
3. Sata-dharma-vidya-dvara, N. 1213, by Vasubandhu, quoted as S. Dh. These three are short works containing lists of technical terms, with the orthodox interpretations thereof.

PARTICULAR AUTHORITIES SELECTED FOR THE PRESENT WORK

It has been thought wise to give a more or less complete list of the principal authorities or sources for each of the three schools in question, as it may serve others in conducting more extended research into Buddhist philosophy, more particularly as Nanjio s Catalogue of the Chinese Tripitaka gives no hint as to the order of arrangement, or the chronological philosophical development. Almost each of the above-mentioned works, however, has slight and, in some cases, important points of differentiation, even from other authorities belonging to the same school. Although a study of these variations is of importance for the understanding of lines of doctrinal evolution, in a book of the present size and scope it is impossible to give due emphasis to every phase, and in consequence the principle adopted has been to choose the final phase of each school, and mention briefly the points of difference from the preceding ideas.

The word final needs, perhaps, some explanation. As long as each school continued its corporate existence a certain amount of change or development took place, but within each school a time was reached when the number and order of the categories became fixed. Subsequent development was largely in the nature of reinterpretation of the minor terms and hidden meanings contained in the fixed list of categories. Thus, for example, the early works of each school are by no means in agreement as to the number of the Caitasikas, but after a certain period each school formulated a definite and fixed list (fifty-two for the Sthaviravadins, forty-six for the Sarvastivadins, and fifty- one for the Yogacarins). Thereafter this number was rigidly adhered to.
For our present purposes, therefore, one work from each of the three schools has been chosen which may be regarded as a representative authority, and which at the same time may be said to mark this final phase. This has been used as the

standard of interpretation, though constant reference has been made to the other works mentioned above, and some of their important statements quoted, as may be seen more particularly from the foot-notes appended to each page.
For the Sthaviravadin school the standard taken has been the Abhidhammattha Sangaha, and by way of commentary the notes and introduction to the English translation of this work by S. Z. Aung.

For the Sarvastivadins the standard taken has been the Abhidharma Kosa, together with the criticisms of Sanghabhadra in his Nyayanusara, and Abhidharma Prakarana. Of great value in this connexion have been the classical Chinese commentaries on the Abhidharma Kosa, all of which are to be found in the Su-tsang-ching (Supple mentary Canon).
For the Yogacarin school the standard taken has been the Vidyamatra Siddhi of Dharmapala, 1 etc., together with the classical Chinese commentaries, also to be found in the Supplementary Canon. 2

Whenever reference has been made to these commentaries or to any works other than these chiefly miscellaneous Chinese works mention has been made of them in foot-notes. Two works, however, require special attention. These are two Chinese compendiums dealing with Cosmic Synthesis. Here are gathered together all the important references to cosmographic matters in the various Mahayana and Hinayana Sutras and Sastras in the Chinese Canon. Their names are :
1- ft BL Sfc ft Fo-tsu-t ung-chi, N. 1661. This is a general history of Buddhism, but two sections deal with cosmology.
1 More especially have I used a commentary called Pf| f j| fjj^ | f|| Wei-shih-lun-chiang-i, not itself in the Supplementary Canon, but which embodies the important ideas of all the classical commentaries. Quoted as Wei-shih.

2 On many cosmological points where V.M.S. is silent I have quoted from Ab. Sam. San., mentioned above. The pagination of references to this and other Yog. sastras corresponds to the Kyoto edition of the Tripitaka.

2. & ^ % A Fa-chieh-an-li-t u. Not in Nanjio or the Tripitaka, but consisting almost entirely of quotations from the Canonical works.
The pagination (in notes) follows the edition in the School of Oriental Studies.
Occasionally I have referred to and quoted from A ^ US |g |g j|f| Pa-tsung-kang-yao-chiang-i.
As Sanskrit is the lingua franca of all Buddhist scholars, in most cases I have restored the Chinese terms to their original Sanskrit form. In certain cases this has been done with the aid of the Mahavyutpatti, quoted as M.Vy. This is a mediaeval and anonymous Sanskrit - Tibetan - Chinese technical dictionary. For the history of this work see the introduction to the Kyoto edition of the book itself.

PART I
1. COSMIC SYNTHESIS METHODS OF COMPUTATION

Numerical Expressions.

The detailed examination of Buddhist cosmography must begin with a consideration of the most frequently employed numerical terms. Unfortunately there is very little uniformity in this matter, and each school seems to have its own numerical sequence. With most of these we need not be concerned, but it may be of advantage to cite two of the most common and important.

In the first of these, taken from the Mahavyutpatti (p. 514), we find the following comparatively simple scheme :
10 tens = 1 hundred (Sata).
10 hundreds = 1 thousand (Sahasra).
10 thousands = 1 ayuta.
10 ayutas = 1 laksa.
10 laksas 1 niyuta.
10 niyutas = 1 koti.
10 kotis = 1 arbuda.
10 nyarbudas = 1 padma.
10 padmas = 1 kharva.
10 kharvas = 1 nikharva.
10 nikharvas = 1 mahapadma.
10 mahapadmas = 1 sanku.
10 sankus = 1 samudra.
10 samudras = 1 madhya.
10 madhyas = 1 atta.
10 attas = 1 parardha.
The enumeration found in the Abhidharma Kosa (12-56) is as follows :
10 ones make ten.
10 tens a hundred.
10 hundreds a thousand.

40 MANUAL OF BUDDHIST PHILOSOPHY

10 thousands
10 prabhedas
10 laksas
10 atilaksas
10 kotis
10 madhyas
10 ayutas
10 mahayutas
10 nayutas
10 mahanayutas
10 prasutas
10 mahaprasutas
10 kamkaras
10 mahakamkaras
10 bimbaras
10 mahabimbaras
10 aksobhyas
10 mahaksobhyas
10 vivahas
10 mahavivahas
10 utsangas
10 mahotsangas
10 vahanas
10 mahavahanas
10 titibhas
10 mahatitibhas
10 hetus
10 mahahetus
10 karabhas
10 mahakarabhas
10 indras
10 mahendras
10 samaptas
10 mahasamaptas
10 gatis
10 mahagatis

make a prabheda.
a laksa.
an atilaksa.
a koti.
a madhya.
an ayuta.
a mahajruta.
a nayuta.
a mahanaynta.
a prasuta.
a mahaprasuta.
a kamkara.
a mahakamkara.
a bimbara.
a mahabimbara.
an aksobhya.
a mahaksobhya.
a vivaha.
a mahavivaha.
a utsanga.
a mahotsanga.
a vahana.
a maha vahana.
a titibha.
a mahatitibha.
a hetu.
a mahahetu.
a karabha.
a mahakarablia.
an indra.
a mahendra.
a samapta.
a mahasamapta.
a gati.
a mahagati.
a nimbaraja.

COSMIC SYNTHESIS 41

10 nimbarajas make a mahanimbaraja.
10 mahanimbarajas a mudra.
10 mudras a mahamudra.
10 mahamudras a bala.
10 balas a mahabala.
10 mahabalas a samjna.
10 samjnas ,, a mahasamjna.
10 mahasamjnas a vibhuta.
10 vibhutas a maha vibhuta.
10 mahavibhutas a balaksa.
10 balaksas a mahabalaksa.
10 mahabalaksas an asamkhya.

Incidentally, asamkhya means " countless ", and accordingly we find that this title is not strictly true, since its exact significance can thus be discovered.
Among the many other well-known numerical sequences are those found in the Lalita Vistara, the Saddharma Pundarika, and the Avatamsaka, 1 but the fact that such schemes were purely fanciful, and were never used for practical purposes, permits us to leave them unnoticed.
The Computation of Size and Distance.

The ultimate or smallest unit in the computation of size and distance was the paramanu or atom. For the calculation of size from this paramanu to the anguli (digit or Buddhist inch) we frequently find the following curious old table 2 :
7 paramanus make 1 anu.
7 anus ,, 1 loharajas or metal dust with 49
paramanus.
7 loharajas ,, 1 abrajas or minute particle of water,
with 343 paramanus.
1 Lai. Vist. (L. s ed.), p. 149, agrees with Ch. Sad. Pun. Burnouf s trans, append, agrees with Ch. Avat. f. 48 of 60 f. ed. and f . 65 of 80 f. ed. (both Ch.). For Ceylon M.B. 6, K. i, 336.
2 A.K. 12-16 ; M.V. 136-16 : for South K. i, 335. M.Vy. 520.

7 abrajas

7 avirajas

gorajas

7 vatayanacchi-
drarajas

make 1 sasarajas or hare or rabbit dust, or minute portion of rabbit s dung, with 2,401 paramanus.
,, 1 avirajas or sheep dust, etc., with 16,807 paramanus.
,, 1 gorajas or cow or ox dust, etc., with 117,649 paramanus.
a vatayanacchidrarajas or the mote in a sunbeam (entering by a hole in the window), with 823,543 paramanus.

7 liksa

7 yuka
7 yava

1 liksa or nit with 5,764,801
paramanus.
1 yuka or louse with 40,353,607
paramanus.
1 yava or barley grain, with
282,175,249 paramanus.
1 anguli or digit, with 1,975,226,743.
A digit is about f in.

Up to this point the units are largely of theoretical interest. The larger figures, used in everyday life, are enumerated in the Abhidharma Kosa 1 as follows : 3 anguli-parva or
= 1 anguli or digit.
= 1 hasta, fore-arm, or cubit.
= 1 dhanu or bow, a fathom or 6 feet.
= 1 krosa, 500 fathoms or 3,000 feet.
= 1 yojana, the Buddhist mile. This, though the orthodox computation, was by no means universal. More particularly the yojana, like the Chinese li, had a very fluctuating value. Sometimes it is spoken of as equivalent to 4,650 feet, others give 4J or 5 miles, the nearest to the Abhidharma Kosa figure ; other authorities give 6J,

finger- joints
24 anguli
4 hasta
5,000 dhanus
8 krosas

12-2a.

7, 7J, 7f , 9, 12J, 16 miles. The use of the term by the Chinese pilgrims was inconsistent and contradictory.
The Smaller Periods of Time.

Discrepancies concerning the computation of time also occur, but are less frequent. The ksana is always the smallest unit. Some say that it is the 90th part of the duration of a thought, or the 4,500th part of a minute, while the Mahavibhasa states that 6,499,099,980 such ksanas comprise one day.
The Abhidharma Kosa 1 gives us the following list : 120 ksanas make a tatksana. 60 tatksanas ,, a lava. 30 lavas ,, a muhurta (48 Euro, minutes).
5 muhurtas ,, a kala (Bud. hour).

6 kalas a day (of 24 Euro, hours). These kalas were grouped as follows 2 : (1) Trisandhya or
three day watches, forenoon or purvahna, noon or Madhyahna, and afternoon or Aparahna. (2) Three night watches, first watch Prathama-yama, midnight watch or Madhyama-yama, and the last watch or Pascima-yama. Incidentally, Hsiian Tsang 3 adds that though this is the official Buddhist calculation, many Buddhists in India adopted an eight-period instead of a six-period day, four periods in the daytime and four in the night.

The Months and Seasons.*

Thirty Vasantas or days and nights make one Masa or month, this being subdivided into two paksa or half months : (a) Sukla paksa, the white period, when the moon is waxing ;

1 12-2a. See also Loka-prajnapti, 5-5ot, for full account.
2 M.Vy. 522 ff.
3 Watters, " On Yuan Chuang," section on Indian time.

• For all such calculations M.Vy. 522, for Ceylon M.B. 23. For the North account of seasons see particularly appendix to I-ching s travels, Takakusu s translation.
  

 
and (6) Krsna paksa, or dark period, when the moon is waning. There are, further, twelve Varsa or months in a year, viz. :
1. Caitra. 7. Asvina.
2. Vaisakha. 8. Karttika.
3. .Tyaistha. 9. Margasirsa.
4. Asadha. 10. Pausa.
5. Sravana. 11. Magha.
6. Bhadrapada. 12. Phalguna.
 
We find that there were several ways of classifying the months into seasons. There was, in fact, a two-fold, a three fold, a four-fold, a five-fold, and a six-fold classification.
1. The Two-fold Classification. This consists of the Ayanas or marches, so called because during the one the sun gradually moved to the north, and during the other it gradually moved to the south. This division will be given more detailed consideration when dealing hereafter with Buddhist astronomy. Each of these seasons contained exactly six months.
2. The Three-fold Classification. Hsiian Tsang informs us that the most common division of the year was into three seasons, winter, spring, and summer, or, better, the cold, the rainy, and the hot seasons.

3. The Four-fold Classification. This was the classification of the months upon which most theoretical emphasis was laid. It corresponds to our own division of the year into spring, summer, autumn, and winter, each season containing three months.

4. The Five-fold Classification. The Vinaya, or Canon Law, treats this classification as the most orthodox. It consists of (1) a winter season of four months, (2) a spring of four months, (3) a rainy season of one month, (4) a last season of only one night and day, and (5) a summer season of five months.
(5) The Six-fold Classification. This division, of which

quite frequent mention is made, consists of six seasons, each one containing two months. Their names are :
1. Vasanta, Spring.
2. Grisma, or later Spring.
3. Varsa, Summer, or the rainy season.
4. iSarat, or Autumn.
5. Hemanta, or Winter.
6. Sisira, or later Winter.
The Yugas, Smaller, Middle, and Larger Kalpas.

In later times Buddhism tended to accept the secular chronology of the various countries which it overran as regards the smaller periods of time, but the larger units continued to be an integral part of Buddhist philosophy. These are :

The Yugas. According to Buddhism the average duration of a man s life, far from being constant, is constantly fluctuating, as is his stature also, since it varies between the average height of one foot and a half and an average life of ten years, until his stature is overwhelming, and his life approximates 80,000 years, and then as gradually dwindles to its original proportions and duration. The periods of increase, or Utkarsa, and decrease, or Apakarsa, are both divided into four sub-periods, known as Yugas, 1 and are called :

1. Krta yuga, the age of perfection, when man lives an asamkhya, and all are blissful.
2. Treta yuga, when the size, duration of life, and happiness of beings is diminished.
3. Dvapara yuga, when all these features are but half of that of the Krta yuga.
4. Kali yuga, or the age of degeneration and quarrelling.

Life begins with a Krta yuga, and then degenerates to a Kali yuga, then starts with a second Kali yuga before going up the scale again. According to some accounts the longest
1 M.Vy. 527, A.K. 12-12o ; for Ceylon M.B. 7.

period of human life is an Asamkhya. Very frequently, however, it is epitomized by the figure 80 or 84,000, the usual Buddhist manner in speaking of large numbers.
Small Kalpa. 1 The whole of the eight yugas taken together forms a small or antara kalpa. This small kalpa, then, is equivalent to the period which it takes for man s life to increase and decrease to the maximum, the rate of such increase being computed at one year in every hundred. It is not, however, necessary for such an increase and decrease to take place in an antara kalpa, for, as we shall see later, there are certain epochs when for a whole antara kalpa there is no sentient existence at all. It is merely a fixed period of time.

Middle Kalpas. These smaller kalpas are further grouped together in sets of twenty, the whole of which period is called a Middle or Asamkhya kalpa. These middle kalpas are, again, of four kinds, which succeed one another in the following order :

1. Vivarta kalpa, or the middle kalpa of formation, during which the world comes into existence, and its first inhabitants are spontaneously and automatically born.
2. Vivarta-siddha kalpa, or the middle kalpa of the con tinued formation, or the prime of the world.
3. Samvarta kalpa, or the middle kalpa of destruction, during which the world gradually degenerates, and is finally destroyed.
4. Samvarta-siddha kalpa, or the kalpa of the continuance
of destruction, during which the world is non existent and all is void.

A Samvarta-siddha-kalpa is followed by another Vivarta kalpa, and so on eternally. It should be noted that the destruction of the world does not mean the destruction of the universe, since, as we shall later have occasion to observe, Buddhism postulates the existence of an infinite number of inhabited worlds.
1 For the Kalpas F.T. 30-2a. F.Ch. 26 and 6a. A.K. 12-12 ff.

Great Kalpa. The entire cycle of the four middle kalpas is called a Mahakalpa or great kalpa, which is the largest unit of calculation. Each such great kalpa is the cyclic period of a world, during which time the whole drama of creation and destruction is played. The great kalpas consist of four middle kalpas, and as each middle kalpa consists of twenty small kalpas, a great kalpa contains eighty small kalpas. The following chart may serve as an aid in memorizing this list :

(a) Great Kalpa. (6) Middle Kalpa. (c) Small Kalpa.

Other Chronological Cycles.

In addition to this more or less mathematical computation of time, Buddhism makes frequent mention of various other cycles, mostly with a more religious significance. Of cosmological reference are the three Great Calamities, 1 cycles of destruction of the world through fire, water, and wind ; the three Smaller Calamities, 2 cycles of evil which occur whenever the duration of human life reaches its lowest ebb. Again, kalpas are grouped into (a) those in which Buddhas do and (6) do not appear. Again, inside of each kalpa there are only certain epochs 3 when Buddhas, Pratyeka Buddhas, and
1 A.K. 12-16a; N.A. 32-20; D.A. HOa. 8 A.K. 12-146 ; N.A. 32-186. 3 A.K. 12-8a ; N.A. 32-9a.

Cakravartins may be born. Finally, in connexion with the duration of the doctrine of each Buddha there is a cycle of the three-fold law, 1 viz. : (1) the period of the True Law, the Buddha doctrine in its purity, lasting for five hundred years after the death of the founder ; (2) the period of the Image Law, when men lose the spirit and cling to the letter of the law, lasting for a thousand years ; (3) the period of the Decay of the Law, during which the doctrine becomes corrupt and defiled, lasting, according to some, one thousand years, according to others ten thousand.

2. COSMIC GEOGRAPHY

World Systems.

The interaction of the atoms and elements brings about the formation of the universe. In common with other Indian philosophies, Buddhism taught that the universe consists of an infinite number of worlds, all more or less on the same plan as our own, with the same number of mountain ranges, continents, oceans, etc., together with the other features of Buddhist cosmography, including the hells in the interior of, and the heavens above, each world.

The number of these worlds is really incalculable. Their number runs into infinitude, and they are scattered through all the six directions of space. 2 These worlds are grouped in various kinds of chiliocosms. A small chiliocosm consists of a thousand cakravalas or worlds, and is encompassed by a gigantic wall. A middle chiliocosm consists of a thousand small chiliocosms, with a similar wall. A large chiliocosm consists of a thousand middle chiliocosms, and is likewise furnished with an encompassing wall. A small chiliocosm contains 1,000, a middle chiliocosm contains 1,000,000, and a large chiliocosm 1,000,000,000 worlds. 3 The Abhidharma Kosa informs us
1 F.Ch. 3)-4a.
2 M.B. 2. [M.B. = Spence Hardy s Manual of Buddhism.}
3 M.B. 8; A.K. ll-15a; N.A. 31-20a ; F.A. 32-7a ; D.A. 92a ; A.V.P. l-5a.

that the distance between two worlds is 1,203,450 yojanas. Our own chiliocosm is known as the Saha world, which means either the place of suffering or the capital of a chiliocosm, and is that to which iakyamuni limited the revelation of himself.

Classification of the World Systems.

The infinity of space and the countless number of worlds which it contains, being admitted, the next task is to discover how Buddhism sets about classifying them.
Hinayana, with its doctrine of the rarity of the Buddhas, divides the universe into three categories :

(1) Cakravalas which appear to the Buddhas, but which do not receive his Dharma, and in which a Buddha is never born ; (2) Cakravalas receiving the Dharma of the Buddhas, but in which a Buddha is never born ; (3) Cakravalas in which Supreme Buddhas themselves appear. This division is late and applies only to certain branches of Hinayana, chiefly the southern branches. 1 The tendency of Northern Buddhism, especially Mahayana, was to universalize Buddhahood.

Perhaps the most important division of the universe into the three regions of Kama or Desire ; Rupa or Form ; and Arupa or Formlessness. 2 This classification is common to all forms of Buddhism, and must therefore be considered somewhat more in detail.

The Kama dhatus or worlds of desire consist of those realms where the inhabitants still suffer from various passions. Those in the lower regions are still subject to pain ; those in the upper regions enjoy a physical and sensuous pleasure. This realm includes the material worlds, or earths with the various forms of life contained therein, and the six Kama heavens.

The Rupa dhatus are those realms in which the lower forms of desire are exterminated, but in which the inhabitants are
1 In Ceylon (cf. M.B.) it is said that the Cakravalas always go in groups of three, arranged as follows o This doctrine is unknown in the North.
2 A.K. 8-la.

still possessed of figure and shape, and are still susceptible to the finer forms of intellectual desire. This realm consists of sixteen, seventeen, or eighteen Brahma heavens situate above each Kama dhatu.

The Arupa dhatus are those realms wherein even figure and body disappear, and only life remains. These realms are generally enumerated as four, in the first three of which there still exists some form of consciousness, while in the fourth there is neither consciousness nor unconsciousness.
In addition to these, Mahayana postulates an infinite number of Buddhaksetras or Buddha lands in which the highest followers of Buddhism are reborn.

A more detailed examination of these three realms had best be divided in the following way : (1) A description of the cosmic geography of the Buddhists, or a chart of the con figuration of the earth. (2) A table of the various heavens and hells, etc., which are to be found in the bowels of or above the earth. (3) Description of the various beings who are supposed to inhabit these regions.

The Earth s Foundation.

The basis of each world is a material earth, and since each world and therefore each earth is uniform, the general plan of one earth holds good for all others.
All the Sutras and Sastras agree that the earth is in the shape of a cylinder, the visible world being on the flat section on top, while the cylinder proper consists of layers of different kinds of materials. These layers are composed as follows x :

1. At the very bottom of each earth, and thus supporting is the infinite and thus unmeasurable world of Akasa, space or ether, which at the same time is above, around, and in each of the other spheres.
2. Above this there is a layer of air or wind, solid and immovable to such an extent that it cannot be penetrated
1 A.K. ll-la; N.A. 31-la ; M.V. 133-4a ; F.A. 32-15a ; D.A. 92a; Y.B. 27a ; cf. Ceylon accounts M.B. by diamonds. It is 1,600,000 yojanas in depth, but its diameter and circumference cannot be measured.
3. Above this is the layer of water, 800,000 yojanas deep, with a circumference of 3,610,350 yojanas.
4. Above this is a layer of gold, or hard rock, in contra distinction to the soft mould of which the earth proper is composed. It is 320,000 yojanas deep, and has the same circumference as that of water.
5. Finally, there is the layer of earth proper, 80,000 yojanas deep, on top of which are the various mountains, oceans, continents, etc. Its circumference is the same as the preceding two.

The earth is supposed to be created, held together, and generally sustained and supported by the aggregate result of the actions of all sentient beings. The manner of its creation and destruction will be described hereafter.

Mahdmeru, the Central Mountain. 1

In the exact centre of this and all other worlds is situate a huge mountain, which is the navel of the earth, and con cerning which many legends have arisen. Various supernatural beings are on its sides, on its summit, as well as immediately above it.

Its total height above the surface of the central sea which surrounds it is 80,000 yojanas, and below its surface it reaches down to the bottom of the earthly layer (another 80,000 yojanas). Regarding its diameter, there is much difference of opinion. 2 All agree that the greatest diameter is at its base (i.e. sea-level) and at its summit, the smallest portion being between these two parts. The maximum diameter is the same as the height, 80,000 yojanas. On the sides are certain excrescences, which are the abode of certain of the deities, such as the four tiers of heavens ruled over by the
1 A.K. ll-2a ; F.L. 32-86 ; Y.B. 2-7a. For Ceylon M.B. 10 ; K.I. 187.
2 For Ceylon M.B. 10.

Catur Maharajikas (the Four Great Kings), or the guardians of the four directions.
Each of these rulers has his own especial colour, which he imparts to the territory over which he rules, so that each side of the central mountain has a different hue, as have the seas, rocks, atmosphere, etc., in each direction. By reason of the intermingling of these four primal colours in those parts where they overlap, eight shades are enumerated, viz. :
 
1. In the north, Gold.
2. In the north-east, Virgin Gold.
3. In the east, Silver.
4. In the south-east, Pale Blue.
5. In the south, Blue.
6. In the south-west, Purple.
7. In the west, Red.
8. In the north-west, Golden Red.

To the Buddhists, the whole of the known world formed but a part of Jambudvlpa, the great southern continent, which accounts for the fact that to us the sky seems blue, but for the people inhabiting other continents in other directions the sky has a different colour.

The Seven Rocky Circles.

Ranged around Mahameru (at intervals between which are oceans) are seven rocky or mountainous circles, gradually decreasing in altitude and increasing in circumference. As regards altitude, the extent of each mountain under the water is the same (80,000 yojanas). 1 The altitude of each mountain above the water decreases by half, and the diameter of each is the same as the extent of its elevation above water. Thus these circles 2 are :

1. Yugamdhara. It surrounds Meru on all sides, though separated from it by an ocean 80,000 yojanas wide. Being
1 Here and elsewhere A.K. has 80,000, D.A. has 84,000, as the basis of calculation. Other works vary between the two. I follow A.K.
2 M.Vy. 280. No two accounts give the same list, see comp. chart, K.I. 186; alsoM.B. 12; A.K. ll-2a; M.V. 133-136; Y.B. 2-76.

half the total height of Mem, its altitude above the water and its diameter is 40,000 yojanas.
2. Isddham, separated from Yugamdhara by a sea 40,000 yojanas wide, has an altitude above the water and a diameter of 20,000 yojanas.
3. Khadiraka is separated from Isadhara by a sea 20,000 yojanas wide, has an altitude and diameter of 10,000 yojanas.
4. Sudarsana is separated from Khadiraka by an ocean 10,000 yojanas wide, has an altitude and diameter of 5,000 yojanas.
5. AsvaJcarna is separated from Sudarsana by a sea 5,000 yojanas wide, has an altitude and diameter of 2,500 yojanas.
6. Vinataka is separated from Asvakarna by a sea 2,500 yojanas wide, has an altitude and diameter of 1,250 yojanas.

7. Nimimdhara, the last of the rocky circles, separated from Vinataka by a sea 1,250 yojanas wide, has an altitude and diameter of 625 yojanas.

Stretched around this last rocky circle is the great salt ocean, gradually decreasing in depth from 625 yojanas, near Nimimdhara, to where it is but 1 inch deep. This is the circumference of the earth, and here there is another encircling mountain, this time made of iron, called the Cakravalagala, or Cakravada. This is but a half of the altitude of Nimimdhara, and is therefore 312J yojanas high. According to other accounts, there are two such mountains, one separated a short distance from the other by a space reaching down to the circle of water in which are found some of the most dismal of the hells.

The Great Oceans. 1

We have already seen that between each of the rocky circles
1 A.K. 11-36; N.A. 31-3a ; F.T. 32-9a = Y.B. 2-76. = (M.B. 12; K.I. 183 for South).

is an ocean. The salt ocean, or that between Nimimdhara and the outer circle, is the ocean in which are placed all the inhabited continents, and so is the only one known to mankind. All of the first seven oceans are 80,000 yojanas deep. Their diameter has been given above. The depth of the salt ocean has also been stated. Its width is 322,000 yojanas. The circumference of the whole earth or Cakravala, and therefore the outer circumference of the salt ocean, is 3,610,350 yojanas.
Incidentally the figures for the diameter of the earth may be of interest. The Abhidharma Kosa gives the following enumeration. Mem having a diameter of 80,000 yojanas, its radius is 40,000. From the outermost point of Meru to the outermost point of Nimimdliara is 158,750 yojanas (79,375 being land, and an equal amount water). Across the salt ocean it is 322,000 yojanas, or more exactly 322,312 yojanas, so that the total diameter of the earth is 1,042,124 yojanas.

According to some accounts, the interior seas are filled with various kinds of perfumed waters. The Abhidharma Kos*a, however, merely says that their waters possess the eight magic qualities :

(1) purity,
(2) coolness,
(3) sweetness,
(4) softness,
(5) lightness,
(6) fragrance,
(7) while drinking it does not injure the mouth,
(8) when drunk it does not injure the stomach.

Various reasons are given for the saltness of the eighth or great ocean. The following are the most frequent : (1) In the middle of the sea is a great fish, whose pollutions cause the salt taste ; (2) in the ancient days a Rsi used his magic powers to effect the brackishness ; (3) the salt taste is caused by the earth s impurities, which have been washed away into the sea.

The ocean does not overflow in spite of the water constantly being added by the rivers, because there are four jewels at the bottom of the ocean, which absorb all the surplus water. These change the water into various things, which eventually disappear, leaving no residuum. The tides were explained

by the theory that the water at stated periods flows in and out of the palace of the Naga king, situate at the bottom of the sea.

The Four Great Continents. 1

Located in the eighth, or outermost, or salt ocean, and outside Nimimdhara or the seventh rocky circle are the four great continents, one on each side of the world.
1. On the North, 8,000 yojanas in extent, is the ideal continent, Uttarakuru, square in shape, like a chair or the lid of a box having four equal sides, each side 2,000 yojanas long. 2

2. On the East, 7,500 or 7,000 yojanas in extent, is Purva- videha, in the shape of a half-moon. According to the Abhi- dharnia Kos*a, it has three sides of 2,000 yojanas each, and a fourth of 350 yojanas.

3. On the West is Aparagodanlya or Godaniya, in shape round, like the full moon, 7,500 or 7,000 yojanas in circumfer ence, with a diameter of 2,500 yojanas.
4. On the South there is the great continent Jambudvlpa, in shape like a triangle (with the point facing south), having a circumference somewhat over 6,000 yojanas. The Abhidharma Koa states that three of its sides are 2,000 yojanas long, and a fourth only 3J.

Most accounts further add that in the vicinity and on each side of these four continents there are two large islands or sub-continents (i.e. one on each side), making in all twelve large inhabited portions of the world. Connected with the northern continent are the islands Kurava and Kaurava ; with the eastern continent Deha and Videha ; with the western continent Satha and Uttaramantrina ; with the southern continent Camara and Aparacamara. All are inhabited by some species of man, though, according to one account, Camara is inhabited only by raksas or demons.
1 A.K. ll-4a ; F.T. 32-86 ; Y.B. 2-76 ; M.Vy. 217.
2 For Ceylon amplifications M.B. 15, for the North D.A. 94 ff.

In the early stages of Buddhist cosmology, Jambudvlpa consisted only of India and the immediately surrounding countries, the other portions of the world being unknown. Later, when Japan, China, Central Asia, etc., came to be known, they were likewise counted as part of the same continent. Still later, Europe, Africa, and even America were put in the same category.

A great many mythical details were added to the Buddhist descriptions of Jambudvlpa, chiefly with reference to the principal mountains, lakes, rivers, and forests. 1 With these, however, we need not concern ourselves, and may rest content with referring the curious to the principal original authorities.

Buddhist Astronomy. 2

All schools of Buddhism taught that attached to each Cakravala there is a sun and moon, as well as a multitude of stars, all comparatively small bodies which move in their orbit around Mount Meru, causing the division into days, nights, months, years, etc. Both the sun and the moon were supposed to be about 40,000 yojanas above the level of the sea or the same height as Yugamdhara, the first of the rocky circles, so that this circle and Meru itself hid their rays from the continents lying on the other side from which the sun and moon happened for the moment to be. Both sun and moon continually revolve around Meru, the path of their orbit being between Nimimdhara and the outer circle, or, in other words, above the great salt ocean. Consequently, when it is day in the southern continent it is night in the northern continent, sunset in the eastern continent, and sunrise in the western continent.

The Abhidharma Kosa states that the disc of the sun is 51 yojanas in diameter, that of the moon 50. The sun is composed of gold and crystal ; the moon of silver and lapis
1 For Ceylon amplifications M.B. 15 ff. ; for the North D.A. 93 ff. 54 A.K. 11-86; N.A. 31-12a; F.A. 32-16a. Cf. M.B. 20, showing differences of Southern accounts, where sun = 50 yoj. and moon = 40 yoj.

lazuli. According to some accounts, the sun is really square, and only its movement and its distance make it seem round. In the sun is the palace of the sun god, Surya, and in the moon that of the moon god, Candra. Each palace is 16 yojanas high, and 8 yojanas square.

Frequent mention is made of the two-fold path or marches of the sun, the northern and the southern. 1 For the six months of the southern march the sun gradually passes 5J yojanas beyond the southern limit of the southern continent Jambudvipa, and shines directly over the sea, while during the northern march its orbit is such that the sun shines directly upon the continent itself. For this reason the days are colder in winter, since the rays of the sun fall upon the ocean rather than upon the continent itself, while for the inverse reason it is warmer in summer. For the same reasons, when the orbit of the sun gradually passes to the south the days little by little become shorter and the nights longer, while during the period when the sun swings to the north the days are longer and the nights shorter. At the time when the days are longest, out of the thirty muhurtas into which the day and night are divided, the days have eighteen muhurtas and the nights twelve, and vice versa when the orbit is changed. During the middle course the days and nights have fifteen muhurtas each.

The rays of the sun are always the same, neither increasing nor decreasing in heat. The seeming difference in their intensity is due to the change in orbit, the intervention of clouds which act as a screen, to mists that arise from the ground, and to the actions of Rahu the great Asura, who occasionally eclipses the sun by swallowing it.
With regard to the moon, several reasons are advanced to account for its phases. Most important, however, is the fact that when it is near the sun the overpowering light of the latter prevents its own light from being seen, etc. While it

1 According to M.B. 21, the Ceylonese count three paths instead of two marches.

takes the sun six months to change from its northern to its southern path, the moon undergoes its entire cycle of change in one month.
With regard to the stars, we are told that they are innumerable, there being 80,000 of especial note. They are made of a very pure material, and are inhabited by devas belonging to the realm of the Four Great Kings. In ancient days the stars were arranged into a number of different constellations, a certain number being assigned to each country, whereby it and its inhabitants might be protected. Consequently, each quarter has its own protecting stars ; in the east six asterisms, and in the other three quarters seven each. There are, in addition, nine planets, counting the sun and moon as two. Apart from these last two, there are l :

1. Angaraka = Mars.
2. Budha = Mercury.
3. Brhaspati = Venus.
4. Sukra = Venus.
5. Sanaiscara = Saturn.
6 and 7 Rahu and Ketu, the two great Asuras, some times called the ascending and descending mode respectively.

As regards the size of these stars, we find widely differing accounts. According to the Lokaprajnapti, the largest are 18 krosas, the smallest 3 krosas, the average size being 10 or 12 krosas.

The Signs of the Zodiac.

Along with other astronomical ideas, the Buddhists adopted the usual enumeration of the twelve signs of the solar, and the twenty-seven or twenty-eight signs of the lunar zodiac. Strangely enough the latter was much better known. I have found no mention of the solar zodiac in any old Hinayana
1 M.Vy. 225. Cf. M.B. 24.

work. 1 In the Sannipata Sutra (a late Mahayana work) the twelve signs of the solar zodiac are :

1. A red ram = Aries.
2. A white bull = Taurus.
3. A man and a woman = Virgo.
4. A red crab = Cancer.
5. A red lion = Leo.
6. A black virgin = Gemini.
7. A pair of scales = Libra.
8. A black elk = Scorpio.
9. A centaur = Sagittarius.
10. A sea monster = Capricorn.
11. A white man = Aquarius.
12. Two fish = Pisces.

The fact that there must have been some historical connexion between the eastern and western zodiacs is obvious.
The twenty-eight members of the lunar zodiac are also found among the early Hindus, Arabians, and Chinese. Their Sanskrit names are as follows 2 :

1. Krttika. 15. Anuradha.
2. Rohim. 16. Jyestha.
3. Mrgasiras. 17. Mulam.
4. Ardra. 18. Purvasadha.
5. Punarvasu. 19. Uttarasadha.
6. Pusya. 20. Sravana.
7. Aslesa. 21. Abhijit.
8. Magha. 22. gatabhisaj.
9. Purvaphalguni. 23. Dhanistha.
10. Uttaraphalguni. 24. Purvabhadrapada.
11. Hasta. 25. Uttarabhadrapada.
12. Citra. 26. Revatl.
13. Svati. 27. AsvinL
14. Visakha. 28. Bharam.

1 But M.B. 23 gives it for Ceylon.
1 M.Vy. 225, cf. M.B. 24, on the influence of these signs F.T. 32-17a.


3. THE DIVISIONS OF THE THREE DHATUS

The Relation between the Dhdtus and the Earth.
Having thus briefly disposed of the receptacle or material world, we are now free to examine the spacial relation that exists between the earth and the various divisions of the three dhatus. This may be expressed in the following way x :

The Arupa Dhadtu

The four arupa heavens have no spacial abode, and, consequently, have no place in cosmic geography.

The Rupa Dhadtu

The sixteen, seventeen, or eighteen heavens of the Rupa Dhatu are all arranged in tiers of three or more, high above the earth.

The Kama Dhadtu

1. The six Kama heavens. Four are above the earth, and below the Rupa heavens. One is on the summit of Mount Meru. One is half-way down its sides.
2. Mankind inhabit the four continents.
3. The Pretas are sometimes just under the earth, some times on its surface.
4. The Animals are sometimes in the air or on the earth, but mostly in the water.
5. The Hells are for the most part in the bowels of the earth, directly under Jambudvipa.

The prominent part which each of these divisions and sub divisions of the three dhatus has played in Buddhist speculation renders necessary a somewhat more detailed explanation of each.
Kama Dhdtu. The Hells. 2
Buddhism postulates an infinite number of hells, with various degrees of torture according to the amount of evil
1 A.K. 8-la. 2 A.K. 11-56 ; N.A. 31-5a; D.A. 976 ff. = Y.B. 4-14a.

wrought by the person concerned. In addition to certain hells outside the limits of this Cakravala, these Narakas, or places of suffering, may be grouped under four heads, viz. :
1. The 84,000 smaller Lokantarika, or frontier hells, situated upon the face of the earth, such as (a) on the mountains, (6) on the water, (c) in the desert.
2. The dark hells, eight in number, placed on the outskirts of the Cakravala ; also called the vivifying hells, because any being dying in the first of these hells, is reborn in the second, and so on.

3. The cold hells, eight in number, situated either on the outskirts of the earth, or underneath the southern continent.
4. The hot hells, 1 also eight in number. All accounts agree in placing them under Jambudvipa. We find but casual mention of the first two types of hell, so that we may pass them over in silence. The cold hells, also, are of less importance, but are sufficiently frequently enumerated to make their names important. They are 2 :

1. Arbuda, where the cold is so great that the flesh breaks
out in sores.
2. Nirarbuda, where the whole body swells and blisters from
the cold.
3. Atata, where the culprit s lips are so frozen that he can
utter but this one sound.
4. Hahava, where only this sound can be uttered.
5. Huhuva, where no articulate sound is possible, but the
cold wind in the throat imitates this sound.
6. Utpcda, where the cold sores resemble the buds of the
blue lotus (utpala).
7. Padma, where the cold sores become red and inflamed
like the red lotus (padma).
8. Mahdpadma or Pundanka, where the sores resemble
great lotuses (mahapadma) or white lotuses (pundarika).
1 These are the only ones commonly known in the South. Cf. M.B. 27.
2 M.Vy. 326.

According to most accounts, including the Abhidharma Kosa, these hells are underneath the southern continent, near the hot hells, and ranged shaft-like one underneath another, but in such a way that the middle or fourth hell is widest, and the top and bottom hell the narrowest. Other accounts claim that these cold hells are on the outer circum ference of the world, between the two Cakravalas.

The hot hells are as follows :
1. Samjwa, where the victims tear one another s flesh
by means of metal claws, and are then destroyed by fire, but are revived by a cool wind in order to undergo further torture. The reward of evil action as regards the body, tongue, and mind.
2. Kalasutra, the hell of black ropes, so called because the
victims are loaded with fiery chains. The reward of evil actions against one s parents, a monk, or the Buddha, etc.
3. Samghdta, or the hell of collected misery, the reward of
accumulated sins, especially arising from the three fold bonds : ignorance, lust, and anger.
4. Raurava, or the hell of lamentations, from the shrieks
uttered by the victims thrust into fiery iron cauldrons. The reward of murderers and poisoners.
5. Mahdraurava, the hell of great lamentations, where
similar but more intense forms of suffering to the preceding are undergone. The reward of heretics and malefactors.
6. Tdpana, the hell of burning heat, where the victims are
burnt in iron ovens. The reward of those who have burnt living beings.
7. Pratdpana, the hell of extreme heat, where the victims
are thrust in a lake of fire, and pierced with iron spikes when they attempt to escape. The reward of habitual malefactors and apostates.
8. Avid, or the non-intermittent hell, so called because
there is no intermission in the suffering, and because the suffering is unmixed with any form of pleasure. The reward of the most serious offences.
With regard to the location of these Narakas, the com mentaries say that immediately below the earth s surface there is an earthy loam of 500 yojanas depth, then another layer of white clay likewise 500 yojanas deep. Below this is Samjiva, and the six following hells, one under another, occupying in all the space between 1,000 and 19,000 yojanas below the surface of the southern continent. Each of these hells has a diameter of 10,000 yojanas.

Below this is the roof of Avici, which is 20,000 yojanas below the surface, and is itself 20,000 yojanas broad, wide, and deep, so that its floor is 40,000 yojanas from the level of the earth. The Abhidharma Kosa adds that according to some accounts all the hells, instead of being in tiers, are on a level. Furthermore, all the accounts agree that each hell possesses sixteen annexes, which makes the total number of hot hells into 136.
There was some difference of opinion 1 as to whether the fiends who inflict torture on the damned are really sentient beings or soulless automata created byKarmic power. The consensus of opinion was in favour of the latter theory, except for Yama, the king of the dead, before whom the dead appear, are judged, and their just punishment assigned. Some accounts state that the judgment of women is in the hands of his sister. In any case, Yama and his retinue are living beings born there as the result of a vow registered in the past. He has eighteen chief ministers and thousands of retainers. The position of Yama s court is also disputed. The Abhidharma Kosa states it to be 500 yojanas under the earth. Others claim that it is on the circumference of the globe.
The Kama Dhdtu. The Other Divisions.
Above the hells come the realms of (1) the Pretas or ghouls, who, according to the Abhidharma Kosa, have their head-
1 A.K. 11-76.


quarters in the underground palace of Yama, but also inhabit cemeteries, and dark caverns, etc. (2) The animals of all kinds found all over the surface of the earth, and in the air, but most of all, so say the commentaries, in the water. (3) Mankind, who inhabit the four great continents and the eight sub-continents. (4) The Asuras, 1 the inferior deities, frequently represented as opposed to the superior deities, who occasionally appear upon the earth, but who for the most part inhabit the lower regions of Meru, and the seven rocky circles. (5) The Devas who inhabit the upper regions of Meru, and the heavens immediately above its summit.
The Deva heavens are of sufficient importance to merit separate attention.

(a) The Heaven of the Four Great Kings. 2 The inhabitants of the lowest of the six deva heavens, the Catur-maharajika Heaven, inhabit the mansions of the sun and the moon, and the summits of the seven interior rocky circles, but their headquarters are the four tiers or excrescences on the lower half of Mount Meru. The lowest of these tiers is 10,000 yojanas above sea-level, the second 10,000 above the first, the third 10,000 above the second, and the fourth 10,000 yojanas above the third, so that the highest point of this realm is 40,000 yojanas above sea-level, or just half the total height of the mountain. These four tiers protrude from the side of Meru 16,000, 8,000, 4,000, and 2,000 yojanas respectively. In the first tier dwell the Yaksas, known as the Karotapani, or the firm-handed. In the second the Maladharas or the holders of crowns. In the third, the Sadamadas, or the ever-intoxicated (with pride), and in the fourth and highest tier the Four Great Kings themselves and their immediate retainers. The four great kings are the guardians of the four

1 See especially D.A., p. 1036.
2 A.K. ll-10a; N.A. 31-14a; F.T. 32-18a ; Dt., p. 1046; Y.B. For the South, M.B. 24.

quarters, and so one is placed on each of the four sides of Meru. Their names 1 are :
(1) Dhrtardstra, guardian of the east ; (2) Virudhaka, guardian of the south ; (3) Virupdksa, guardian of the west ; (4) Vaisravana, guardian of the north. The retainers of the first are Gandharvas, of the second Kumbhandas, of the third Nagas, and of the fourth Yaksas.

Each of these four kings has ninety-one sons who share in the name, privileges, and duties of their parents, and help to guard the ten regions of space. Further, each king has eight all-powerful generals, who act as agents of the four kings and have charge of the lesser deities of the mountains, rivers, forests, etc., all over the world. The chief of these generals is Vaideha, who acts as the special protector of Bhiksus, and is frequently portrayed at the end of Chinese Sutras as the symbol of guardianship against all evil. The Abhidharma Kosa adds that among all the Deva realms the inhabitants of this heaven are the most numerous.

(b) The Heaven of the thirty-three Gods, 2 or the Trayastrimsa heaven, is the second heaven of the Deva-lokas. All traditions are in agreement that this heaven is placed on the summit of Meru, occupying the whole of the space thereon, each of whose sides is generally stated to be 80,000 yojanas long. At each of the four corners there is a peak 500 yojanas high, where reside the Vajrapani Yaksas, who act as guardians of this heaven. In the middle of the flat summit table of Meru is the royal city of akra, the chief of the Devas. This city is called Sudarsana, the beautiful to see. Each side is 2,500 yojanas long, and its circumscribing wall of gold is 1J yojanas high. The floor of this heaven is of a hundred colours. Nevertheless, it is as soft as cotton, following the foot as it ascends and descends. In the middle of this city is the palace of Sakra, called Vaijayanta. The length of each of its sides is 250 yojanas. On each of the four sides of the
1 M.Vy. 223.
a A.K. 11-106 ; F.T. 32-20a ; D.A., p. 105a ; Y.B. 4-146.

jewelled city is a pleasure park with a magic lake. The names of these frequently described parks are : (1) Caitraratha, on the east ; (2) Purusyaka, on the south ; (3) Misrakavana, on the west ; and (4) Nandana, on the north. On the north east of the city there is a magic tree 100 yojanas high, emitting a beautiful perfume ; on the south-west is the meeting hall of the gods of this realm where they gather together to discuss the law.
(c) The other four DevaloJcas. 1 The remaining heavens of the Kama dhatu are :

1. Ydma, 80,000 yojanas above the second heaven and 160,000 yojanas above sea-level. Here there is no division of day and night, it being perpetual day-time.
2. Tusita, 160,000 yojanas above the Yama heaven, and 320,000 above sea-level, where reside the heavenly illuminating deities who shed light upon all the world. This is a very popular heaven among the Buddhists, for here went Mahamaya, the mother of Gautama, on her death, and here reside the Bodhisattvas before their final incarnation on earth as Buddhas. For this reason Maitreya, the next Buddha, resides at present in this heaven.

3. Nirmdnarati, 320,000 yojanas above the Tusita heaven, and 640,000 yojanas above sea-level. The name means the heaven of transforming pleasures, so called because subjective desires are at once transformed into objective pleasures, thoughts as well as wishes being creative forces.
4. Paranirmitavasavarti, 640,000 yojanas above the preceding heaven, and 1,280,000 yojanas above sea-level. It is the heaven of the freedom of transformations, and the highest of the Kama heavens. Here it is that a single look may generate new Karma. Strangely enough, either in this heaven or immediately above it is the abode of Mara, the Buddhist Satan, who is the king of lust and cupidity, and is therefore the ruler of the Kama dhatu, or the realm of desire.
1 M.Vy. 219; A.K. 11-126; N.A. 31-16a ; D.A. 926; M.B. 25; C.P. 139 ; Y.B. 4-14a.

The Abhidharma Kosa states that there was a difference of opinion regarding the dimensions of these heavenly mansions. According to one opinion, Yama and the others are four times the size of the Trayastrimsa heaven, while others state that each heaven is twice the size of the one immediately below it.
Rupa Dhdtu.

Above the heavens of the Kama dhatu, where both cupidity and form remain, are the heavens of the Rupa dhatu, frequently called the Brahma heavens, where cupidity no longer has a place, but where the inhabitants still have bodies, and so shape or form (rupa) . These heavens are variously enumerated as sixteen, seventeen, and eighteen. In the Abhidharma Kosa, 1 which gives the number as seventeen, the heavens are enumerated as follows :

1. Brahmakdyika, the heaven of Brahma s retainers.
2. Brahmapurohita, the heaven of Brahma s ministers.
3. MahabraJima, the heaven of Brahma himself.
4. Panttabha, the heaven of lesser light.
5. Apramdndbha, the heaven of infinite light.
6. Abhdsvara, the heaven of universal light.
7. Panttasubha, the heaven of lesser purity.
8. Apramdnasubha, the heaven of infinite purity.
9. tSubhakrtsna, the heaven of universal purity.
10. Anabhraka, the cloudless heaven.
11. Punyaprasava, the heaven of fortunate birth.
12. Brhatphala, the heaven of great results.
13. Abrha, the passionless heaven.
14. Atapa, the heaven without heat or affliction.
15. Sudrsa, the heaven of perfect form.
16. Sudarsana, the heaven of perfect vision.
17. Akanistha, the highest heaven.

Southern accounts 2 generally agree in enumerating the Rupa heavens as sixteen, agreeing on this point with the
1 8-2a, cf. M.Vy. 219 ; D.A., p. 926.
2 M.B. 26, C.P. 138.
 
Kasmira Sarvastivadins, though there are points of divergence on names. Mahayanists 1 for the most part, and practically all schools in China and Japan, give the full number, eighteen.

In this connexion the Buddhist theory of Dhyana must be taken into consideration, for the Dhyanas are intimately associated with the Rupa heavens, and the various heavens are classified according to the Dhyana to which they appertain. Dhyana means meditation or contemplation, but later the word was used to indicate a special type of meditation, in which four or five states were distinguished.

When only four Dhyanas are spoken of, they are as follows : 2

(1) supernatural ecstasy associated with vicara and vitarka ;
(2) ecstatic contemplation no longer associated with either vicara or vitarka, or, in other words, where reasoning gives way to intuition ; (3) contemplation where ecstasy gives way to serenity ; (4) deep meditation where the mind becomes indifferent to pleasure and pain. According to the Abhidharma Kosa, 3 the first three rupa heavens are gained as the result of the practice of the first dhyana, and so are called the First Dhyana Heavens. In like manner, the next three are called the Second Dhyana Heavens ; the next three the Third Dhyana Heavens ; and the last eight the Fourth Dhyana Heavens.

The division of the Dhyanas into five instead of four is frequently made in the later schools, more especially in those inclined to esotericism. 4 In this case the second of the four Dhyanas, " Intuitive Meditation," is divided into two parts, the first corresponding more or less to the occidental " instinct ", and the second to " spiritual perception not derived from intellect ". In the schools which enumerate both five Dhyanas and eighteen rupa heavens, the relation of the Dhyanas and the heavens is as follows : The first three to the first Dhyana ; the next three to the second Dhyana ;

1 Y.B. 4-14a. 2 M.Vy. 112, Y.B. ll~43a.
a 8-2a, cf. also A.S.P. 2-76. * But also in Neo. Stu. C.P. 141.

the next three to the third Dhyana ; the next four to the fourth Dhyana ; and the last five to the fifth Dhyana. Associated with the five Dhyanas are the five Dhyani Buddhas, the supreme lords of Nepalese and Tibetan Buddhism, as well as the so-called esoteric branch of Mahayana generally. These, however, as accepted by none of the three schools under consideration, may be ignored.

As regards the size of these various heavens, we find two principal accounts. According to one, the first Dhyana heavens are in size the same as the earth ; the heavens of the second Dhyana are in size equal to a small chiliocosm ; the heavens of the third Dhyana to a middle chiliocosm ; and the heavens of the fourth Dhyana in size equal to a great chiliocosm. According to the other account, the heavens of the first, second, and third Dhyanas are respectively equal in size to a small, a middle, and a great chiliocosm, while the extent of the heavens of the fourth Dhyana is measureless.
 
Arupa Dhatu. 1

Higher than the Rupa Dhatu, or the realm of form, is the Arupa Dhatu, or the realm of formlessness, in which the inhabitants have neither desire nor such a thing as a body, consciousness alone remaining. Nearly all the schools enumerate the heavens of this realm as follows :

1. Akdsdnantydyatana, the heaven of boundless space.
2. Vijftdndnantydyatona, the heaven of infinite con sciousness.
3. Akimcanydyatana, the heaven of absolute non-existence.
4. Naivasamjndnasamjndyatana, the heaven of neither consciousness nor unconsciousness.

Rebirth in these realms is said to be due to meditation on the four immeasurables, one for each heaven. Yet, notwith standing their many merits, rebirth in one of these abodes was deprecated, as the duration of life is so inconceivably long that progress to the supreme goal of Nirvana is seriously
1 M.Vy. 221 ; M.B. 26 ; C.P. 139 ; A.K. 8-2a ; Y.B. 4-14a.

delayed. In fact, rebirth here is one of the eight Akusalas or evil moments, eight times of birth, not conducive to enlightenment.
The Buddha Ksetras.

So far, the Mahayana and Hinayana accounts of the various realms of existence have been more or less in agreement, but on one point there is wide divergence. We have already noticed that while the performance of good deeds will result in rebirth in one of the Devalokas or Brahmalokas, etc., yet as such a rebirth is frequently a hindrance rather than a help, one wonders where a man should strive to be reborn in order to reap the greatest advantage where there is a place free from affliction and yet is one where spiritual progress may be made.

Practically all forms of Hinayana are silent on this point, but in Mahayana we find an attempted solution in the doctrine of the so-called Buddha Ksetras or Buddha lands, reference to which will be found in practically all the sutras of this school. The basic idea of this dogma seems to be that every man upon reaching supreme and perfect enlightenment acquires a spiritual realm, to which he repairs after death and in which he continues to instruct his Bodhisattvas and other persons who may be born there, leading them to supreme enlightenment for themselves. These manners of rebirth naturally appealed very strongly to the Mahayanists, and consequently in many of their devotional writings we find the authors piously desirous that the merit which they may have acquired by instructing the world through their written works may secure for them at death rebirth in one of those lands.
As usual, there is great inconsistency in the various accounts and enumerations of these Buddha Ksetras, since every Buddha field is really so great and so large as to be co-existent with the universe, and yet at the same time there are as many Buddha fields as there have been Buddhas. In the earlier days of Mahayana a man endeavoured to be reborn in any

Buddha land, or at least in the realm of the Buddha who took his especial fancy. As time went on, however, Mahayana became more monotheistic, particularly in its Chinese and Japanese phases, and the belief grew that there was but one supreme and universal Buddha, who included and over shadowed all the rest. Consequently, the Buddha land of this Buddha, whoever he happened to be, came to be the goal of ordinary ambition. Historically, we find that there were two monotheistic strains, which finally resulted in two different schools, in one of which the supreme being is known as Vairocana, and in the other Amitabha. Consequently, we find likewise two main lines of development in the Buddha Ksetra theory.
(a) Buddha Ksetras in the Vairocana schools. The followers of Vairocana were for the most part the esotericists and occultists who were generally more concerned with the control of one s destinies in this life than in the world to come, so that the paradise doctrine of this school did not receive the same development as did the Amitabha paradise doctrine. Nevertheless, the Buddha Ksetra idea was not wholly neglected. In certain cases the ten or thirteen stages of Bodhisattvahood were symbolized as material heavens, in the highest of which dwelt Vairocana himself. Even more interesting are the thirteen stages of the path of progress of the righteous departed as taught in the esoteric school of China and Japan :
 
1. Acala, where the soul remains for the first week after
death.
2. Sdkyamuni, for the second week.
3. Manjusri, for the third week.
4. Samantabhadra, for the fourth week.
5. Ksitigarbha, for the fifth week.
6. Maitreya, for the sixth week.
7. Bhaisajaguru, for the seventh week.
8. Avalokitesvara, for another hundred days.
9. Mahasthdmaprdpta, for another year.
10. Amitabha, for another three years.
11. Aksobhya, for another seven years.
12. Akdsagarbha, and
13. Vairocana for ever.

(b) Buddha Ksetra of the Amitabha school. Turning now to the other school, we find that the Amitabha cult is based almost exclusively on rebirth in paradise through self- renouncing adoration to Amitabha, so that among the followers of this school the Buddha Ksetra doctrine received very great development. Originally but one out of many equally important realms (the Avatamsaka Sutra assigns to it a still lower position), it came gradually to assume pre-eminence among all the paradises, possibly because by chance it was placed in the west. The principal scriptural authorities for the doctrine are :

(1) The larger Sukhdvati-vyuha Sutra, which, dealing with the past, tells how this marvellous land came to be created through the great vow of Dharmakara Bodhisattva, who later became Amitabha Buddha ;
(2) the Amitdyur Dhydna Sutra, or the sutra of the meditation upon Amitabha, which deals with the present, or how men should so conduct themselves as to secure rebirth in Amitabha s land ;
(3) the smaller Sukhdvati-vyuha Sutra, which deals with the future, i.e. the future condition of those who gain this Ksetra, and gives a long description of the great bliss that is to be found there. Apart from occasional mention elsewhere, two works in the Sastra literature of India deal more particularly with the doctrine. One of these is a Bodhisattvabhiimi, or discourse on the stages of a Bodhisattva, ascribed to Nagarjuna, the other a com mentary on the Sukhavati-vyuha called the Amitdyur- sutropadesa, ascribed to Vasubandhu. In both of these works the purely material side of the paradise doctrine is retained, and Amitabha is still but a single and semi-historical Buddha, and his paradise but a single even if important Buddha Ksetra. The rather monotonous tone of the descriptions of delights to be found in these various works may be seen from the

translation of the three Sukhavati Sutras in the sacred books of the East.

4. THE WORLD OF SENTIENT BEINGS 1

Classification of Sentient Beings.

We have already seen the geographical position of the various divisions of the three dhatus. We must now study the persons who inhabit these realms.
First of all in this connexion must we give the principal categories into which sentient beings are divided. These are : (1) the five or six gatis, or destinies ; (2) the seven vijnana sthitis, or bases of consciousness ; (3) the nine abodes of sentiency.
The Five or Six Gatis. This is the most important of all Buddhist classifications of sentient beings, and is the basis of the various Buddhist wheels of life or charts of existence. The five-fold division is made by most branches of Hmayana, the six-fold division by a few branches of Hmayana and most branches of Mahayana. The five gatis 2 are :

1. The inhabitants of the Narakas or hells.
2. Preta, ghouls, goblins, or demons.
3. Animals.
4. Mankind.
5. Devas or gods.

Where a sixth gati is added, it consists of the Asuras titanic, demonaic monsters, somewhat akin to the Devas, with whom, however, they are constantly at war. Those who defend the five gati theory never doubted the existence of the Asuras. They were merely not of sufficient importance to be given a place as a separate destiny, and were grouped either with the pretas or animals or both. Sentient beings when they die are reborn into one or the other of these five destinies. There is no other form of existence possible. The geographical
1 For Ceylon, M.B., chap. ii. 2 A.K. 8-5a.

relationship between the three dhatus and the five gatis has already been given.

The Seven Vijnana Sthitis. The three dhatu and the five gatis are likewise divided into those realms which do and those which do not support consciousness. Those which do are divided in a seven-fold manner, forming the seven Vijnana sthitis, which are x :

1. Those realms in which the bodies as well as the thoughts of the inhabitants differ from one another. This includes the whole of the Kama dhatu and also the first Dhyana heavens of the Rupa dhatu, except during the period of creation.
2. Those realms in which the bodies are diverse but the thoughts uniform. This consists of the first Dhyana heavens during the period of creation, inasmuch as at that time all are filled with the single thought, " We have been created by Brahma."
3. Those realms in which the bodies are the same but the thoughts diverse. This consists of Abhasvara and the other heavens of the second Dhyana.
4. Those realms in which both the bodies and thoughts of the inhabitants are uniform. This consists of the heavens of the third Dhyana.
5. 6, 7. The last three vijnana sthitis consist of the three lower divisions of the Arupa dhatu.

The three evil places (hells, pretas, and animals), the heavens of the fourth Dhyana, as well as the fourth heaven of the Arupa dhatu, are not classed as vijnana sthitis, inasmuch as life and conditions therein do not serve as stimuli or supports (sthiti) of consciousness.
The Nine Abodes of Sentiency. 2 The nine so-called abodes of sentient beings consist of the above seven vijnana sthitis, with the addition of the fourth, Arupa dhatu, and the unconscious deities of the fourth Dhyana heavens. These realms are so called because sentient beings exist there willingly, or, in other words, they are suitable abodes for 1 A.K. 8-7a. a M.Vy. 169, A.K. 8-96.

living beings. The evil realms are not included in this category, since they may be likened to prisons rather than abodes. Nor, for somewhat more metaphysical reasons, are the other sections of the fourth Dhyana included.

The Four Kinds of Birth, and the Stages of the Foetus. Sentient beings revolve in an eternal circle through the five gatis. Dying in one gati, they are reborn in another. There are four methods whereby this rebirth may take place. 1 They are : (1) Birth from an egg ; (2) birth from a womb ; (3) birth from slime ; (4) apparitional birth. Instances of egg birth are various kinds of birds ; of womb birth, such animals as the horse, cow, dog, etc. ; of slime birth are mosquitoes, flies, and various other insects. Apparitional birth is so called because it is miraculous, without visible support, and with all the organs instantaneously perfectly formed. Of such birth are the inhabitants of the various heavens and hells. The normal method of birth for man is from the womb, but occasionally the other three methods occur. The animal gati also includes all four methods of birth. Three have already been mentioned. Apparitional birth takes place in the case of dragons, etc. All the denizens of the heavens and hells have no other method of birth than apparitional. Pretas are sometimes born from the womb, sometimes apparitionally. Of all the various kinds of birth, apparitional birth is the best, but Buddhas and Bodhisattvas receive womb birth that their humanity may be complete.

In the case of womb birth, the foetus goes through a regular order of development. Five principal stages of embryo development are usually enumerated. These are 2 : (1) Kalala ; (2) Arbuda ; (3) Pesi ; (4) Ghana ; (5) Prasakha.
The Doctrine of the Intermediate Existence? There was a good deal of difference of opinion among the various Buddhist sects as to whether or not at death a being passed at once
1 M.Vy. 168; A.K. 8-116.
2 A.K. 9-86.
J Long discussion, A.K. 8-136 ff.

into Ms new existence, or whether there is a short period of existence in an intermediate state. The Mahasanghikas denied the doctrine of the intermediate existence, while the Sarvastivadins and most of the later schools accepted the doctrine, which later was worked out in some detail. The following points concerning the Sarvastivadin doctrine on the subject may be of interest.

The being in the intermediate state is possessed of a definite body, but of a very subtile kind. In shape it is of the same kind as his future existence. Thus, in the case of a person destined to be born as a man. The intermediate body is of the same size as a child 5 or 6 years of age. The organs are always complete. The intermediate bodies of those destined for the Rupa dhatu are possessed of clothes as are Bodhisattvas and some others, but the majority of beings destined to the Kama dhatu are nude. Such beings can see other beings of the same class, in certain cases of the classes below them. They are possessed of certain magic powers, such as being able to pass through space, and on the death of the previous person the new intermediate body is attracted to the place of its new birth through desire, going there miraculously. Its destiny is fixed. The body destined for human birth never develops into another existence. It partakes of no solid food, but receives its sustenance from smell. The duration of its existence is indefinite, existing until such time as the new body is prepared for habitation, though ordinarily, say some, it lives for only one week or seven weeks. If a male it enters the right side of the mother s womb and faces the back, if a female it enters the left side of the womb and faces front. In the case of twins the child born last is the oldest.
Size and Duration of Life of the Various Realms. 1 A word must be said concerning the nature of life in each of the various gati through which sentient beings pass. First as regards their stature. No exact figures are given for the destinies lower
1 A.K. 11-156 ff. ; N.A. 31-21a ff. ; C.P. 142 ; Y.B. 4-14a and 4-146.

than man. In the southern continent (Jambudvrpa) the size of men varies from time to time, but the average size, at least for the present, is 3|- or 4 hasta or cubits. For the eastern continent the average size is 8 cubits, for the western continent 16 cubits, and for the northern continent 32. In the heavens of the Kama dhatu statures are as follows :
 
1. Heaven of the four Great Kings . J krosa.
2. Heaven of the thirty- three Gods . \ krosa.
3. Yama heaven f krosa.
4. Tusita heaven .... 1 krosa.
5. Nirmanarati heaven . . 1J krosas.
6. The highest Kama dhatu heaven . \\ krosas.
/
In the Rupa dhatu statures are as follows : In the first J yojana, in the second 1 yojana, in the third 1 J yojanas, in the fourth 2 yojanas, in the fifth 4 yojanas, and so on, doubling (except for Anabhrakas) so that the highest has 16,000 yojanas for its average stature. In the Arupa dhatu the inhabitants have, of course, no bodies.
As regards duration of life there is no constant for the southern continent, as it varies between a asamkhya and ten years ; for the eastern continent, 250 years ; for the western continent, 500 years ; for the northern continent, 1,000 years.

In the first Devaloka, the heaven of the four great kings, a day and night is equal to 50 human years. Their months consist of thirty such days, their year twelve such months, and the average duration of life is five hundred such years. In the second heaven a day and night are equal to a hundred human years, a month has thirty such days, a year twelve such months, and the inhabitants live for a thousand such years. In each of the higher realms the numbers are doubled. Thus, in the Yama heaven a day and night are equal to 200 human years, a month contains thirty such days, a year twelve such months, and the inhabitants live for 2,000 such years. Since there is no sun nor moon in these realms, day and night are marked by the opening and closing of sacred lotuses.

In the Rupa dhatus there is neither day nor night, and the duration of life is measured by Kalpas. In the first the average duration of life is one middle kalpa, or J a mahakalpa ; in the second 2 middle kalpas, or | a mahakalpa ; in the third 1 kalpa ; in the fourth 2 kalpas ; in the fifth 4 kalpas ; and so on, doubling in such a way that the number of yojanas for average height is the same as the number of kalpas, the highest being 16,000 kalpas.
In the Arupa dhatu, or realm of Formlessness, we have even greater figures. In the first the duration of life is 20,000 kalpas, in the second 40,000 kalpas, in the third 60,000 kalpas, and in the fourth 80,000 kalpas.

Finally, the duration of life in the three evil gatis (animals, pretas, and hells) must be taken into consideration. Animals have no fixed duration of life. Some live only for a moment and some for centuries. The greatest span of life is enjoyed by a species of dragon which lives for one middle kalpa. The day of a preta is equal to a human month. Thirty such days make a preta month, twelve such months a preta year, and a preta lives for 500 such years. No exact figures are given for any of the hells, save the hot hells. The whole duration of life of the heaven of the four great kings is equal to a night and a day of Samjiva, and the inhabitants live for 500 years of such days. The duration of life in the next six hells corresponds in a similar manner to life in the remaining five deva heavens. In the seventh hell the duration of life is half a middle kalpa, and in Avici the duration of life is a whole middle kalpa.

Nature of Life in other Realms} A few out of the many other such details of life in the other gatis as found in the Buddhist books are as follows : Every form of sentient being is under the necessity of taking food, though there are four
1 Cf. especially A.K. ll-13a ff. ; Y.B. 4-146 ff.

kinds of food : l (1) corruptible food, i.e. food capable of being digested, which is the food used by all forms of men and by the Devas of the six Kama heavens ; (2) food that is partaken by contact only, which appertains to the upper divisions of the Kama heavens and the lower regions of the Rupa dhatu ; (3) food that is partaken of by contemplation, as is the case with the upper regions of the Rupa dhatu ; (4) food that is partaken of by the knowledge of it. This applies only to the inhabitants of the Arupa dhatu.
In all of the heavens save those of the Arupa dhatu clothes are used, though sex remains only in the heavens of the Kama dhatu. There are five ways of satisfying sexual desire : 2

(1) by copulation ;
(2) by embracing ;
(3) by the holding of hands ;
(4) by laughing or smiling at one another ;
(5) by looking at one another. Living in contact with the earth, the inhabitants of the heaven of the four great kings and also of the heaven of the thirty-three gods, unite by copulation.

In the Yama heaven a single embrace produces a new being. In the Tusita heaven the mutual holding of hands suffices. In the Nirmanarati heaven smiling, and in the highest Kama heaven a single glance constitutes sexual union. Birth (which is apparitional in all the heavens) takes place in the Kama heavens as follows : Shortly before a deva is born one of the devis finds a flower in her hand. She knows by this fact that a child is to be born to her, and accordingly after seven days birth takes place, often, however, by suddenly appearing on her knees. At the time of their birth, the children are as if five to ten years of age. There then appears spontaneously a precious vessel filled with divine food, partaking of which the new-born being grows in size like the rest of the devas, while magic trees provide them with necessary garments. The inhabitants of the Rupa and Arupa dhatus are born fully grown and without the aid of any sort of parent, and in the Rupa dhatu are born fully clothed. We are also informed that the gods of all three realms speak
1 M.Vy. 169. a A.K. 11-126; N.A. 31-16a.

only the Aryan language, a sort of heavenly Sanskrit, which moreover they speak correctly without having to learn it.
In the Kama dhatu there are three ways of enjoying pleasurable objects, the first of which applies to men and the four lower deva lokas, and is by ruling over and enjoying the sense pleasures they find around them. The second, which applies to the fifth deva loka, by creating pleasurable objects and then ruling over and enjoying those things which they themselves have created. The third, which applies only to the sixth deva loka, is by ruling over and enjoying the sense objects especially created for their enjoyment by others.
Even in the most pleasurable regions (apart from Sukhavati) life must come to an end, and the devas must die here to be born elsewhere. As they begin to grow old five signs of decadence begin to appear. These are : (1) The flowers upon their heads begin to decay ; (2) their eyes grow dim and move uneasily in fear of the change which they know must come ; (3) the lustre of their bodies begins to fade ; (4) perspiration begins to exude from under their arms ; (5) they listlessly absent themselves from their proper places.
In the various Buddha ksetras the pleasures are even greater and more lasting, but of a less sensual nature. Most of the pleasure is of a spiritual nature, and consists of listening to the holy law. According to the older schools of Buddhism, these Buddha ksetras themselves are not permanent and life therein is only the preparation for Nirvana or Buddhahood, but in the Shin sect of Japan rebirth into Sukhavati is itself the highest goal, and is final and complete happiness.

PART II
COSMIC ANALYSIS

Subjective and Objective Classification.

We now come to the consideration of the ultimate elements into which the Buddhists thought the universe could be decomposed a subject which promises to be of greater interest. Here, at the very outset, we are faced with a very curious situation, for we find a two-fold analysis of the universe, one subjective and the other objective. In the former instance the nature of any one personality is examined, and by a process of analysis the seemingly unified personality is broken up into a number of component parts, which are ultimate or elemental, and since every personality is a microcosm closely corresponding to the macrocosm it follows that the component parts of the personality are also the component parts of the universe.

The objective classification is merely a re-arrangement of these component factors in a more scientific way, i.e. by examining the whole universe, irrespective of any one personality. It follows from this that the objective and subjective classifications are mutually inclusive, and that the difference between them is merely one of standpoint.
In point of fact the origin of the two categories is due to the peculiar nature of the Buddhist historical development. Primitive Buddhism was founded upon an agnostic basis, 1 certainly as regards the external world. The Buddha declined to state whether it was infinite or finite, whether it is eternal or non-eternal. Consequently, for primitive Buddhism a complete list of the elements of being, approached from an objective point of view, was impossible.

Nevertheless, a certain amount of subjective analysis was permitted, and, in fact, encouraged. In order to eliminate
1 Cf. especially the agnostic passages from the Sutra Pitaka collected by Warren in his B. in Trans., chap. 2.

the belief in the atman as taught by the Upanisads, the Buddha is said to have stated that the personality is not a unit, but a compound of various factors, such as the material body, consciousness, feeling, ideas, volitions, etc. In the early days little further analysis seems to have been attempted. But this was sufficient start for the Indian mind, always given to analytical subdivisions. Each of the main groups became many times divided, until a very complex chart of the factors of life was eventually tabulated.
Once this had been done, it is easy to understand how the next stage, the re-grouping of these component parts from the objective point of view, came to be undertaken. The early agnostic position of primitive Buddhism was soon neglected, 1 and, the categories already enumerated including all forms of life and all aspects of the external universe, it soon became obvious that it was more logical to re-arrange the categories in such a way as to form a complete philosophic analysis of the factors of being.
We see, therefore, that the subjective analysis was earlier and less systematic than the objective analysis, and that subsequently the latter tended, as more logical, to over shadow the forme\ Having been embodied in the sutras, however, the subjective classification was never lost sight of, and we find that occasionally the later philosophers, including Buddhaghosa, preferred to revert to the earlier grouping, although acknowledging the validity of the objective classification.

(A) THE SUBJECTIVE CLASSIFICATION
Owing to its priority in point of time, and also to its greater simplicity, it is advisable to consider the subjective classification first. This consists of three categories, viz. 2 :
1 See e.g. f. 30 of A.K., where some of the older agnostic passages con cerning the soul are used by Vasubandhu to deny its existence.
2 C.P. gives these categories but scant attention, but they are enumerated pp. 182-3. An excellent discussion of each category from the Pali point of view will be found in Mrs. Rhys Davids Buddhist Psychology, where original sources are quoted.

1. The Five Skandhas.
2. The Twelve Ayatanas.
3. The Eighteen Dhatus.

In the present instance it is more convenient to consider the five skandhas as one division and the twelve ayatanas and the eighteen dhatus as a second division. 1

1. The Five Skandhas

General Discussion.

The five skandhas constitute the component parts of a personality, though, certainly in later Buddhism, they are not ultimate factors, inasmuch as each of them is subject to sub division. It is probable that the very name implies that each of them was considered a complex group rather than a unit, for the word means heap, collection, group. Vasubandhu (A.K. 1-146) cites the following passage from the sutras : " All things possessed of form, whether past, present, or future, whether internal or external, whether coarse or fine, whether mean or great, whether distant or near all such things constitute one skandha, called the rupa skandha." "

All the ultimate factors are classified into five groups or skandhas, and these five groups constitute the personality. This doctrine of the five skandhas is undoubtedly very old. Frequent mention is made of them in the sutras, where they are chiefly cited in disproof of the atman theory. The soul (atman) is not, nor is it the possessor of, the body, feelings, ideation, etc. In fact, there is no soul at all, but the personality consists of these groups and nothing more.
Nowhere in the sutras is there a categorical or logical definition of the five skandhas, and it would seem as if each of the terms taken separately formed part of the common property of contemporary Indian thought, and that the only

1 For Sarvdstivddins see especially A.K., the whole of the first and second fasc. ; Sang. Par. 6-la ; D.Sk., f. 17 for indriyas, f. 18 for ayatanas, and f. 19 for skandhas. For Yogdcdrins see especially P.Sk. and P.Sk.Vai. ; Ab. Sam. San., 1-16 ff. V. M. S., like C. P., pays little attention to the subjective classification.

originality displayed by Buddhism was the doctrine that these skandhas and nothing more constituted the personality.
Unfortunately, for us the names of the five skandhas are not so illuminating as they seem to have been to the ancient Indians, and we find in the Occident a wide variety of terms used to translate them. The names and arrangement of the skandhas is almost invariably as follows :

1. Rupa. 3. Samjna.
2. Vedana. 4. Samskara.
5. Vijnana.

1. Rupa, literally form or shape (sometimes colour), corresponds roughly to our matter, and in the personality implies the physical body. Among the many other definitions are ruppati ( = whence rupa), a most difficult word to translate, but which implies " subject to transformation ", "affected", "disturbed", "disintegrating", "modified" (A.K. l-10a).
The other definition, more in accord with our own ideas on the subject, is " that which resists " or " impenetrable ".

Primitive Buddhism gives no definite subdivision of this skandha, other than that stated in the verse quoted above, but we know that the term rupa referred to the matter constituting our own and other bodies, as well as inanimate matter. An enumeration of the thirty-two parts of the body is frequently given, of which the five sense organs are the most important, while external matter is also given a five-fold classification (corresponding to the five sense organs). We know, moreover, that all matter, whether internal or external, was supposed to be derived from the four elements, earth, water, fire, and air, which constitute the four ultimate states of matter. 1
The Sthaviravadins subdivided matter a little more systematically. 2 Four divisions were ultimate or non-
1 Cf. the famous Kevaddha sutra in D.N. (D.B. 1-276), corresponding to D.A. 81a.
2 For authorities on these and the following statements see the list given under objective analysis below.


derived, while there were twenty-three or twenty-four factors derived from the four elements. These and all the following subdivisions will be considered at length when we come to deal with the objective analysis of the universe. The whole of the classical metaphysical literature of this school, as well as Buddhaghosa and the Abhi. Sang. (C.P.), follow the sutras in ignoring the doctrine of paramanus or atoms, but in modern times the atomic theory seems to have been adopted to explain the older doctrines.
The Sarvastivadins accepted the atomic theory (as we shall see hereafter), and the idea of the four ultimate elements from which are derived eleven fundamental material factors, a somewhat more systematic list than that of the Sthaviravadins.

The Yogacarins, being idealists, thought that all matter is but the creation of the mind. Nevertheless, from the relative point of view, they followed the Sarvastivadins very closely and accepted the four elements and the eleven derivatives (though the Yogacarin eleventh factor differs from that of the Sarvastivadin). The early Yogacarin philosophers, such as Asanga and Vasubandhu, likewise accepted the atomic theory, but this was denied by later thinkers such as Dignaga and Dharmapala as being inconsistent with idealism.
2. Vedand, the first of the four immaterial skandhas, is sometimes translated sensation, but careful study of the texts shows that it corresponds more closely to our own term feeling, for, in the first place, sensation in the sense of awareness is not Vedana but Vijnana, and, secondly, the fundamental division of Vedana into pleasant, unpleasant, and neutral, or sometimes into pleasant, unpleasant, joyful, sorrowful, and neutral (the first two physical, the next two mental, the last both), shows that the hedonistic side of Vedana is emphasized. The fact that it is correlated with the sense organs (thereby giving it a six-fold classification) shows that it may be more adequately defined as " sense feeling occasioned by sense impressions ", which is almost word for word the definition given it by Vasubandhu (A.K. 1-116).

3. Samjna is sometimes rendered " perception " and some times " conception ". That it is not mere sense perception is, I think, clear, when Samjna is compared with Vijnana. The perceptual side seems to be more emphasized by the Pali or Sthaviravadin commentaries, as when the Vibhanga, to give Mrs. Rhys Davids translation, 1 divides Samjna into

(1) cognitive assimilation upon the occasion of sense, and
(2) cognitive assimilation on the occasion of naming.

The Sarvastivadins and the Yogacarins, on the other hand, emphasized the conceptual aspect. Vasubandhu (A.K. 1-116) defines it as " the grasping of the differences of characteristics ", and, again, " Samjna skandha has for its essence the grasping of images, i.e. it seizes hold of the attributes, blue or yellow, long or short, male or female, pleasant or unpleasant, antipathetic and sympathetic, etc." Personally, I favour the term " ideation " as a translation of Samjna, and Mrs. Rhys Davids tells me that this will also cover the Pali use of the term.

The correlation of Samjna with the sense organs is also seen by its division into six categories, as in the case of Vedana.
4. Samskara. Unquestionably, the most difficult term to explain is Samskara, and a large number of different renderings have been given, ranging from Spence Hardy s " conscience ", which is certainly wrong, to Professor Ehys Davids " confections ", which is perhaps the most correct etymologically, though personally I should prefer " co efficients ". In point of fact, however, Samskara early became associated with Karma, or action, as may be seen by its position in the Pratitya Samutpada. It has thus been rendered into Chinese (f^, to do, to perform), and in the Samyukta Agama it is said " all creatively active Sasrava
1 B.Psy. sect, on Skandhas.

samskrta dharmas are called the Samskara-upadana- skandha " (S.A. f. 13).

In many of the sutras some attempt seems to have been made to limit Samskara to volition (Cetana volitional mentation). Thus the Samyukta Agama (op. cit.) says : " The group of six volitions (corresponding to the six sense organs) constitutes Samskara skandha " (also quoted by A.K. 1-lla).

The definition of Samskara as volition would have rounded off the Buddhist list of the five skandhas very well, but as psychological analysis continued, and further factors in the mental process were formulated, a place had to be made for them in the classification of the factors of the personality. Here there was a difficulty. The later Buddhists dared not add to the five-fold classification which they believed to have been laid down by the founder, so that the newly postulated factors had to be arranged somewhere inside the five already existing skandhas. The most convenient dumping ground was found to be Samskara, which thus became a weird medley of otherwise unclassified mental factors. Thus Mrs. Rhys Davids (Bud. Psy., 51), speaking of this matter, says : " The constructive aspect (of this skandha) was reserved for ... volition. The other fifty-one factors (of the Pali enumeration) are rather co-efficients of any conscious state than pre eminently active or constructive functionings."

Vasubandhu feels that an apology is needed for including this additional material, though his excuse is rather lame. He says (A.K. 1-116) that the reason the Buddha stated that Samskara skandha consists only of the six-fold Cetana is because this factor is the most important, since Samskara means creative activity, and in Cetana this attribute is more predominant than in the other factors. Nevertheless, the other factors must not be excluded, " for if this were done the remaining Caitasikas, and Citta Viprayukta dharmas would not be included in any of the skandhas, and consequently they would be independent of suffering and the cause of suffering, and hence could not be cut off and could not be known. For, as the Blessed One has said, " If a single thing be not analysed, or remain unknown, I say that it is impossible to attain the end of suffering ..." Consequently, all samskrta activities not included in the other four skandhas are placed inside the Samskara skandha.

Samskara thus came to be used as a term denoting all the mental concomitants which are at any time associated with the arising of Vijnana or consciousness. Consequently, since Vedana and Samjna come under this category, they also were enumerated a part of Samskara, so that from the absolute point of view the five categories were reduced to three, viz. :

(1) The Body.
(2) Mental properties, or concomitants of consciousness.
(3) Consciousness.

Nevertheless, respect for tradition prevented the older and clumsier five-fold classification being disregarded, so that we find the later commentaries trying to explain why Vedana and Samjna are included in Samskara and yet also listed as separate skandhas. Vasubandhu has the following remarks on the subject (A.K. 1-166) : " Among the Caitasikas, Vedana and Samjfia are considered separate skandhas because they serve as causes of pugnacity and of the wheel of birth and death, and because of the due order of their functioning. There are two bases of pugnacity. One is the expression of various desires, the other is the expression of various opinions. Vedana and Samjna act respectively as their predominating cause. Owing to the emotions (Vedana) all the desires are made manifest. Owing to erroneous ideation (Samjna) all theories are made manifest. Both birth and death have Vedana and Samjna as their principal causes, since the wheel of birth and death is set rolling owing to man s enslavement to the emotions and his attachment to erroneous theories. For these two reasons and because of the due order of their causal functioning, to be explained hereafter, Vedana and Samjna are considered separate skandhas."

A word must be said concerning the subdivision of Samskara. Apart from Cetana, which is specifically mentioned, many of the mental properties later included in this skandha are found separately enumerated in the sutra Pitaka, but no attempt seems to have been made to give a detailed or definite categorization.
In the Hinayana Abhidharma period, elaborate charts of the Samskaras were compiled, 1 the individual items con sisting of the various mental properties casually mentioned in the sutras, together with a certain number of factors deduced by individual introspection. It is to be regretted that so many are of the former type, since the use of terms is very vague in the sutras and many psychological terms were but as obiter dicta, which were later enshrined in the Abhidharma works as final revelations of ultimate truth. The right of such items to be considered ultimate factors was therefore very acutely defended by minds able to formulate a far more scientific analysis of mental components.

The Sthaviravadins enumerated fifty- two such Samskaras. 2 Among the Sarvastivadins there was much variation in number for some considerable time, and it was not until the time of Vasubandhu that the number was definitely put at forty-six, which afterwards remained the standard figure. The Yogacarins hovered between fifty-one, fifty-two, and fifty-three, but fifty-one was eventually considered the orthodox figure. 3 Incidentally, the Yogacarin fifty-one is by no means in agreement with the fifty-two Caitasikas of the Sthaviravadins.
5. Vijndna, the last of the skandhas, is usually translated " consciousness " or " cognition ". This definition is quite in accord with all the commentaries, and many references to
1 Of. e.g. Dham. San., part ii (Book I of R.D. trans.).
2 At least after Buddhaghosa, see Expositor.
3 Y.Bh. = 53 ; Ab. San. = 55; A.V.P. P.Sk. S.dh. = 51 ; and V.M.S.

the term shows that it denoted for the Buddhists merely " awareness " in the broadest sense of the term. Hence it is associated with much which we should call sensation, save that it lacks the hedonistic element which is given to Vedana. Again, it is associated with the perceptual aspect of Samjna, save that it is ampler in its scope, implying not merely the " seeing "of a thing but the full awareness of it, or the absorption of the image into the conscious mind.

The distinction between Vijnana and Samskara, particularly in the later use of the latter term, is that Vijnana is " con sciousness " or the " various aspects of consciousness ", and Samskara is the contents or functional phases of con sciousness. Thus the awareness of a visual object implies attention (manaskara), sensation or contact (sparsa), etc. And further, as Vedana and Samjna are considered as Samskaras, we find that even these are phases or functions of consciousness rather than independent realities.
That Vijnana involves both the sensatory and ideation aspects of consciousness can be seen from the minimum six-fold classification of Vijnana common to all forms of Buddhism, however many additional aspects may be added. The six divisions consist of five kinds of consciousness dependent respectively upon the five sense organs, and one type of consciousness dependent upon the operation of the mind (manas). This last possesses several functions peculiar to itself, such as intellection, reasoning, and memory.
The most important addition to this six-fold category was made by the Yogacarin, who added two more, making eight in all. To this, certain other authorities, chiefly Chinese, have added a ninth. 1

Among the Sthaviravadins, in addition to the six-fold group, Vijnana was also classified in a different way under eighty-nine different heads. This arrangement is peculiar to this school, and had no effect upon later philosophic speculation.
1 Cf. sect, on Vij. in ^ g| and jj, fc ^.

Concluding Remarks.

Turning now from the skandhas taken separately, let us consider them for a moment as a whole. To us there seems little logical basis for this five-fold division of the personality, and it would seem as if the Buddhists themselves were struck by its pragmatic nature, and that the non-material part of the personality was arbitrarily broken up into four co-ordinate parts chiefly in order to emphasize the complex, compound nature of the mind.

Nevertheless, the five-fold division is not altogether lacking in psychological insight, and, more particularly if the volitional aspect of Samskara had been retained, we should have had a certain correspondence between the east and those western psychologists who divide the mind into (1) feeling, (2) reason, and (3) volition, corresponding respectively to Vedana, Samjna, and Samskara, which, with Vijnana, consciousness considered as a whole, and Kupa, the body, would give us the following classification :

1. The body, including sense organs.
2. Consciousness, or awareness, reception of the sense
impressions transmitted by the sense organs.
3. Resultant feeling of like or dislike of these impressions.
4. Ideation or the formation of mental images concerning
the nature of the external world from which sense impressions are derived, including the classification (naming) of those objects which are pleasant and those which are unpleasant.
5. Volition or will with respect to choosing as far as
possible those objects which are pleasant and those which are unpleasant. Later, as we have seen, mental activities other than volition were added, but were placed in the same category.

It was only the early mistranslation of the names of the five skandhas which prevented it being seen that some such

 MAOTAL OF BUDDHIST PHILOSOPHY

scheme was in the mind of the early Buddhist philosophers. In this connexion, and more or less in defence of this position, one or two points deserve attention.
Vijnana follows immediately upon the interaction of the sense organs and sense object, and without the intermediary action of a separate faculty such as Vedana or Samjna. This is obvious from the frequently repeated texts to the effect that " as a result of visible object and organs of sight visual con sciousness comes into being ", and also from the correlation of the sixth sense objects, six sense organs, and six aspects of consciousness to form the eighteen dhatus or factors of existence. From this fact, the sensatory aspect of Vijnana becomes obvious, a fact that is sometimes overlooked, owing to Vijnana s place among the skandhas, Vedana and Samjna intervening between it and Rupa, the body. We find, however, that while retaining the traditional order in the bare enumeration of the skandhas, both Buddhaghosa and Nagarjuna give their exposition of Vijnana immediately after their explanation of Rupa.

Vedana, Samjna, and Samskara do not first arise as a result of bodily functioning and then produce Vijnana, for it is expressly said that these cannot exist independent of consciousness. 1 Rather are they accessories, even though necessary accessories of consciousness which arise simultaneously with it. Since they are thus but concomitant phases dependent upon Vijnana, which arises directly from the operations of the sense objects and sense organs, the later Buddhists grouped them together as Caitasikas or " mentals ", and the Sthaviravadins and Sarvastivadins gave them, in their external classification of the universe, the following invariable order :

1. Rupa.
2. Citta or Vijnana.
3. Caitasikas.
1 A.K. 4-3a ; M.Sh., p. 173.
(a) Vedana.
(6) Samjna.
(c) The remaining Samskaras.

While, however, the Caitasikas are mental properties owing to their existence to Citta or Vijnana, and hence cannot arise before the latter, yet, on the other hand, pure Citta devoid of all of these Caitasikas cannot arise. So that it is said that the origination of Citta and the Caitasikas is simultaneous.

A word must be said concerning the traditional order of the skandhas. In the sutras they are invariably enumerated in the order given above. The philosophers might see fit, as did Buddhaghosa and Nagarjuna, to depart from this order when dealing with them philosophically, but as regards mere enumeration they held fast to the traditional order. Never theless, in the sutras themselves no reasons are ever given for this order, but in the Abhidharma Kosa we find an attempt made to defend the time-honoured enumeration (A.K. 1-176).
" The order of the skandhas is given in accordance with their relative coarseness and impurity with reference to the plate simile and in accord with the enumeration of the dhatus. Rupa being impenetrable, it is the coarsest of the skandhas. Among the non-material skandhas Vedana is the coarsest, just as everyone speaks of his hand, etc., being painful. Samjna or ideation ... is the next in coarseness, since the notions, man, woman, etc., are easy to understand. Samskara is coarser than Vijnana, as the forces of hate, etc., are easy to understand. Vijnana is the finest of them all, and since it universally grasps the characteristics of all other objects it is difficult to understand. Thus the order of the skandhas is in accord with their relative coarseness.

" From the very beginning until now men and women have been mutually attracted by one another s rupa. This arises from abandoning oneself to passionate feelings (Vedana). Giving oneself over to pleasure is due to erroneous ideation (Samjna). This erroneous ideation is dependent upon the Klesas (Samskaras). The Klesas arise in dependence upon Vijnana, and on the other hand the first three defile Vijnana once they have arisen. Thus the order of the skandhas is in accord with their mutual defilement.

" Again, rupa is like a plate. Vedana is like food or drink contained in the plate. Samjiia is like a sauce, Samskara is like the cook, and Vijnana is like the eater. Thus the order of the skandhas is in accord with the simile of the plate, etc.
" Again, the order of the skandhas is in accord with the relative position of the dhatus (realms of existence) ; e.g. in the Kama dhatu there are various subtle forms of desire where the rupa characteristics are made manifest. In the Dhyanas of the Rupa dhatu one finds sukha and priti, etc., where Vedana is particularly prominent. In the three first Arupa heavens they grasp the idea of space, etc., so that here Samjiia is the most prominent. In the highest heaven Cetana is the dominating factor, so that this realm is characterized by Samskara. All these four are supports of Vijnana."

2. The Twelve Ayatanas and the Eighteen Dhatus

The other categories of the subjective classification of existence consist of the twelve ayatanas and the eighteen dhatus. These had best be considered together. Both are categories which were formulated not from an analysis of the human personality nor from an objective analysis of the external world, but as the result of the investigation of the functions of consciousness, and the means whereby con sciousness is produced. Here all the component parts of being are grouped together with reference to the part they play as consciousness producers, for Buddhism starts with the assumption that consciousness is not an eternal self- existent thing but is the temporary product of certain pre-existent material factors.
Ayatana, says Mrs. Khys Davids, means " Place or sphere of meeting, or of origin or the ground of happening ". 1
1 B.Psy. sect, on Ayat.

Vasubandhu (A.K. l-15a) renders it the gate of production of the Citta and Caitasika dharmas. More freely we can explain the term as being the basis of consciousness, or the factors which bring about consciousness. In this category the various aspects of consciousness themselves are not included.
The Ayatanas are twelve in number, and are as follows :

1. Object of sight
2. Object of hearing
3. Object of smell
4. Object of taste
5. Object of touch
6. Object of thought .
7. Organ of sight
8. Organ of hearing
9. Organ of smell
10. Organ of taste
11. Organ of touch
12. Organ of thought

- Sense object.

- Sense organ.

Dhdtu, like dharma, is denned as that which bears its own attributes, but Vasubandhu (op. cit.) says that the term means species or genus, or even element, just as one says that a mountain consists of certain elements : gold, silver, copper, etc. In like manner does the world consist of so many dhatus. Consequently, we may call the dhatus the factors of con sciousness, or more correctly the elements of existence, regarded from the standpoint of consciousness and its causes, since the dhatu category contain all the twelve ayatanas, and in addition the six major divisions of consciousness itself, making eighteen in all. These are :

1. Sight object

2. Sound object
3. Smell object
4. Taste object
5. Touch object
6. Mental object J

Sense object.

7. Sight organ
8. Sound organ
9. Smell organ
n A m \- Sense organ.
10. Taste organ
11. Touch organ
12. Mental organ
13. Consciousness dependent upon sight
14. Consciousness dependent upon sound

15. Consciousness dependent upon smell
16. Consciousness dependent upon taste

Kecipient consciousness.

17. Consciousness dependent upon touch
18. Consciousness dependent upon mentation,

A few words concerning each of these factors will not be out of place.
1. The Fifteen Sensuous Factors, consisting of the five sense objects, the five sense organs, and the five-fold sense perceiving aspects of consciousness. These are not so likely to be misunderstood, but to each term a somewhat peculiar interpretation was given.

(a) The Five Sense Objects

These consist of visible objects, audible objects, etc. Consider carefully the word object. The substantialist would say that there is but one substance, which is perceived in different ways by the five senses, is seen by the eye, is touched by the hand, etc. Not so for the Buddhist, particularly after the atomic theory had been accepted. 1 Every material object consists of molecules (Samghata paramanu, or Kalapa). Each molecule contains at least one visual atom (dravya paramanu), i.e. one atom which affects the eye and no other sense organ, one taste atom which affects the tongue and no other sense organ, etc. Consequently, the sense organs receive in reality impressions from different objects, even though these objects or atoms all form part of a single molecule, or atomic group. Concerning the nature of these atoms and
1 Cf. A.K. 4-laff., where all the following points are discussed ; also f. 30. molecules, and the way in which they are produced and destroyed, we shall speak more at length hereafter. Suffice it for the moment to say that these atoms are not eternal, but are derived from the four elements. The four elements are themselves atomic, are cognizable by touch alone and not by any other sense organ, though their existence may be inferred by the reason acting upon the data given by the other senses.
 
The fact that the object of sight must be a different substance from the object of taste, etc., is really inherent in the word dhatu, element, or factor, for if the object of sight and the object of taste were really one and the same they would constitute one dhatu and not two.

(b) The Sense Organs

In the same way the sense organs, according to the Buddhists, are not what we usually mean by the term. The Caksur- indriya is not the eyeball, but certain atoms of a peculiar kind scattered over the ball of the eye and possessed of the faculty of vision. The nose sense-organ consists of specific kinds of atoms scattered inside the two nostrils and possessing the faculty of smell, and so on with the others. The organ of touch, for example, consists of a large number of atoms scattered throughout the body, and possessed of the faculty of (tactual) feeling.

These five kinds of atoms are quite distinct in kind from the five sense-object atoms, though they are equally derived from the four elements. 1 Those parts of the body which do not form part of the sense organs consist of molecules or atom groups of the five sense atoms, with their attendant element atoms.
The sense organs in the ordinary sense of the word, i.e. the sense orifices, are sometimes called the auxiliary indriya or faculties. 2 Incidentally, the double nature of the orifices of the eye, ear, and nose called for comment from the Buddhist
1 Ab. Hr., l-2a. 2 A.K., A.K. 2, last part.
H

philosophers, and we find them stating that the number of the dhatus is not thereby increased because (A.K. 1-136) " though twofold, yet as their species, their sense-object, and their resultant consciousness is the same, their nature is one ". When, further, they paused to consider why these particular orifices were two-fold, the older explanation was that these orifices were made in pairs for the sake of adornment, as with but a single orifice sentient beings would appear hideous. But the philosophic Vasubandhu objected to this explanation, and argued that their two-fold nature (A.K. 1-14&) was due to more practical reasons, viz., that we may have a clearer or more exact impression of the external object. When we look at an object with one eye closed, we are unable to see it clearly.

(c) Sense Perceiving Aspects of Consciousness

The five aspects of consciousness concerned only with the cognition of sense data arise from the interaction of sense object and sense organ, or more accurately the sense-object atoms and sense-organ atoms. The transient nature of consciousness, and incidentally of the sense objects and sense organs, is seen from the following citation from the Samyutta Nikaya : " Consciousness comes into being, monks, because of two things. Because of the sight organ and visual object there comes into being visual consciousness. The organ of sight is impermanent, fleeting, constantly changing. Visible objects are the same. Visual consciousness arising from such a conditioning relation which is impermanent, fleeting, constantly changing, is itself no less so, for how can it be that consciousness arising from such an impermanent relation can itself be permanent " (S.N. iv, 67).
Two further details call for mention. Later Buddhists, when dealing with sense perception, laid great emphasis on the distinction between individual perception (svabhava- laksana) and aggregate perception (samjnana-laksana). The sixth aspect of consciousness, mano-vijnana, was considered to have aggregate perception, or to be cognisant of objects, common to all the same organs. But the first five vijnanas were limited in their scope to their own field of action, they had only an individuate cognizance of the homogeneous object immediately in front of them, and did not encroach upon other fields.

This individuality of perception, however, only referred to each vijnana cleaving to its own ayatana. Inside of its own sense group, every vijnana and aggregate perception as may be seen from the following discussion in the Abhidharma Kosa (l-8a). Objects of touch were divided into eleven divisions. Some teachers thought that at any one time tactual consciousness could only perceive one derivative and four elemental groups, owing to the doctrine of svabhava-laksana, but Vasubandhu says that tactual con sciousness can perceive all eleven simultaneously, because, all being within one ayatana, they did not constitute a case of samjnana-laksana.
Again, Buddhism asserted that Vijnana has three kinds of cognizance or Samkalpa. 1 One, Svabhava samkalpa, is immediate sense perception of the object presented to it ; the second, Smrti samkalpa, is memory or the cognizance of present sense data associated with the memory of former similar and dissimilar sense impressions. The third, Samprayoga samkalpa, may be called reasoning or intellection based upon sense data. Of these three functions of consciousness, Mano- vijnana possesses all three. The first five vijnana are possessed of Svabhava samkalpa alone.

2. The Five Non-sensuous Dhdtus present some of the most interesting problems of Buddhist philosophy, though for the present we can but consider their bare outline, (a) The Sense Object
The mind proper or mental cognition is considered a sense in the same way as the first five vijnana. Consequently it,
1 I have not found this doctrine in Pali tradition, but it io common to both Sarv. and Yog. schools, see A.K. 2.

too, is possessed of the three-fold distinction of sense object, sense organ, and sense perceiving aspect of consciousness. The sense object in this case consists of those things which are the subject of mentation. Thus, in so far as we reason or think about Nirvana, Nirvana becomes a sense object, or the object of Mano-vijnana. The formulation of the group of factors constituting this dhatu was obviously subsequent to the Sutra period, for we find some difference of detail among the various schools.
The Sthaviravadin l enumeration is as follows :

1. Nirvana, the highest reality.
2. The subtle forms of Bupa matter which cannot be
perceived by the physical sense organs, or their associated vijnana.
3. Citta, or mind itself.
4. Caitasikas, or mental properties.
5. Prajnapti, or concepts or notions.
The Sarvastivadin 2 in enumeration was :
1. The Vedana, Samjna, and Samskara skandhas, with all
their numerous subdivisions.
2. Avijnapti Rupa, or unmanifested rupa.
3. The three Asamskrta dharmas, or the transcendental
and permanent factors of existence, making seven divisions in all. The Yogacarin 3 enumeration was :
1. The Caitasika dharmas.
2. The Citta-viprayukta dharmas.
3. The six Asamskrta dharmas.
4. The subtle forms of matter. These four categories
when subdivided make eighty-two dharmas in all. At first sight there appears to be a great deal of divergence of opinion concerning mental object among the three schools
1 C.R, p. 3 and p. 120.
2 A.K. 1-10.
3 A.V.P. l-lo, see disc. * ft $C

which we are considering. But a good deal of the difference is more apparent than real, as will be seen when we come to deal with each of the above-mentioned groups. It will be better, however, to postpone all discussion of the matter until we come to examine them from the point of view of objective analysis.

(b) Sense Organ

In many of the later Sthaviravadin commentaries it is said that the sense organ of the Mano-vijnana is the physical heart, which takes the place of the brain in most Oriental calculations. This, however, is not mentioned in the Pitakas themselves, and in their enumerations of the twelve ayatanas and the eighteen dhatus they agree with the other schools of Buddhism in making the manas (or mind) an abstract entity the organ of Mano-vijnana. What is this curious thing called the mind or manas which is thus contrasted with mental cognition, or Mano-vijnana ? At present I am aware of no Pali text dealing with the matter, but Vasubandhu tells us (A.K. l-12a) that the " flux of the six vijnanas constitutes Manas ", and, again, (b) " the continuous passing away of the six vijnanas forms manas . . . The continuous passing away of the group of six vijnanas, causing later vijnanas to arise in their place, is called Mano-dhatu, just as the child of this (man) is called the father of that (man), or as the fruit of this tree is called the seed of that tree." In other words, each of the six vijnanas has only a momentary existence. Never theless, there is a Karmic or causal affinity between the various groups of consciousness of one moment and the next. The group of this moment inherits the tendencies, etc., of the immediately preceding group, and as the chief function of Mano- vijnanas is memory and reason, both inseparately con nected with the continuity of the mental process, it is said that the constantly dying away vijnanas of the past moment constitutes the base or organ for the activity of the Mano- vijnana of the present moment. Just as the activity of the caksur or indriya brings about the arising of Caksur- vijnana or the visual consciousness, so does the transmitted energy of all the immediately preceding vijiianas bring about the arising of the Mano-vijnana.

(c) The Sense-perceiving Consciousness

At the present moment little need be said about the nature of the Mano-vijnana. We have already said that this aspect of consciousness possesses aggregate as well as individual perception, as opposed to the other five types of conscious ness, and, moreover, while the latter has only svabhava samkalpa, the former has samprayoga samkalpa, and smrti samkalpa as well. Nor must it be forgotten that the Yogacarins added two other types of consciousness, which shared some of the attributes and functions of the old Mano- vijnana.
The Buddhists laid great stress upon their division of the universe into the eighteen dhatus or elements, and used these categories to explain many of the functions of life, and just as the Vaisesikas, after enumerating their list of dravya or substances, proceeded to expound the list of gunas or qualities or attributes possessed by each of these substances, so did the Buddhists give a good deal of attention to the characteristics of each of the eighteen dhatus, even though the non-substantialist position of Buddhism made it impossible for them to carry out the theory of inherent attributes possessed by self-existent substances. Consequently, in place of a list of fundamental inherent gunas, the Buddhists, after enumerating their list of dhatus, merely attempted to place them in different logical groups ; e.g. those which are visible, and those which are invisible ; those which are denied (sasrava), and those which are not defiled (anasrava) ; those which are objective and those which are subjective, etc. For the most part, however, these classifications were of little philosophic import, so that we may rest content with referring the curious to the original authorities.

(B) THE OBJECTIVE CLASSIFICATION

We now come to the consideration of the component parts of the universe classified and analysed from the objective standpoint. Here no longer does the human personality or the process of consciousness serve as the starting-point. Rather are the phenomena of life, both mental and physical, considered universally.
Here, also, do the component parts of the universe receive a new designation. It is curious to note how many terms there are which we are forced to lender " elements " or " factors of existence". The difference in terminology has almost exclusive reference to the standpoint from which the analysis of the universe was made. A word must now be said con cerning their relationship.

The skandhas or component parts of the human personality are for the most part aggregates or compounds and not them selves ultimate or simple factors of existence. Vedana and Samjna, to be sure, are treated as ultimates. The Sarvastivadins state, moreover, that all the various classes of vijnana constitute but one ultimate factor, but the other schools consider even the vijnana skandha as composite. Rupa and Samskara are certainly names for groups of elements and not true ultimate factors themselves.
The dhatus come nearer to a scientific conception of ultimate factors, since all but one (dharma dhatu) are discrete, simple, ultimate elements. Dharma dhatu, however, is a generic term, and includes a great many discrete factors.

With the enumeration of the dharmas, however, an attempt (even though not altogether successful) was made to enumerate those factors which are themselves the underlying units of the other groups, and to them, therefore, may we properly assign the term element or ultimate factor. The word factor rather than element has been chosen, inasmuch as the latter term has been usurped by the four Mahabhutas. The dharmas include not only these elements, but apply also to the mental and other spheres.

The dliarma classification came to have more and more importance assigned to it in the subsequent developments of Buddhist philosophy, and finally usurped most of the attention previously given to the skandhas and dhatus. It is, therefore, imperative that careful investigation be made both of the meaning of the term and the objects denoted by it.
The Meaning of Dharma.

Dharma has been used in a great variety of senses. It has been well called the blank cheque of Indian thought. In the present instance, however, it has no relation to the more usual significance of truth, law, religion, duty, etc. Along with dhatu, and for similar reasons, it has been defined as " that which bears its own attributes ", meaning thereby ultimate entities possessed of their own characteristics, as opposed to groups of phenomena whose characteristics are derived from ultimate substances which underlie them. Being thus con trasted with phenomenal groups and constituting the factors of which such groups are composed, Mrs. Rhys Davids rendering of the term by " phenomena " (in her trans, of the Dham. Sang.) is by no means happy, while her " mental states " is still less so, since many of the dharmas are not mental (as she herself admits).

On the other hand, " ultimate factor " and " element " are equally liable to be misunderstood, for the doctrine of the dharmas never conflicts with the anti-substantialist position of Buddhism. Only the Asamskrta dharmas (Nirvana, etc.), which are purely transcendental, are uncaused and underived. The other dharmas are not like the eternal substances of the Jainas and Vaisesikas, which are uncaused, eternal, and possessed of certain inherent attributes which remain unchanged even though all of their manifested attributes undergo transformation. Every single one of the Samskrta dharmas or factors of phenomenal life are impermanent, caused, conditioned. Thus, for example, the atoms of the five sense organs and the five sense objects (each sense organ and sense object is counted a dharma) are derived from the four elements, are in continuous dependence upon them, and are frequently destroyed and recreated. Moreover, the Buddhists do not like to admit that the dharmas are invisible substances (things in themselves) possessed of a number of inherent attributes which alone appear to us. That thing which we see is the sense object, and not merely some one aspect inherently existing in a self -constituted substance. It was probably for this reason that the name dharma (bearer of its own attribute) was given to it. As has been very wisely pointed out by Mrs. Rhys Davids, " This, to us, very obscure characterization may very likely mean that dharma, as phenomenon, is without substratum, and is not a quality cohering in a substance." 1

Dharmas themselves are phenomena in the sense that they are not eternal substances, but they are not phenomena in the usual Western sense of modes of other underlying substances. The sense organs are derived from and dependent upon the four Mahabhutas, but they are separate entities possessing their own characteristics, and not aspects, attributes, phases or modes of the Mahabhutas.

They are ultimate in the sense that they are simple and not composed of finer heterogeneous units. Thus, for example, though the sense organs may be composed of atoms, every caksur indriya paramanu is uniform and homogeneous, so that all such paramanus together constitute but one dharma or factor of existence. They are thus ultimate when con trasted with the physical body, let us say, which consists of sense organ atoms, and sense object atoms, as well as the four Mahabhutas themselves.
The propriety of the term " factor " as a rendering of dharma is also to be seen from two further considerations. One is the so-called permanence of the dharmas, and the other is the list of things which are called dharmas.

1 Dh. San. Intro., p. xxxiv (trans.).

First, as regards their permanence ( -0 J *$ ^)- Supposing that no other form of organic existence were possible other than that which we see around us to-day, we could say that the eye is a permanent factor in human life, even though every single eye is necessarily impermanent, and subject to decay. In the same way the various dharmas are permanent factors in life, even though every one of them is constantly being destroyed and recreated.

Second, as regards the curious list of dharmas which both the Sthaviravadins and Sarvastivadins enumerate. Both contain certain things which can be called neither mental nor material, though many of these are inconsistently classed by the Sthaviravadins under rupa skandha, while the Sarvastivadins enumerate them, more logically, as a separate group unconnected with either matter or mind. Among these are homogeneity, decay, vitality, birth. The Sarvastivadins say that these are not merely modes of other substances, but are themselves separate dharmas or ultimate factors of life.

From the foregoing we may see that the conception of " dharma " is one of the most original contributions of Buddhism to the history of human thought, much in the same way as Plato s " Ideas " and Aristotle s " Forms ", or the doctrine of substance. Many Buddhist ideas are derived from, or at least shared with other systems of thought, but the doctrine of dharma has no exact parallel with any other conception, so that it is most curious that no one has as yet dealt with the curious philosophic position which the doctrine involves.
Classification of the Dharmas.

Having thus examined the meaning of the word dharma, let us now turn to the enumeration of the dharmas themselves. First as regards their grouping.
In the Abhidhammattha - sangaha, 1 representing the Sthaviravadin tradition, we find the following arrangement :
1 C.P., p. 81.
1. Citta, or mind, consciousness.
2. Caitasikas, or mental, properties.
3. Rupa, or matter.
4. Nirvana, or the highest reality, sometimes denned as Asamskrta.

The Sarvastivadin classification (A.K. 4-la) is l :
1. Rupa, or matter.
2. Citta, or mind.
3. Caitasikas, or mental properties.
4. Citta- viprayukta, or miscellaneous factors.
5. Asamskrta, or unconditioned factors, one of which is Nirvana.

The Yogacarin 2 classification is :

1. Citta, or mind.
2. Caitasikas, or mental properties.
3. Rupa, or matter.
4. Citta-viprayukta, or miscellaneous factors.
5. Asamskrta, or unconditioned factors.

All three schools agree in calling those dharmas which are not definitely Asamskrta, Samskrta, so that in reality there are but two great categories : (1) The Asamskrta or Eternal, unconditioned elements of existence which do not enter into combination ; and (2) Samskrta, conditioned and impermanent elements which enter into combinations to form the phenomenal world around us. This second group, as we have seen, the Sthaviravadins divide into but three categories, while both the Sarvastivadins and the Yogacarins postulate four. This fourth category, Citta-viprayukta, means, literally, not connected with mind, but it is also defined as being equally disassociated with matter or rupa. This category consists of certain dharmas which are somewhat incongruously placed under rupa by the Southern Buddhists,

1 Also Pr. Pad., f. 1-la.
2 A.V.P. 1-la, also Sat. D. la.

together with, certain dharmas which are not found in the Southern list at all.

Although there is some disagreement as to the exact number of dharmas included under each category, all the schools agree that rupa dharmas consist of the component units of the Rupa skandha ; Citta dharmas of the component parts of the Vijnana skandha ; while the Caitasikas consist of the Vedana and Samjna skandhas, together with the component parts of the Samskara skandha. According to the Sarvastivadins and Yogacarins, the Citta- viprayukta dharmas are also included in the Samskara skandha.
There is also comparative unity of opinion concerning the relationship between the dharmas on the one hand and the dhatus and ayatanas on the other. The five sensuous sense organs and sense objects belong to the Rupa dharmas. The five sensuous forms of consciousness belong to the Citta dharmas. To this category also belongs the sixth or Mano- vijnana, and Mano-dhatu, while the remaining dhatu or ayatana, the object of mentation, includes all the other dharmas.

The accompanying chart will serve to make clear the relationship supposed by all the schools to exist between the skandhas, the ayatanas, the dhatus, and the dharmas.
In the earlier stages of all three schools there seems to be a good deal of deviation and inconsistency in the enumeration of the dharmas. In this period no exact numerical definition seems to have been given, but in the later stages of each school an attempt was made to fix artificially the number of each group as well as the total number of dharmas. 1
Thus in the Sthaviravadin school, Rupa consists of twenty- seven or twenty-eight, Citta of eighty-nine, Caitasikas of fifty-two dharmas, while Nirvana consists of an additional dharma, but no especial mention seems to be made of the total number of dharmas taken as a whole.
1 For author, see list above.

Among the Sarvastivadins subsequent to the Abhidharma Kosa, the total number of dharmas was fixed at seventy-five, three being allotted to the Asamskrta dharmas, eleven to Rupa, one to Citta, forty-six to Caitasika, and fourteen to Citta- viprayukta .

Among the Yogacarins subsequent to the Vijnana-matra Siddhi, the total number of dharmas was fixed at a hundred, six being allotted to Asamskrta, eleven to Rupa, eight to Citta, fifty-one to Caitasika, and twenty-four to Citta- viprayukta.
This gives us the following comparative tables :
Category. Sthavir. Sarvdst. Yogdc.
 
Asamskrta ..13 6
Rupa ... 28 11 11
Citta ... 89 1 8
Caitasika 52 46 51
C.-viprayukta . 14 24
Total . 170 75 100
1. The Asamskrta Dharmas

All forms of Buddhism agree in dividing the dharmas into those which are Asamskrta and those which are Samskrta. Samskrta comes from the root Sam = with ; and krta, done or performed. Accordingly, the word is literally the same as the Latin confectus. The Pali commentaries take samskrta to mean conditioned, implying that the Samskrta dharmas are caused, conditioned, mundane, temporal, impermanent, non-eternal, active, and associated with the Asravas or taints. As opposed to these are the Asamskrta dharmas, which are not subject to cause, condition, or dependence, and are therefore transcendental, out of time, unchanging, eternal, inactive, and free from the Asravas. Mrs. Rhys Davids, in an appendix to her translation of the Dharma Sangani, has collected a list of adjectives applied to the Asamskrta dharmas. These may be studied with advantage.

1. The Sthaviravadins, 1 however, know of but one Asamskrta dharma, Nirvana. Nirvana is to be gained by the annihilation of the roots which lead to rebirth, and may, therefore, be called the cessation of phenomenal life, though a person who attains to Nirvana continues in bodily existence until his span of life is exhausted, after which the phenomenal dharmas which compose his personality disintegrate, having no further creative force to keep them together. Nirvana, then, is of two kinds : (1) Nirvana with a subsidium, or Nirvana associated with a still existing personality ; and (2) Nirvana without a subsidium, or the state of Nirvana after the phenomenal personality has disintegrated. The fact that Nirvana is called a dharma shows that it was considered a positive concept and not a mere negation of life.

2. The Sarvastivadins enumerate three Asamskrta dharmas. These are : (1) Akasa, (2) Pratisamkhya Nirodha, and (3) Apratisamkhya Nirodha.
(a) Akasa corresponds to what in the West is called either space or ether. Vasubandhu says (A.K. 1-36) it has for its characteristic non-impeding, and since it offers no obstacle, matter (rupa) freely functions therein. The fact that Akasa is always considered a substance and an eternal and unchanging unity shows that Akasa is not merely empty space, or lack of matter, but a positive entity having many attributes common to the old idea of ether.

An important point to notice in this connexion is that the Sthaviravadins give Akasa among their list of derived material dharmas produced by the four Mahabhutas, while with the Sarvastivadins it is elevated to the rank of an Asamskrta dharma. In point of fact, however, the Sarvastivadins seem to distinguish between two kinds of Akasa, the first the eternal and omnipresent ether, the other to empty space, with which is frequently associated the sky. The first is translated by the Chinese as j Q (hsii k ung),
1 See C.P., p. 168.

and the second by ! (k ung) alone. The second concept agrees more or less with the Sthaviravadin conception of Akasa. Thus the Mahavibhasa (75-96) : " Hsii-k ung is not rupa, while k ung is rupa (i.e. lack of rupa, or interstices between rupa). Hsii-k ung is invisible, k ung is visible. Hsii-k ung is Anasrava, k ung is sasrava. Hsii-k ung is asamskrta, while k ung is samskrta." In each case the definition of k ung corresponds to the Sthaviravadin Akasa.
 
In another passage Vasubandhu (A.K. 1-216) brings out the strong difference between the Asamskrta Akasa and the Akasa, which is merely empty space. " K ung is an interstice (or hole) . . . The apertures in a gate (as regards inanimate objects) or in the mouth or nose (as regards animate objects) are called k ung."
From this it would seem clear that the Sarvastivadins regarded the Asamskrta Akasa as ether, and the Samskrta Akasa as space.

(b) Pratisamkhya Nirodha is conscious cessation. Vasubandhu (A.K. 1-36) says : "It has for its nature freedom from bondage . . . Pratisamkhya means conscious deliberation, and is a type of intelligence since it deliberates upon each of the four Noble Truths. The attainment of cessation (Nirodha, i.e. the cessation of the taints and passions) by means of the power of deliberation is therefore called Pratisamkhya Nirodha, just as a cart pulled by bullocks is called a bullock- cart by the elimination of the middle term." Vasubandhu then goes on to discuss whether this Pratisamkhya Nirodha is uniform and homogeneous, i.e. whether the attainment of freedom from one bond implies simultaneous emancipation from all others, and he answers (op. cit.) : " By no means, for it differs according to the basis of the bondage ; that is to say, the emancipation from bondage is in accord with the extent of basis of the bondage. If this were not so, when one annihilates one klesa ... all other klesas would thereby be destroyed, and, consequently, training in order to master the others would be unnecessary " (which is false).

Nevertheless, although emancipation from the klesas must be accomplished individually, yet this process leads to complete freedom from all bonds in the end, and so Nirvana, or complete freedom from all bonds by means of conscious effort, comes within this category, and thus Pratisamkhya Nirodha is identified with the only Asamskrta dharma known to the Sthaviravadins.

(c) Apratisamkhya Nirodha, in contrast with the preceding, is " unconscious cessation ", and is explained as the non- arising of consciousness, not as the result of conscious effort, but by lack of the necessary sufficing conditions. Vasubandhu (A.K. l-14a) says : " Just as when the mind is intent upon one object, all other objects of sight, sound, smell, taste, and touch are lost, because the group of the five vijnanas remains in the not-yet-arisen (or future) state and consequently never arise at a later moment since they are unable to perceive past-sense objects. In like manner, because of the incom pleteness of sufficing causes, Apratisamkhya Nirodha is brought about."
Each of the three Asamskrta dharmas, according to the Sarvastivadins, is separate and discrete. Consequently, they are pluralists, inasmuch as they accept more than one eternal unconditioned element.

3. The Yogacarins. In contrast to the Sarvastivadins, the adherents of the Yogacarin school enumerate six Asamskrta dharmas. 1 These are : (1) Akasa, (2) Pratisamkhya Nirodha, (3) Apratisamkhya Nirodha, (4) Acala, (5) Samjna-Vedana Nirodha, (6) Tathata. The first three are the same as the Asamskrta dharmas of the Sarvastivadins, and, therefore, need not be discussed. The additional thiee are :
(a) Acala is, literally, immovability. This is a mental state identified with Upeksa, or indifference, or the trans-

1 Sat. dh. dv., p. 2, V.M.S. f. 1, G.V.P., and A. Sam. San. enumerate eight Asainsk. by dividing Tathata into three classes.

cending of both pleasure and pain, and is associated with the third and fourth Dhyanas of the Kupa dhatu ; it is also associated with the mental states of the first three divisions of the Arupa dhatu.

(6) Samjna-Vedana Nirodha. This is the state of trance in which both samjfia (ideation) and vedana (feeling) cease (nirodha). At this point the aspirant enters into the Nirodha samapatti. This stage is associated with the mental condition prevailing in the fourth Arupa heaven.
(c) Tathata, literally suchness, or the true nature, or the Absolute, which, according to the Yogacarins, is the ultimate reality underlying all phenomena. " This is the ultimate essence of everything, and it is termed Bhutatathata or permanent reality, because it is both real and eternal, though its true nature cannot be grasped by words or ordinary conceptions."

The addition of this dharma made a complete revolution in Buddhist philosophy, particularly as compared to the ontology of the Sthaviravadin and Sarvastivadin schools, for, whereas according to these latter systems each of the dharmas is eternally distinct, according to the Yogacarins all dharmas are but modes of the one fundamental " essence of mind ".
Consequent upon this idea, the Yogacarins go on to say that the enumeration of the preceding five Asamskrta dharmas belongs to the realm of relative truth, since in reality there is but one Asamskrta, namely the last or Tathata, while the others are but different ways of conceiving it.
The Absolute, as the ever-present, non-impeding reality, is called Akasa ; as free from the limitations of the phenomenal world (of which, however, it is the underlying reality) it is called " Pratisamkhya Nirodha ", especially when this state is regarded as the result of conscious effort. The term " Apratisamkhya Nirodha " is also used to show that the Absolute manifests itself when the conditions which obscure it are absent, that its existence continues even when other conditions bringing about phenomenal existence do not arise. The fourth and fifth Asamskrta dharmas are but inferior stages leading to noumenal existence.

2. Rupa, or Matter l
Having finished our survey o
f the Asamskrta dharmas we now turn to the Samskrta dharmas, or the ultimate factors, which enter into combination in the phenomenal world ; and of which they are the proximate cause. Consequently, they are defined (A.K. l-4b) as " mundane, temporal (i.e. belonging to the world of time), possessed of causes and conditions, phenomenal, capable of being described by words, tainted (sasrava), and associated with the two truths Duhkha and Samudaya ". Since these are the ultimate parts of phenomenal life, and such life consists of matter and mind (including the phases of the mind), such dharmas are classified into Rupa, Citta, and Caitasika dharmas, to which other schools would add Citta- viprayukta dharmas.
Let us first consider the Rupa dharmas, or the material factors of existence. All schools of Buddhism agree that every material dharma falls into one or other of two categories : (1) The ultimate or underived, and (2) the derived. They are generally agreed that the ultimate dharmas are four in number and consist of the four Mahabhutas, or four material elements, which are earth, water, fire, and air, though each of these terms has a very special significance for the Buddhists.
Regarding the enumeration of the derived material dharmas, there is not quite such uniformity, though all lists include the five sense objects and the five sense organs, however many other dharmas may be given.

1 For Str. p. 3 of Dh. San. ( Book 2 of Rt. Trans.), C.P. partvi, p. 154 ff. For Sarv. Wu-shih-p i-p o sha lun, first half of first fasc., A.K., first half of first fasc. ; for Pr. and M.V. fifth grantha. For Yog. A.V.P. 1-36 and 4a, A.S.S., p. 2a.

(a) The Ultimate Material Elements

These, the four Mahabhutas, require, as we have said, especial attention, as they are easily misunderstood. The use of the terms in the Hinayana sutras is somewhat vague, so that we are left uncertain as to what interpretation was given to them in primitive times, but all the later schools agree that the elements are themselves only perceptible through touch. What we see, hear, smell, or taste, are only sense objects derived from the elements. Even as regards the faculty of touch, certain of the things sensed are only derivative sense objects, though, in addition, the touch sense organ can perceive the elements themselves.
The Abhidharma Kosa (l-9a, where a detailed discussion will be found) gives us the following chart of the attributes and functions of each of the four elements :

Name. Attribute. Function.
1. Earth Solidity Supporting
2. Water Moisture Cohesion
3. Fire Heat Ripening
4. Air Motion Expanding

Every molecule of every material object contains at least one atom of each of these four elements. Thus a molecule of what we call water contains atoms of earth, water, fire, and air, in addition to the atoms of the sense objects, but it so happens that for the time being the water element has a predominant influence, so that we call the molecule in question water. The Abhidharma Kosa (4-16) has the following state ment concerning the matter. " Query : If the four Mahabhutas always arise together, in all molecules, whether hard or moist, hot or mobile, how is it that we can sense one and not the others ? Answer : Because for the time being the nature of one happens to be predominant, so that our senses notice one and not the others." The Mahavibhasa is even more emphatic on this point, and in one passage


(31-156) gives long arguments to support the Buddhist position.
In this connexion one doctrine peculiar to the Sthaviravadin school must be noticed. They believe that only three of the Mahabhutas are tangible, i.e. that the water element cannot be felt, and, as no element is perceptible to the other senses, it follows that the very existence of this element can only be known by a process of inference. Thus we find S. Z. Aung (C.P. 155%) saying : " Particles of matter are held together by Apo (cohesion) (water) which cannot be felt by the sense of touch e.g. when one puts his hand into cold water, the softness of the water is not apo but pathavi (earth), the cold felt is not apo but tejo (fire), the pressure felt is not apo but vayo (air) . . . From this one can easily see that the Buddhists are not dealing with Thales water, Anaximenes air, Heraldeitus fire, or the Peripatetics matter of Greek philosophy."
For this reason we should prefer to call the mahabhutas forces rather than material elements, except for the fact that they are considered atomic and therefore are obviously material.

What is the connexion between the four ultimates and the various derivative forms of matter ? Buddhaghosa, repre senting the Sthaviravadins, merely says that a derivative form of matter is aided by the four elements supporting (earth), binding (water), maturing (fire), and moving (air) it, like an infant prince being fed, bathed, dressed, and fanned. (Att. Sal. quoted Dham. Sang. K. D. trans., p. 174.)
The Sarvastivadins and the Yogacarins deal more explicitly with the causal relation between them, giving a five-fold relationship, one for each of the mahabhutas taken separately, and one for them altogether. The Mahavibhasa (127-66) gives us the following chart :
1. Producing, as a mother gives birth to a child all four.
2. Reliance, as a disciple relies upon his teacher fire.

3. Establishing, as the earth holds or sustains
things . . . . . . . earth.
4. Sustaining, as food supports food . .< water.
5. Nourishing, as water fertilizes the roots of
trees ....... air.

The Abh. Sam. San. (l-2a), representing the Yogacarin school, defines each of these causal relations in the following way :
1. Producing. This is the originating cause, because the derivative dharmas cannot arise apart from the activity of the Mahabhutas.
2. Reliance. This is the transforming cause, because, apart from the Mahabhutas, the derivatives have no intrinsic qualities of their own.
3. Establishing. This is the co-ordinate transforming cause, because whenever the Mahabhutas undergo modification the derivatives undergo a like transformation.
4. Sustaining. This is the cause of continued existence, because, depending upon the mahabhutas, the derivatives do not cease, but have a continuous development.
5. Nourishing. This is the cause of growth, because increase (or stimulus) takes place owing to the nourishing force of the Mahabhutas.
(b) The Derivative Dharmas

The Sthaviravadins give twenty-seven or twenty-eight rupa dharmas in all, of which twenty-three or twenty-four are derived from the four Mahabhutas, while the Sarvastivadins and Yogacarins have only eleven. It is probable, however, that the twenty-seven represent an earlier tradition, owing to the greater crudity of the method of enumeration. It seems likely that the more philosophic thinkers of the Sarvastivadin and Yogacarin schools eliminated those dharmas which were inconsistent with a more logical, systematic, and scientific view of the universe.

The following is a comparative list of the derivative material dharmas, as taught in the three schools :

2. Sound
3. Smell
4. Taste
5. Touch
6. Sight object.
7. Sound
8. Smell
9. Taste
10. Touch
11. Unmanifested
matter.

(3) Yogdcdrin.
1. Sight organ.
2. Sound
3. Smell
4. Taste
5. Touch
6. Sight object.
7. Sound
8. SmeU
9. Taste
10. Touch
11. Matter included
under dharma-
dhatu.

(1) Sthaviravadin. (2) Sarvdstivddin.
1. Sight organ. 1. Sight organ.
2. Sound
3. SmeU
4. Taste
5. Touch
6. Sight object.
7. Sound
8. SmeU
9. Taste
10. Touch
11. Female Sex.
12. Male Sex.
13. Food.
14. Organ of Life.
15. Space (akasa).
16. Gesture.
17. Speech.
18. Buoyancy.
19. Elasticity.
20. Power of Adaptation.
21. Power of Aggregation (birth).
22. Duration.
23. Decay.
24. Death.

The curious crudity and iUogicality of the Sthaviravadin school is at once obvious. Here are placed, side by side, material objects and the modes or activities of these organs, such as buoyancy or elasticity, birth, death, etc.
The Sarvastivadins have taken some of the categories out of this list and placed them, more appropriately, among the Citta-viprayukta dharmas. Akasa, in the sense of mere empty space, the apertures of the body, etc., being but the absence of a thing, is not considered a separate dharma, although, as we have seen, Akasa, in the higher sense, is placed among the Asamskrta dharmas. The remaining dharmas of the Sthaviravadin list are ignored, for though masculinity and femininity were classed among the twenty- two indriyas yet, since they were not considered ultimate elements, they were omitted from the lists of dharmas.

It may, at first sight, seem strange to us that even the logical Sarvastivadin and Yogacarin philosophers should have been content to keep the sense organs and sense objects as ultimate factors of existence, but the reasons for their doing so become obvious when the Buddhist theory of matter is taken into consideration. As we have already remarked in dealing with the dhatus, sight objects, smell objects, etc., are for the Buddhists really existing objects, separate from, though produced by the four mahabhutas. Thus, what we consider a single material object, such as a lump of earth, is really something very complex. It contains not only minute particles of all the four elements which are invisible, but also minute visible particles, another set of particles which can be tasted, others which can be smelt, etc., and since, according to the premises, the whole universe of inanimate matter consists of compounds of these ultimate sense particles in addition to the four mahabhutas, it would be inconsistent if they were not counted as separate dharmas.
It is the same way with the sense organs. The particles of the caksur-indriya are quite different from the four mahabhutas, and equally distinct from the five kinds of sense object atoms. Hence they are ultimate factors of existence.

Let us postpone for a moment examination of the eleventh categories of the Sarvastivadins and Yogacarins and deal somewhat more in detail with each of the sense organs and sense objects.

(i) The Sense Organs

We have already learned to distinguish between what the world knows as sense organs and what the Buddhists consider to be the true sense organs. The physical eye is not the true sense organ, though it contains minute percipient particles which constitute the organ of vision. Thus the physical eye is complex. Its molecules contain atoms of the four mahabhutas, and atoms of four out of the five sense objects (sound is generally excluded). It also contains the body organ (kaya indriya) atoms, and finally the atoms of the caksur indriya itself. This idea, common to all forms of Buddhism, is expressed by Buddhaghosa in the following words :

" First the aggregate organism (physical eye) . . . There are fourteen constituents : the four elements, the six attributes dependent upon them . . . vitality, nature, body sensibility (kaya indriya particles), and the visual sentient organ (the true sight sense organ) . . . When the world seeing an obviously extended white object fancies it sees the eye, it only perceives the basis of the eye . . . But that sentient organ which is there bound inherent, derived from the four great principles (mahabhutas) this is the visual sense." (Att. Sal. quoted R.D. trans. Dh. San., 173.)
 
And so it is with the other sense organs, the matter of the five indriyas being quite different from the other kinds of matter. Sometimes the sense organs proper are called the principal indriya, while the sense orifices (e.g. the physical eye) are called the auxiliary indriya.
All accounts agree that the particles constituting the five indriyas are incredibly minute. Concerning their shape and their position we find the following statement :
1. The Eye Indriya. Regarding this the commentators quote the following verse :

" The visual sense by which he beholds forms, Is small and delicate, comparable to the louse s head."
(op. cit.)
Vasubandhu adds that the atoms of the caksur-indriya are diffused over the surface of the eye. f: Just as the particles of flour poured over the surface of the water would scatter on the surface of the water, so do the atoms composing the caksur- indriya spread over the pupil of the eye" (cf. A.K. 2, 19).

2. The Sound Organ. Buddhaghosa says that the units of the sound organ are situated within the cavity of the physical ear, and are well furnished with fine reddish hairs. They are in shape like a little finger-stall (? anguli-thanaka), or a finger covered with rings (op. cit. 178). Vasubandhu, on the other hand, states that they are screwed up just as the bark of a cherry-tree rolls up as soon as it is detached from the trunk (A.K. 2-19a).

3. The Smell Organ. Buddhaghosa says that these particles are inside the nostrils, and in appearance are like a goat s hoofs. Vasubandhu (op. cit. 196) adds that they are like claws whose points face downwards.
4. The Taste Organ. Buddhaghosa says that these particles are above the middle of the tongue, and in shape are like the upper part of the leaf of a lotus. Vasubandhu says they are in shape like a half -moon. It is interesting to note that the tongue and not the palate was considered the basis of the organ of taste.
5. The Touch Organ. Regarding these particles, the commentators merely say that they are scattered all over the body. Vasubandhu states that the number of kaya indriya atoms is equal to those of the body proper, or more correctly that every molecule of the body contains at least one touch organ atom. He further tells us that the shape of these minute particles is not the same for men and women.
In this connexion one or two additional points call for consideration.
1. The later Buddhists distinguish between the eye and ear on one hand, and the nose, tongue, and body on the other. The latter can sense only that which is in immediate contact with them, while the former can sense that which is at a distance. The distinction does not seem to be met with in the original Sthaviravadin Abhidharma works, and among the

Southern Buddhists the distinction is found for the first time in Buddhaghosa. S. Z. Aung is mistaken, however, when he tells us (C.P. 160) that the distinction began with Buddhaghosa, as it is insisted upon by the Sarvastivadins both in the Mahavibhasa and the Abhidharma Kosa (M.V. 13-76 and A.K. 2-7). Vasubandhu tells us, moreover, that the scope of the eye is even greater than that of the ear. It is to be regretted that the Buddhists did not tell us more concerning the nature of the medium between the sense organ and the sense object.
2. The Sarvastivadins draw a distinction between the eye and ear on the one hand and the nose, tongue, and body, on the other, from another point of view. The former are able to sense any number of paramanus at the same time, as when we see either a mountain or a flea (A.K. 2-19a). The latter class of sense organs, however, can at any one moment sense only the same number of objective atoms as their own. If a certain sense object contains a greater number of atoms than the sense organ atoms with which they come into contact, the sense organs only sense at one moment the equivalent number of sense object atoms, and then sense the remaining fraction a moment later, the interval between the two periods being so small and the process of double sensation so rapid that we usually consider the whole affair simultaneous.

(ii) The Sense Objects

We have already seen how the Buddhists, in their rejection of the doctrine of substance, were forced to deny that what we experience are various qualities inherent in an external and unified object, but rather that when e.g. we perceive " red ", there is really a separate atom of redness impingeing on the caksur-indriya, and that this atom is a co-existing but independent atom in a molecule of objective matter. This doctrine was carried to its logical extreme no matter how strange the conclusions may seem to us. Thus, they

claimed that there were atoms of shape as well as colour (though this was denied by the Sautrantikas, A.K. 13-6), and that when we experience with the eye the sensation round, it is because there is an atom of roundness in the impingeing molecule. This can be seen from the discussion carried on in the first Chuan of the Abhidharma Kosa : " Some teachers say that some objects are possessed of both shape and colour. . . How can this be ? (Attempted answer) : Because we can see therein both (shape and colour). (Objection) : This sensation refers to our cognition, and not to the external object." (A.K. l-7a.)

Consequently, when we see something that is both long and white, in reality we are sensing two separate atoms, and their unity is purely subjective. Most of our complex notions are thus reduced to a subjective compounding of a few simple external factors.
Needless to say, this philosophy, once accepted, led to a careful examination of just what things are ultimately objective and distinct. The final enumerations were only reached after considerable thought and controversy ranging over a long period of time, beginning with the crude and hesitating categories found in the Dhamma Sangani (which consequently had more or less to be followed by all the later Sthaviravadins) down to the logical lists of the later Sarvastivadins. The Sarvastivadins had, in fact, a much more definite list of categories than the Yogacarins, for, inasmuch as the latter thought that all matter is the product of the mind, the problem of the ultimate units of the objective universe was not for them an urgent one.

In the present instance, therefore, we may rest content with quoting Vasubandhu s 1 enumeration of the divisions of each sense object which represent the Sarvastivadin opinion on the subject, referring the curious to the other original authorities 2 :
1 A.K., f. 1, first half. Also M. Vyu., p. 195.
2 For Str. Shan-Sang, tr., p. 183 ff. For Yog. Ahh. Sam. San., 1-26 ff.

1. Types of Sight Objects. (a) There are eight kinds of Samsthana rupa, or shape objects :
1. Dirgha, or long.
2. Hrasva, or short.
3. Vrtta, or square.
4. Parimandala, or round.
5. Unnata, or high.
6. Avanata, or low.
7. Sata, or even.
8. Visata, or uneven.
(b) There are twelve kinds of varna rupa, or colour objects, of which four are primitive and eight derivative :
1. Nila, or blue.
2. Pita, or yellow.
3. Lohita, or red.
4. Avadata, or white.
5. Abhra, or cloud colour.
6. Dhuma, or smoke colour.
7. Rajas, or dust colour.
8. Mahika, or must colour.
9. Chaya, or shadow (where objects can be seen).
10. Atapa, or (dazzling) brightness (like sun).
11. Aloka, or light colour (like moon stars).
12. Andhakara, or darkness (objects invisible).

2. Types of Sound Objects. These are of eight varieties, obtained by dividing sound in two different ways : (1) according to the nature of the sound and the object which emits the sound ; (2) according to the nature of the sound as judged by its effect upon the auditors.
The first division is that between (1) Upatta-mahabhuta- hetuka, or sound produced by objects which have the power of perception, i.e. sentient agents ; and (2) Anupatta- mahabhuta-hetuka, or sound produced by objects not having the power of perception, i.e. non-sentient agents. These classes are divided into (a) articulate sound, and (b) inarticulate

sound. Each of these four classes is again divided into pleasant and unpleasant, making the following eight in all :
,. , ;. \(d) pleasant (1).
(1) Articulate . \ ;, L , ;
(1) Sentient . . 6 unpleasant 2.
1(2) Inarticulate - " pl< T f ! ((6) unpleasant (4).
,. , , ((a) pleasant (5).
(1) Articulate . ; * , , ;
(2) Non-sentient . 6 unpleasant 6.
1(2) Inarticulate J " pl< T J > ((6) unpleasant (8).
3. Gandha, or Smell. This has only four classes : (1) Sugandha, or pleasant ; (2) Durgandha, or unpleasant ;
(3) Samagandha, lit. similar smell, but which is interpreted as meaning smell which renders nourishment to the body ;
(4) Visamagandha, dissimilar smell, or smell which does not render nourishment to the body.
4. Rasa, or Taste. This consists of six classes : (1) Madhura, or sweet ; (2) Amla, or sour ; (3) Lavana, or brackish ; (4) Katuka, or acrid ; (5) Tikta, or bitter ; (6) Kasaya, or astringent.
5. Sparsa, or Touch. This consists of eleven classes, viz. : (1) Prithivi, or earthy ; (2) Apas, or watery ; (3) Tejas, or fiery ; (4) Vayu, or airy ; (5) $laksnatva, or smooth ; (6) Karkasatva, or rough ; (7) Laghutva, or light ; (8) Gurutva, or heavy ; (9) Sita, or cold ; (10) Jighatsa, or hunger ; (11) Pipasa, or thirst.
(iii) Atoms and Molecules
In the preceding pages frequent mention has been made of the terms atoms and molecules. It is therefore expedient to examine more closely the Buddhist atomic theory.
While frequent mention is made of the four Mahabhutas, neither the Pali sutras nor the seven Pali Abhidharma works contain any mention of the atomic theory, and the idea seems to have been introduced into Buddhism by the Sarvastivadins, probably as the result of contact with the Vaisesikas, for the Jain atomic theory, like that of the Buddhists, seems a borrowed and not an original doctrine. I have been unable to find any reference to atoms or molecules in the Jnana Prasthana, nor in any of the six pada, which with them constitute the original Abhidharma canon of the Sarvastivadins.

The Mahavibhasa, however, probably composed in the early part of the second century A.D., makes frequent mention of the atoms in its interpretation of the older works, and seemingly with no sense of incongruity. That it was quite prevalent at this time is seen from the fact that the Abhidharma Hrdaya, translated into Chinese in the third century A.D. (one of the earliest Abhidharma works to be translated), contains the whole theory in its developed form. The atomic theory plays an integral part in the philosophy of Vasubandhu and Sanghabhadra. The early and classical Yogacarins, moreover, accept the theory provisionally in spite of their idealism. The later Yogacarins, however, beginning with Dignaga, felt that the doctrine stood in the way of the doctrine that all phenomena spring from the mind, and in his Alambana pratyaya Dignaga wrote a very able refutation of the atomic theory.
The Neo-Sthaviravadin school founded by Buddhaghosa does not seem to have gone, at first, into details of the atomic theory as found in the north, but the kalapa theory (corre sponding to the Sarvastivadin Samghata paramanu) was used by Buddhaghosa himself in the Attha SalinI, and was thereafter considered an integral part of the Sthaviravadin philosophy, the idea being mentioned and considerably developed in the Abhidhammattha-Sahgaha. At a somewhat later time (when, I do not know) the ultimate units of the kalapa or molecule received the name paramanu, and Professor Maung Ting tells me that at the present time it is considered an integral part of Sthaviravadin philosophy.
1. Atoms. The Sarvastivadins (Ab. Hr. 1, etc.) tell us that there are fourteen kinds of atoms, just as the scientist

at least of a generation ago would have said that there are eighty odd kinds of atoms, one kind for each element. The number fourteen is accounted for by one being given to each of the sense organs, and one to each of the sense objects, and four additional ones for each of the four Mahabhutas.

Unlike the atoms of the Vaisesikas and Jains, however, the atoms of the Buddhists are not eternal. They spring into being from time to time, and then are destroyed, lapsing seemingly into nothingness (cf. A.K. 12, latter half). Thus, the atoms of the five sense organs and sense objects originate owing to the atoms of the four elements, and would instantly lapse into decay were it not for the sustaining power of the elemental atoms. Hence every derivative atom has with it, sustaining it, one atom of each of the Mahabhutas. 1 Even these mahabhumika atoms, however, are not themselves permanent, but undergo a four-fold process of birth continuance, decay, and destruction, followed by a new cycle of birth, etc. This phenomenalist view of the atoms is of interest as a contrast to the substantive view of the atoms held by their Indian contemporaries.
 
2. The Molecules. These atoms are grouped together into molecules, and all parts of the material universe consist of these molecules, called Samghata paramanus by the Sarvastivadins, and kalapas by the Sthaviravadins.
Vasubandhu tells us that the molecules of non-sentient matter are the simplest, and that even these are at least eight-fold, i.e. containing at the very least atoms of each of the four elements and four of the sense objects, atoms of sound not being necessarily included. Where a sound atom is produced by the action of the other atoms, the molecules are then nine-fold. Strictly speaking, even an eight-fold molecule contains at least twenty atoms, since there are four sense object atoms and each such derivative atom must have one atom of each of the four elements supporting it, making four
1 A.K. 429.

derivative and sixteen elementary atoms in the simplest molecule. This number is increased to five derivative and twenty elemental atoms in the case of those molecules containing sound atoms. The number eight or nine there refers only to the different kinds of atoms in a molecule, and not to the total number of constituent atoms.
The molecules of every animate body is more complicated. Every such molecule must contain at least nine kinds of atoms, for in addition to the foregoing necessary eight, each molecule contains an atom of the kaya indriya, or touch sense organ. The molecules of the sense orifices (eye, ear, nose, etc.) are at least ten-fold, for such a molecule must contain not only the four elemental atoms, four sense object atoms, and a kaya-indriya atom, but also an additional atom of the sense organ in question.
Since the Neo-Sthaviravadins retained the whole twenty- eight divisions of rupa of the old Pali Abhidharma, their list of the component parts of a molecule differs somewhat from the preceding. Those who are interested in the subject will find particulars in Abhidhammattha Sangaha (C.P. 164, seq.).

(iv) The Eleventh Category of the Northern Schools

Before closing our discussion of the divisions of rupa or matter as known to the Buddhists, we must examine the two as yet unexplained categories, namely, the Avijnapti rupa of the Sarvastivadins, and the mentally perceived matter of the Yogacarins.

1. Avijnapti was a term of much dispute among the Buddhists, and there were even widely diverging opinions as to the general group under which this particular category should come. The term means " not manifested ", or not expressed, or latent. According to the Buddhists, every physical act, word, or thought should have some corresponding result. In many cases the result was open and obvious. An act visibly modified the nature and position of the molecules. Sometimes, however, no such manifested (vijnapti) result

would be observed. The moral theory of the Buddhists would not allow them to suppose that such an action had no result, even though no result were visible, so they invented the category of Avijnapti, which we may say corresponds to the Western idea of the general character of a man, inasmuch as the general character of a man is affected by his past actions, even though it seems to have no result.

This general character was affected by not only definite conscious acts, but also by acts performed unconsciously or in a state of confused thought. On the other hand, for the action to have effect it must be definitely good or bad. This general character, moreover, was of a very general nature, constantly changing and yet remaining with one through the whole course of one s life.
So far, the Northern Buddhists of all schools seem to have been in agreement, but in attempting to give a definite place to this dharma, dispute arose. Harivarman s Sattvasiddhi insists that it must be relegated to the Citta-viprayukta, or miscellaneous dharmas neither mental nor physical. The T ien T ai and the Hua-yen schools of Chinese Buddhism said that the character arising from physical and verbal action was material (rupa), but that mental action was not. The Yogacarins stated that conventionally, or from the relative standpoint, all three might be called physical. The Sarvastivadins, however, claimed that all such character is ultimately material, and dependent upon the four mahabhutas. In consequence, they classified this dharma as the eleventh of their material factors. (For further details of the Sarv. theory, A.K. 1-86 ; also the first half of 13, on Avijnapti karma.)

2. The material properties classified under mental object. 1 The first ten dharmas, including the four mahabhutas, the five sense objects, and the five sense organs, all belong to the realm of sensuous or immediately experienced matter. In addition, however, there are certain aspects of matter of
1 A.S.S., f. l, p. 2fc.

which we have no sensuous knowledge, and which are not the objects of the five sense organs but of the mind. Thus, atoms, etc., cannot be seen, they can only be inferred. The Yogacarins have placed in a separate category all those aspects of matter which are thus defined, whether these aspects be objectively existent or merely the result of mistaken creative imagination on the part of the mind. This includes, but is not confined to, the Avijnapti rupa of the Sarvastivadins, and constitutes the eleventh category of the Yogacarins.

III. CITTA, OR MIND
Having completed our survey of the Buddhist theory of matter, let us now examine their theory of the mind. This is, as we have seen, discussed in a two-fold way in Buddhist books : (1) Mind itself in its various divisions, known as Citta or Vijnana ; and (2) mental qualities evoked by mind in the process of its origination, and which are known as Caitasikas.
In the present instance we are concerned with the first alone. But in order that we may understand the various types of mind enumerated by the Buddhists it is necessary, once more, to emphasize the caused and conditioned nature of all mentation. Buddhist books are constantly repeating the Buddhist adage that there is not ego entity, no self- existing mentator, and that not only is the mentating personality evoked by a combination of causes and conditions, but also that mind ceases to exist when sense object and sense organ cease to interact. Strangely enough, even the Yogacarins, who were idealists, refused to believe in a permanent individual mind-substance, but stated that though mind is the only ultimate reality, every individual mind is constantly changing and being remodelled under the influence of causal law.

All this, of course, is in accord with the curious anti- substantialist position assumed by Buddhism all along the line. There is no eternal self -existing matter. Likewise there


is no eternal self-existing quiescent substance known as mind having a prior existence and which is merely stimulated into activity when brought into contact with the sense objects by means of the sense organs. Rather, certainly according to Hinayana Buddhism, is it a definite product created out of nothing by the interaction of the indriyas and visayas.
This consideration is important, inasmuch as many of the so-called divisions of mind are types of mentation classified in accordance with the nature of the stimulus which brought them into being. Consequently, we find two different methods of classification. One is according to the physical basis from which mentation arises. This method of classification is common to all forms of Buddhism, including the three schools with which we are at present more particularly concerned. The other method of classification is more complicated, and deals with all possible states of mind which can arise at any given place, e.g. on earth, the Rupa dhatu, etc., or with the conditions (as opposed to the bases) under which they arise. This latter mode of classification seems to be peculiar to the Sthaviravadin school. Possibly, the clumsiness of the Southern list caused the philosophers of other schools to think such attempts at classification ill-advised.

1. Mind Classified according to the Bases which Evoke it
Before proceeding to consider the divisions of the mind, however, it will be well to take into consideration the different terms that are used for mind, and see what they mean. Apart from Nama (name) in the compound Nama-Rupa, which covers the whole of the non-material part of the human personality, there are three terms which are most frequently employed. These are Citta, Manas, and Vijnana. 1 Western scholars have not yet come to a standard translation of these important words, but provisionally we may say that Citta is mind, Manas is reason, and Vijnana is consciousness.
1 See discussion of each term, A.K. 4-13a.

Citta, the Buddhists derived, probably wrongly, from Citra, or variegated (Exp., 1-85), a term which the Chinese render J| i|C, or arising by compounding. This derivation was probably given by way of added emphasis to the doctrine of the non-substantial nature of the mind. The term is very little used in the sutras, but came to be the standard word in later days for the whole of the subjective life, as opposed to rupa or even to the Caitasikas, and in some ways corresponds to our " soul ", " heart ", or " spirit ", provided that all these terms are de-atmanized.

Manas, the Buddhists derived from a term to measure (ma) . It implies the calculation, evaluating, j udging of a thing. Consequently, it is frequently used when the mind is considered a reasoning factor. When used technically, therefore, we may call it Reason, but frequently it, like . Citta, may be rendered Mind. It is interesting to note, however, that when the Mahayanists say that the whole universe is but the creation of Mind, or that nothing exists outside the " mind ", it is Citta and not Manas that is used. Very occasionally Vijnana takes the place of Citta in this connexion.

Vijnana is perhaps the oldest of the three terms, and in the early sutras, when the reincarnating personality is spoken of, it is Vijnana and not Manas, or even Citta, that ia mentioned. In later times, however, it came to be used almost exclusively for the sensatory, experiencing aspect of the mind. In this way Citta, or the mind as a whole, is said to be divided into so many Vijnanas or types of data-receiving forms of cognition.

In the midst of all this confusion, it is agreeable to find that the Buddhists, at least the Sarvastivadins and Sthaviravadins, agree that there is only a difference in terminology, and that the thing spoken of is the same (C.P. 234 ; A.K. 4, 13). The slightly different use of the terms by the Yogacarins will be dealt with presently.
As regards the actual number of divisions of Vijnana, the Sthaviravadins and the Sarvastivadins l are in agreement 1 A.K., 1-16, for full discussion.

in postulating but six, as opposed to the eight divisions of the Yogacarins. The six-fold division need not detain us long, for it has already been discussed in connexion with the dhatus and the fifth skandha. There is one vijnana for each of the five material sense organs, and one vijnana (Mano- vijnana) of a more general character, which exercises the functions of reason, judgment, memory, planning, etc. In addition to the other differences between the first five Vijnanas and Mano-vijnana which have already been given, the Buddhists assert that the former are purely passive, while the latter is active. Consequently, while certain phases of the former may come under the category good or bad, it is only because they have been influenced in a moral or immoral way by the former. Again, only a small number of the Caitasikas are associated with the first five vijnanas, while all of them are to be found in association with the Mano-vijnana.
 
One of the points on which Buddhism is singularly obscure is the psycho-physical relation between the first five vijnanas and the all-encompassing Mano-vijnana. The early works do not state where the first five vijnanas are, but by their close association with the sense organs it is implied that they are actually resident in them. This idea was expressly stated in later times (cf. Spence Hardy, Man. Bud., p. 434). Con cerning the physical basis of the Mano-vijnana, there is even greater ambiguity. The Sarvastivadin knows of no physical basis, the sense organ of Mano-vijnana being Manas, or the whole of the disintegrating aspects of consciousness itself. The Neo-Sthaviravadins give the physical basis of the Mano-vijnana as the heart, but since the first five kinds of consciousness occur at the sense doors, and Buddhism knows nothing of the nervous system, we are left wondering how the Mano-vijnana becomes aware of the impressions received by the other five. The Sarvastivadins partly answer, or, rather, evade this question, by saying that the six vijnanas form not six dharmas, but together constitute only one dharma or ultimate factor, and that consequently the six vijnanas are but six aspects of one vijnana rather than six separate entities (cf. A.K., 1-116). This doctrine of the unity of vijnana does not, of course, interfere with the anatman theory.

The eight-fold division of vijnana postulated by the Yogacarins l consists of the foregoing, plus two others. These additional two are :
1. Klista-mano Vijndna, literally, soiled-mind consciousness. This may be rendered by self-consciousness (V.M.S. 4-156). Whereas Mano- vijnana carries on the ordinary process of reasoning, it deals with ideas more or less as they come, without consciously or continuously distinguishing between that which appertains to the self and that which appertains to the non-self. This continual distinction is the work of the seventh vijnana, which, according to the Yogacarins, functions even when a man is asleep or is otherwise unconscious. It is the basis of the constant tendency towards the atman theory, for it falsely considers the Alaya vijnana (the eighth vijnana, the basis of all the other vijnanas) to be a real and permanent ego entity, although in reality it is in a constant state of flux.

2. Alaya Vijndna means repository consciousness, since it is the basic form of all other consciousness, and, in fact, of all forms of existence. All the other vijnanas have their origin in the Alaya vijnana, and owing to their activity the construction of the phenomenal world takes place. This is not the place to discuss the whole of the idealistic philosophy of the Yogacarins, but mention must be made of the triple function of the Alaya vijnana. The first we can call the positive (It 3$ neng ts ang), because it stores up the seeds of all the other vijnanas. The second we can call the negative ($F 3K so ts ang), because it receives the influence of all the
1 The bulk of V.M.S. consists of a full discussion of each of these Vij. For Alaya, pp. 6a-156; for Kl. Man. Vij., pp. 156-21a; for the First Six Vij., 2 la seq.


other phenomenal vijnanas. The third is this vijnana con sidered as the object of false belief, 1 because the seventh vijnana constantly considers that this ever-changing Alaya vijnana is an eternal ego entity. 2

In accordance with their postulation of three kinds of mental activity not immediately connected with the senses (Mano-, Klista-mano-, and Alaya- vijnanas), the Yogacarins slightly modify the old definitions of Citta, Manas, and Vijnana. All three terms, say the Yogacarins, may be used of any of the eight aspects of consciousness, but more especially does the eighth or Alaya vijnana take the title Citta ; more especially does the seventh vijnana receive the title of Manas ; while Vijnana applies more particularly to the sixth vijnana. (Cf. Wei shih lun cheng i, 3 a well-known commentary on the V.M.S., where, p. 48 seq., the matter is discussed at length.)

Finally, we may add that, whereas the Sarvastivadins count their six vijnanas together as forming but a single dharma in their list of seventy-five, the Yogacarins count each of the eight vijnanas separately in making up their list of a hundred dharmas. This fact, however, may be easily misunderstood. It is not that the Sarvastivadins believe in the unity of the mind more than the Yogacarins, but rather that according to the Yogacarins there is in reality only one substance Mind, of which all other things are derivatives, and since it attempts to enumerate all the important derivatives, it gives each of the vijnanas a separate place.
2. Classification of the Mind, according to its Place and Condition

This elaborate classification of the mind plays] a very important part in the Sthaviravadin Abhidharma works, but it seems to be confined to this school. Consequently, while it is necessary to give this list, it is not essential that we should go into great detail about the matter.

• t - v.M.s.2-66. Ptte.
  

A few preliminary remarks, however, will not be out of place. The full list consists of eighty-nine dharmas, but within this number certain main types of classification can be discerned. 1
1. Consciousness classified according to whether it appertains to the Kama, Rupa, or Arupa dhatus.

2. Consciousness classified according to whether it is active, passive, or neither. This means : (a) active = that which engenders karma for the future ; (b) passive = consciousness which comes into being as the result of past Karma ; (c) neither = consciousness which, though active, leads to no further rebirth and may, therefore, be considered inoperative. This refers to the actions of the Buddha, Pratyeka Buddhas, and Arhats, who are free from the wheel of birth and death.
Each of these categories may be several times subdivided :

1. Active Consciousness is divided into meritorious and demeritorious. Meritorious is divided into that which appertains : (i) To the kama dhatu ; (ii) to the rupa dhatu ; (iii) to the arupa dhatu ; and (iv) transcendental state or emancipation. Demeritorious is divided into only one category, that which appertains to the kama dhatu, but this is subdivided into (i) that rooted in greed ; (ii) that rooted in hatred ; (iii) that rooted in ignorance.
2. Passive Consciousness is likewise divided into (a) that appertaining to the kama dhatu, (b) the rupa dhatu, (c) the arupa dhatu, and (d) transcendent. Kama dhatu, resultant consciousness, is likewise divided into (i) meritorious, the result of previous good karma, and (ii) demeritorious, the result of previous bad karma. (i) is again subdivided into (a) associated with hetu or root conditions, and (b) dissociated with hetu or root conditions.
This method of classification having been once clearly understood, the accompanying chart giving the whole eighty-nine divisions of consciousness will become intelligible.
1 Cf. first part of C.P., p. 81 ff., and Book I of R.D. trans. Dh.S., pp. 1-165.

3. Caitasika Dharmas or Mental Properties 1 We now come to a classification of the various phases of mental activity which Buddhism believes to be fundamental. To us the fact that the Buddhists gave these items a place in their list of separate and ultimate factors of life may cause some surprise. We should have expected them to be classed as subdivisions of Citta in much the same way as the six, eight, or eighty-nine fold division of vijnana. This, however, would be inconsistent with the Buddhist non-substantialist position. Other schools of thought might consider these caitasikas to be not separate entities but qualities inherent in a simple substance such as mind, but to the Buddhists this savoured too much of atmanism and substantialism. W T herever possible, Buddhism was determined to make qualities into separate entities or dharmas, and though the Caitasikas were necessarily co-ordinate with the mind (citta), they were independent co-products of the interaction of the sense objects and sense organs, and not merely inherent, dependent qualities.
The Sarvastivadins were especially insistent upon their separation and independence, and with them we may be certain that what was enumerated in their list of seventy- five dharmas was not co-incidental with anything else, for they refused to count any entity twice over, as may be seen from their refusal to count citta as six-fold. (For the fact that every dharma given in their list of seventy-five was distinct, see among many other passages, the discussion on the Citta- viprayukta dharmas, A.K. 4, latter half (136 to end), and also the discussion of Moha, A.K. 10-1.)

With the Yogacarins, the Caitasikas along with the other dharmas, were not separate entities, but only real phases of the one true entity, mind, 2 and even the later Sthaviravadins
1 For Sth. cf. C.P., part ii, pp. 94-110 ; also incidental lists in Book I of Dh. San. For Sarv. Dh. Kaya f. 1 ; A.K., 4-26 to 136 ; N.A. 10-30a ff. For Yog. all Sat. dh. and P. Sk. ; A.S.S. l-3a ff. ; V.M.S. 5-22a ff.
2 V.M.S. 1-56.

seem to accord a certain relative unity to the Citta and Caitasika dharmas, where Citta is sometimes said to be considered as a sphere and Caitasikas as its separate sections.
One other thing has to be borne in mind in this connexion. These Caitasikas do not pretend to be a list of all mental complexes which may be found in an individual. Rather are the Caitasikas the elements of which these complexes are com posed, and just as the four mahabhutas and the five sense object atoms may combine in an infinite number of ways to form the complex external world around us, so may the various Caitasikas be compounded in an infinite number of ways, ranging from the simple thoughts and desires of a child to the most abstruse metaphysical inference (cf. C.P. 237 seq.).

Now a word as to the main groups into which the Caitasikas may be divided. All schools are agreed that there must be at least three such groups, 1 viz. :
1. General mental properties which are neither meritorious
nor demeritorious.

2. Meritorious mental properties.
3. Demeritorious mental properties.
1. General Mental Properties. The Sarvastivadins do not further divide this group, but the Sthaviravadins (at least, the Neo-Sthaviravadins) and the Yogacarins state that there are two kinds of general mental properties : (i) universal mental properties, which are found in every form of mental activity and common to all forms of vijnana ; and (ii) particular mental properties, which are found only under certain conditions and limited to a certain type of vijnana.

2. Meritorious. The Yogacarins and Sarvastivadins do not divide this group, but the Sthaviravadins usually introduce a rather pragmatic four-fold division.
3. Demeritorious. The Sthaviravadins do not divide this category, but the Sarvastivadins and Yogacarins go into great detail. The Sarvastivadins give the following three groups :
1 Further details in authorities cited above.
 
(1) The fundamental klesa dharmas ; (2) the Akusala or positively demeritorious dharmas, which are also fundamental or non-derivative ; and (3) the Upaklesa dharmas or sub sidiary klesas, derived from the dharmas of the preceding two categories. Of the two categories, Mula-klesa and Akusala, the latter is the more positively evil.
The Yogacarins, on the otter hand, only make a distinction between the fundamental klesas and the subsidiary klesas, though, as we shall see later, they introduced several internal subdivisions, particularly as regards the second category.
 
4. Indeterminate. In the Sthaviravadin school all of the Caitasikas were included in one or other of the preceding three categories, but both the Sarvastivadins and Yogacarins added a fourth group, consisting of those dharmas which are indeterminate and which may be either meritorious or meritorious according to circumstances, or which for other reasons can come under none of the preceding heads.
From the preceding remarks, it will be seen that the Sarvastivadins l had the following arrangement of the Caitasikas :
1. Maliabliumika dharmas, or general mental properties,
ten in number.
2. Kusala-mahabhumika dharmas, or mental properties
common to all types of meritorious mental activity, ten in number.
3. Klesa-mahdbhumika dharmas, the fundamental passions
or afflictions, six in number.
4. Akusala-bhumika dharmas, or evil mental properties,
two in number.
5. Upaklesa-bhumika dharmas, the subsidiary passions
derived from the preceding two categories, ten in number.
6. Avydkrta-bhumika dharmas, or miscellaneous mental properties, eight in number.
1 A.K. 4-3a.
The Yogacarin 1 arrangement was as follows :
1. Universal mental properties, common to every act of
consciousness, and neither meritorious nor de meritorious, five in number.
2. Particular mental properties, found only in certain
types of mentation, but neither meritorious nor demeritorious, five in number.
3. Meritorious mental properties, eleven in number, all of
them fundamental.
4. Fundamental Klesas, six in number (demeritorious).
5. Subsidiary Klesas, twenty in number (demeritorious).
6. Indeterminate mental properties, four in number.
The Sthaviravadin 2 classification was:
1. Universal mental properties, seven in number.
2. Particular mental properties, six in number.
3. Demeritorious mental properties, fourteen in number.
4. Universal meritorious mental properties, nineteen in
number.
5. The Abstinences, three in number (meritorious).
6. The Illimitables, two in number (meritorious).
7. Reason (meritorious) (1).
Before we can go into further details concerning general relationships it will be necessary to enumerate the constituent dharmas of each of the categories.
I. General Mental Properties
A. The Sarvastivadins made no distinction between the universal and particular, or rather they believed that all of the following dharmas are universal. They are 3 :
1. Vedand, or feeling.
2. Samjnd, or ideation.
3. Cetand, or volition.
4. Sparsa, or sensation.
1 V.M.S. 5-22a. 2 C.P., p. 98, mnemonic verse. 3 A.K. 4-36.
5. Chanda, or will or conation.
6. Mali, or intelligence, or wisdom, or