The 8th International Conference Buddhism & Australia
will be held on 7-9 February, 2019 in Perth, Western Australia.

Chinese Buddhist Encyclopedia Illustrations
Some of the Buddhist Illustrations created by Chinese Buddhist Encyclopedia
FREE for everyone to use

We would also appreciate your feedback on Chinese Buddhist Encyclopedia. Please write feedback here
Here you can read media articles about the Chinese Buddhist Encyclopedia which have been published all over the world.

Articles by alphabetic order
 Ā Ī Ñ Ś Ū Ö Ō
1 2 3 4 5 6 7 8 9 0

The Samputa Tantra: Edition and Translation Chapters I-IV

From Chinese Buddhist Encyclopedia
Jump to: navigation, search
Please consider making little donation to help us expand the encyclopedia    Donate Paypal-logo.jpg    Enjoy your readings here and have a wonderful day  



This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1. The sign or "target" for pages apparently lacking from the docum9nt photographed is "Missing Page(s)". If it was possible to obtai:s the missing page{s) o:r :;et:tkHl, the•,• are spliced into the film along with adjacent pages. This may have necessitated Cii.rtti t!J tiin.i a.; imat}a and duplicating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterate( with a large round black mark, it is an indication that the photographer suspected that the copy may hav moved during exposure and thus cause a blurred image. You will find .a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite methcd in "sectioning" the material. It is customary to begin photoing at ths upper left hand corner of a large sheet and to continue photoing 'from left to right in equal sections with a small overlap. If necessary, sectioning is continued again -beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.

University Microfiims International
300 North Zeeb Road
Ann Arbor, Michigan 48106 USA St. John's Road, Tyler's Green
High Wycombe, Bucks, England HP10 BHR



ELDER, George Robert, 1942-

Columbia University, Ph.D., 1978
Religion, history

University Microfilms International, Ann Arbor, Michigan 4a1os

@ 1978

George Robert Elder


George Robert Elder

Submitted in partial fulfillment of the requirements for
the degree of Doctor of Philosophy in the Faculty of Philosophy

Columbia University


01 1280.JPG



George Robert Elder

This dissertation is primarily editorial in charac­ terl it contains as its central contribution a Sanskrit edition of the first four chapters of an important ancient Indian Buddhist text, the Sam.put.odbhavasarvatantranidana-
.....: _.,..._,,_ mahakalpadUa, a title abbreviated as Samputa Tantra. The portion edited represents an integral unit as the first of ten parts in the entire text of forty chapters. As a tan­
tric Mahayana scripture, the Sam.put.a is of the Anuttarayoga
class, of the Mother type, and is a Shared Explanatory Tan- tra. Specifically, this classification places our text within the literary cycles of the two Root or mula Tantras,
the Cakrasam.vara and the Hevajra. The literary location is
significant in part because the Hevajra Tantra has already been edited and translated into a Western language; our work, then, advances Buddhist scholarship precisely at that point. The materials for editing consisted of three San­ skrit Mss. of late date and the much earlier Tibetan trans-
lation of the Tantra found in the Canon as Xan dag par

sbyor ba zes bya ba'i rgyud chen po: In addition, use was made of a Tibetan translation of an early commentary on the
work by Suravajra -- his Ratnamala or Rin chen uhren ba.

The contents of the dissertation are as followsa

(1) historical considerations providing an overview of the literary history of the Buddhist Tantras and of the Sa puta Tantra in particular, (2) philological considerations de­ scribing the nature of the rnaterials for editing, the char­ acter of corruption, and the special problems involved in the editing of very corrupt Sanskrit Mss., (J) the Sanskrit edition with an apparatus fully discussed, (4) the Tibetan translation, corrected with an apparatus, and (5) an English translation without annotation.
As for the contents of the text, the first chapter opens with the standard Anuttarayoga Nidana sentence. While certain terms within this sentence are explained immediately by the Bhaaavat, a remarkably full tantric explanation of each term and of each syllable is given in the fourth chap­ ter. Following a voidness contemplation, the creation of
the body-mandala is described along with production of the

Bodhicitta element and manipulation of the "winds" among the nine orifices. In a closing section parallel to lines in
the Hevajra Tantra, the Passion deities are presented. Chap­ ter two teaches the thirty-seven Dharmas Accessory to En­ lightenment in a standard non-tantric Mahayana Buddhist fashion but concludes with the tantric teaching of the Joys, emphasis that Buddhahood is located in the "body," and an enumeration of the thirty-two channels within that "body,"
a list essentially identical with that found in the Heva,jra.

03 1280.JPG

The short chapter three explains the meaning of Dakini and the "yoga of illusion" along with the characteristics of the Bodhicitta and nondiscursive dharmas. Finally, the fourth chapter-- the bulk of which is taken up with a multi-leveled explication of the opening sentence of chap­ ter one -- describes mantra placement, offers a beautiful "I am" passage ("I am the Dharma," etc.), and states that there is "no difference between the compiler of the scrip­ tures and the explainer of them."
Our work, then, provides a new source for that area

of Indian Buddhist studies which has only begun to be ex­ plored, permits a more precise examination of the interre­ lationship of tantric texts within the same literary cycle, and encourages appreciation for the uses of symbolic lan guage in religion.

ABBREVIATIONS ,,,,,,•,,•,,•,,,•,•,,,,• ii

PREFACE ••• I••••••••••••••••••••••••• I iv

1. Historical considerations ••••• 1
2, Philological considerations ••• 17
1. Sanskrit edition

a. Chapter one I I I I I I I I I I I I I I

b. Chapter two •• •• •• • •• • •• • •

c. Chapter three I I I I I I I I I I I I

d. Chapter four I I I I I I I I I I I I I

2. Tibetan translation

a. Chapter one I I I I I I I I I I I I I I

b. Chapter two I I I I I I I I I I I I I I

c. Chanter three • •• • • • • • • • • •


d, Chapter four I I I I I I I I I I I I I

J, English translation

a. Chapter one I I I I I I I I I I I I I I

b. Chapter two I I I I I I I I I I I I I I

c. Chapter three •••• •••••• • •

d. Chapter four ••• •• • •••• •• •

BIBLIOGRAPHY •••• • • • •• • •• • ••• • • • •• •• ••












CIVIT Geore:e, eel. and trans., The a a Tantra Chanters I-VIII. American Oriental Society,

Comm. Suravajra, Rin chen phren ba, Tibetan Bun­ dhist Canon, bstan 'g ur, Japanese Photo­ raphic Fdition of the Peking Fdition, 55z 251-3 , Institute for Advanced Studies of World Religions, Stnny Brook, New York.

Das Sarat Chandra Das, A Tibetan-Fnglish Dic­ tionary (Delhi: Motilal Banarsidass; re­ printed., 1970).

Ed e. Franklin Ed erton, Buddhist Hybrid Sanskrit Grammar and Dictionary, 2 vols. (Delhia Motilal Banarsidass, 1970; reprint of e0., 1953).

GST Benoytosh Bhattacaryya, Guhvasamaja Tantra (Barodaz Oriental Institute, 1931 •


HV'I' D. L. Snelle:rove, ed. and trans., The He­ vajra T ntra, 2 vols. (Londona Oxford Uni­ versity, 1959).

Ms. A. Samputodbhavasarvatantranid namah kalnaraja, Sanskrit Mss., No. 428, Tokyo University Library, Tokyo. (Microfilm.)

Ms. B. Samputodbhavasarvatantranidanamahakalpar ja, Sanskrit Mss., No. MBB-1971-17, Institute for Advanced Studies of World Reli ions, Stony Brook, New York. (Microfiche.)

Ms. c. Sarnputodbhavasarvatantranidanamah kalparaja, Sanskrit Mss., No. 427, Tokyo University Library, Tokyo. (Microfilm.)

MW Monie Monier-Williams, A Sanskrit-Fnglish Dictionary (Oxforda Clarendon Press, 1899).

SVT Shinichi Tsuda, ed., and trans., The Samvarodaya Tantraa Selected Chapters (Tokyoa Hokuseido Press, 1974).

Tib. Yan dag par sbyor ba zes bya ba'i rgyud
chen po, Tibetan Buddhist Canon, bka' 'gvur, Japanese Photographic Edition of the Peking Edition. (Xerograph provided by A. Wayman.)

The thesis which follows is an attempt to recover

a portion of an ancient Buddhist text, The Sam.put.a Tantra. While the work has been primarily editorial, we give an

English translation along with introductory materials.

At each step of our progress through a difficult task, we were helped -- tested and encouraged -- by Pro­ fessor Alex W yman, our mentor at Columbia University. We wish to express here a deep gratitude to him for years of his attention but especially for the model of Buddhist scholarship that he provides.
We wish to thank also the Library of Tokyo University

and the Institute for Advanced Studies in World Religions, Stony Brook, New York, for providing the manuscripts.
Historical Considerations


Sometime during the early centuries of our era, in India, a religious synthesis occurred. Traditional forms of Indian religion-- Hindu and Buddhist -- were faced with nontraditional forms of the sort reconstructed for us by
Sa Tsudaa

The male and female followers of the tantric cult (yogins and yoginis, or perhaps they imagine or call themselves Qakas and Dakinis) make pilgrimage to certain countries which are thought to be places of pilgrimage such as pitha, keetra and so on.
When they meet at such places, they recognize each other through secret signs (choma). Their assembly takes place in a secret, inauspicious place such as a graveyard with the superintendent of the assembly (9an dhyaksa) and his female partner as the central f1gures. This leader of the assembly (cakranatha) and his female partner are thought to be identical with Heruka and his partner Vajrav r h!; this assem­ bly is imagined to be no other than the mandala. Here, pleasure is realized through the sexual yogic practice practised by all the members of this assem­ bly at the same time; it is a pleasure which is uni­ versal, and which can be experienced in common by all the members of the assembly, or in ether words, the pleasures experienced by each member are fused into one super-individual pleasure. This pleasure
is identified with the universal and supreme pleasure re l}zed through the sexual union of Heruka and Var­
1The Samvarodaya Tantraa Selected Chapters, edited and translated by'Shinichi Tsuda (Tokyoa Hokuseido Press, 1974), pp. 56-57. Tsuda is actually reconstructing with line ref­ erences to his tantra; we omit these re•ferences in our quo­ tation since their purpose is not eerved here.
While technical terms in this description need to be dis- cussed, the main features of a tantric cult are clear e- nougha secrecy, the presence of the femala, the sexual em­ brace of male and female, pleasure, identification with the divine. To be sure, much if not most of traditional reli- gion rejected these features out of hand as incompatible with itself or as simply not religious. Yet a significant
proportion of the traditions seemed to recognize the missing other half of itself and -- as if in response to the tantric symbolism itself-- embraced these cults in different ways
to bring to birth something new in the religious life of
India, Hindu and Buddhist Tantra.1

For a while at least, the synthesis must have looked like a "religious movement." "For it was really a vogue," says M. Eliade. He says that "quite suddenly, tantrism becomes immensely popular, not only among philosophers and theologians, but also among the acti.,,e practitioners of the religious life (ascetics and yogins, etc.), and its prestige
also reaches the 'popular' strata." 2 No doubt the roots of
1How the term tantra came to be associated with a particular form of religion is not entirely clear. Prior to the emergence of the cults, the term was used in a general sense much like stra; this meaning seems to derive from' the root -tan as "extend" developed as "draw out, show, ex­ plain." See Chintaharan Chakravarti, Tantrasa Studies on their Religion and Literature (Calcutta• Punthi Pustak, 1963), chapter 1.

2Mircea Eliade, Yoga• Immortality and Freedom (Prince­ ton• Princeton University, 1969; first published, Paris, 1954), pp. 200-20).
the movement reach deeply into the primitive fertility rites discussed by Eliade elsewhere.! Historically, however, two quite different origins are given by the scholarly literature. On the one hand, the cults seem to have been one more invasion by way of the Northwest and part of a general religious current at that time, East and West, which also produced the "mysteries" and "gnosticism" of late Greco-Roman culture. In support of this view,
there is some evidence that Hindu and Buddhist tantric forms


flourish first on the borders, in the Northwest but also East, in Bengal. On the other hand, the cults may have been a hidden force within India from pre-Aryan times. Traces
of tantric language and practice have been noted in the Ve­ das, especially the Atharva Veda, and in the much later Yoga Upanishads. Early Buddhists may preserve an historical trace of tantrism in their rejection of p cakamagu adi hadhamma­ nibbanavada, followers of the "five strands of desire;" and the later Mahayana Buddhists do emphasize a feminine element in the presence of the deity Prajfiaparamita. From these traces, the emergence of Tantra can be seen as a natural
and gradual development of things Indian; or, more dramatic­

ally, as a sudden and even violent eruption of pre-Aryan religious life never really conquered by chariot or fire sac-

1see especially Mircea Eliade, Patterns in Comparative Religion (Cleveland• World Publishing Company, 1963; first published, Paris), chapters 7-9.
rifice but seething more or less underground for many centuries.1
Whatever the origins, in the early centuries of our era, both Early Buddhism (Hinayana) and Mah y a Buddhism encountered Tantra. Apparently, the former rejected it, and the latter in part accommodated the new religion; this can be seen in the fact that on points of doctrine Tantric Buddhism claims to be Mahayana. For example, the medieval Tibetan scholar of the Buddhist tantras, Mkhas grub rje (fifteenth century), states unequivocally& "The doctrine (dar ana)of all (four) sections of the tantras is Prasah­
gika." -- meaning the Candrakirti half of the Madhyamika school, which is Mahayana. 2 That this should be the locus
of the synthesis in Buddhism is not surprising given the feminine presence, as already noted, of the "Perfection of Insight." The A asahasrikapra.j!taparami ta, dated first century and perhaps the earliest of Mah y a sntras,tells
Wisdom controls him who gives gifts,
And also morality, patience, vigour, and concentration.
She takes hold of the wholesome dharmas so that they
may not be lost.
She alone is also the one who reveals all dharmas.
1The literature on this topic is reviewed by Chakra­ varti, Tantras, over several chapters. See also Eliade, Yoga, chapter 6.

2Mkhas rub r•e•s "Fundamentals of the Buddhist Tan­ tras,.. trans. Ferdinand D. Lessing and Alex Wayman The Haguea Mouton, 1968), p. 9).
The Saviours of the world who were in the past, and also those that are (just now) in the ten directions,
Have issued from her, and so will the future ones be.
She is the one who shows the world (for what it is), she is the genetrix, the mother of the Jinas,
And she reveals the thoughts and actions of other beings.l


Feminine and beautiful as these lines are, they lack the erotic tone of Tantra. Thus, th& vehic of the Mahayana that emerges from the synthesis with tantric cults will de­ serve a distinct name, the Vajray a or Mantrayana; and its
scriptures, while of the Sutra type, will be called tantras.2 Of the Sutra type, Buddhist tantras are "revealed"
by the Buddhas or, more prosaicallyp appear as sermons taught by Gautama Buddha -- which, of course, is historically im­ possible and, from the point of view of Mahayana doctrine, never simply was the case. We find in these scriptures, as in others, a dialogue between the Lord and one or more of
his disciples; and we find a religious content with an Early

Buddhist core and a Mahayana transformation of that core. If, as we shall see suggested below, the earliest Buddhist tantras were written in the fourth and fifth centuries A.D., they will have had exposure to early non-tantric Mahayana materials but will have been contemporary with the later de-

1The Perfection of Wisdom in Eight Thousand Lines, trans. Edward Conze (Bolinasa Four Seasons, 197J), pp. 17, 31.

2It is interesting to note that both of these terms, Sutra and tantra, can mean "thread," an image providing some 1ns1ght into how ancient India viewed sentences and the "fabric;t of language.

velopments. Thus, we can expect influence from the

Praj aparamita sutras, the Saddharma Pundarika, the Lank ­ vatara sntra, the writings of Nagarjuna of the second cen­ tury; it should be less clear, however, to what extent the tantras are influenced by the Yogacara developments of the kind represented by the brothers Asanga and Vasubandhu, fifth century, or by the Logic school represented by Digna­
ga and Dharmakirti, sixth and seventh centuries. The tantras themselves, however, will not aid the process of sorting out doctrinal influences since it is specifically the practices
of the tantric cults which were accommodated by the new form of Buddhism. The texts tend to assume the Dharma and discuss instead practices which were new and needed to be explained. These practices were also esoteric and could not be explained without the living guidance of a Guru; and so much of what appears in the texts is sketchy or opaque. We consider that this pragmatic issue is really behind much of the obscurity
of tantric texts which scholars too often judge to be in­ tentionally mystifying as part of a tantric technique. No one can deny the use of this technique in Tantric Buddhism for specific purposes, but -- as we have had occasion to
argue elsewhere -- there remains the simple fact that the tan- tras name and sometimes even list their so-called "secrets." What the secrets mean in a yogic setting, however, is another matter.l


1George R. Elder, "Problems of Language in Buddhist Tantra," History of Religi:.Q!!.li, 15, no. 3 (February, 1976),
pp. 231-250.

It was inevitable that the gurus' explanations themselves should reach written form as well; and so an extensive commentarial tradition developed around the tan­ tras. This has provided us with names and these, in turn, with places and datesa so that it is possible, for exam­ ple, to establish a certain popularity of Buddhist tan­ trism in Bengal during the reign of the Pala kings as early as the eighth century. It is interesting to note that if tantric Mahay a Buddhism was on the rise at this time in certain areas of India, it is precisely when non-tantric Mahayana was generally on the decline. A. Wayman has indi­ cated that, for the Guhyasamaja Tantra at least, commen­ taries were written into the twelfth century when the Bud­ dhist religion in India received its decisive blow from
the Muslim invasion.1 Now, according to the Tibetan his­
torian, Taranatha, the Tantric form of Buddhism did not become public for three hundred years.2 While this possi­ bility needs to be squared with the materials on tantric

popularity mentioned above, the tradition leads us back from the earliest known commentaries of the eighth century
to the fifth century when the tantras must have been extant. Indeed, Wayman places the Guhyasamaja Tantra in the fourth
1Alex Wayman, Yoga of the Guhysamajatantra (Delhiz Motilal Banarsidass, 1977), p. 53. The author was grac­ iously provided with the galley proofs of this book by Professor Wayman since it has just been published and is not readily available.

2Ibid., p. 97.

century with most if not all of the revealed scriptures of Tantric Buddhism written in the fourth and fifth cen-
turies. B. Bhattacaryya, editor of this particular tantra, is generally in accord with a fourth century date or, even
earlier, the third century.1 While these early texts may

have enjoyed Chinese translations in the eighth century, it w s definitely in the following two centuries in Tibet after invention of a written language in the seventh
century (l) and state acceptance of Buddhism in the eighth
that many tantras were translated into Tibetan.
The Sa pu a Tantra, with which this thesis is par­ ticularly concerned, must have enjoyed a literary history similar in kind to that just described, We shall not expect, therefore, to discover a human name or names for its authorship since the tradition will not have preserved any; its authorship is traditionally divine. We shall expect to find its contents in the form of a sermon by
the Lord in dialogue with a disciple concerning points of tantric practice; and, in fact, this is the case. The first human name with which this text is associated is
one Kr acarya, who is said by Taranatha to have brought our tantra to light, And, as we would anticipate, Bhat-

1Ibid. p. 99; Guhyasamaja Tantra, edited by Benoytosh Bhattacharyya (Baroda: Oriental Institute, 1931), p. xxxviii.
tacaryya places this acarya or guide historically in the seventh or eighth century.1 Accordingly, the text enjoys a commentarial literature; indeed it is an impressive one
for quality if not quantity. The catalogs list three com­ mentaries on the Sa pu a Tantra: a relatively small work by a Suravajra who has not been identified by us but who is probably dependent upon earlier and more substantial commentaries and, therefore, of the twelfth century; a compendious work by the famous Abhayakaragupta who wrote significant works on the tantras and who is also late, twelfth century; a smaller yet sizeable commentary by one Indrabhuti who, Wayman thinks, may well be Indrabhuti, the Elder, Guru of Padmasambhava -- sometimes credited with establishing Buddhism in Tibet -- and, therefore, signif­
icantly at the head of the commentarial tradition, late seventh or early eighth century.2 In addition, late Tib­
etan scholarship is not averse to citing our tantra to make a point; in particular, numerous references to the
Samputa are found in the Lam rim chen mo and Snags rim
I •
chen mo, compendia on the tantras written by the renowned fifteenth century reformer, Tson kha pa.3
1B. Bhattacaryya, Journal of the Bihar Orissa Research Society, 14 p. J4J, cited by Chakravarty, Tantras, p. 21.
2see A Com lete Catalo ue of the Tibetan Buddhist Canons. edited by Hakuji Ui,et al. Sendais 193 a and Wayman, Yoga, p. 96.
3see the materials on Tson kha pa in Alex Wayman, The Buddhist Tantrass Light on Indo-Tibetan Esotericism


As to the type of tantra with which we have to deal,

the Samputa is classed in a variety of ways by the numer-
• I
ous systems of classification employed by Indian scholars and elaborated by the Tibetans, notably Bu ston of the fourteenth century. Following Bu ston, Mkhas grub rje presents a fundamental fourfold system which we have pre­ sented elsewhere as follows&
An important structure from which to begin examina­ tion of many issues in Buddhist Tantrism is that of the "four divisions" of its literature classified according to their expression of the "two methods" employed by Tantra, "outer action" (ba.hya-kriy ) and "inner yoga" (adhyatma-yoga). Mkhas-grub-rje, dis­ ciple of Tson-kha-pa and thus commenting out of the Gelugpa school, tells usc "The Kriya Tantra was ex­ pressed for subduing the candidates (vineya) who delight in 'outer action,' while the Carya Tantra was elpressed for subduing the candidates who delight in practicing 'outer action' and •inn:er yoga' in equal measure, The Yoga Tantra was expressed for subduing the candidates who delight in the 'yoga of inner ­ madhi,' while the Anuttara Yoga Tantra is the incom­
parable Tantra for subduing.the candidates who delight
in 'inner yoga'." The confusing similarity of method for the last two divisions -- which are supposed to be distinguished by method -- is clarified by Way­ man's statement of this scheme after Tson-kha-pa. According to the Snags-rim-chen-me, the Yoga Tantra is expressed for those "who delight predominantly
in inner Samadhi over external ritual,.. while the
Anuttara Yoga Tantra is for those "who delight in inner sam dhi completely." We observe that the Snags­ rim-chen-me also correlates this scheme with activ- • ities of "deities," The distinction between deities' "holding hands" and being 11united" -- corresponding

(New Yorka Samuel Weiser, 1973), An unpublished English translation of the introduction to the Snags rim chen mo has come our way. While the work is scheduled for publica= tion in India, it is now in manuscript as Jeffrey Hopkins,
"Tsong-ka-pa's Introduction to Tibetan Tantra," with a pref­
atory commentary by His Holiness Tenzin Gyatso, the Fourteenth Dalai Lama. Tson kha pa's reliance upon the Samputa is readily seen here. • •
to the Yoga Tantra and Anuttara Yoga Tantra divi­ sions, respectively -- is a clue to the very real difference between these two divisions and supports the fact that the fourth division does "completely" what is begun in the third.l

Our tantra, then, is of the Anuttara Yoga type (mal 'byor blamed kyi rgyud) and not concerned with practices such as circumambulation, ritual bathing, the construction of ex­ ternal mandalas, and praises; instead, so this scheme claims, only practices which induce yogic states of mind are dis­
cussed in the Samputa. Incidentally, the correlation with

Passion Families made by the Snags rim chen mo in the anal­ ysis given above is based upon a quotation from the Sa nu a Tantra:
Laughing, looking, holQing hands, The two embracing, there are four.


Another system of classification is structurally simple but obscure, the twofold distinction of subject mat­ ter called Mother and Father. Its obscurity is signaled
by the fact that Mkhas grub rje discusses over the length

of some six pages right and wrong interpretations of the classification. In the midst of the argument, he states& "If one has an Anuttara Tantra, it necessarily teaches the
subject matter of the inseparability of 'means' and 'iilsight'

in those senses." And he goes on a

1Elder, "Problems of Language," pp. 234-235, citing Mkhas grub rje's "Fundamentals," p. 219, and Wayman, Buddhist Tantras, PP 32-33.

2The quotation is outside the four chapters we have translated but is provided by Hopkins, "T ong-ka-pa's Intro­ duction," p. 112.

That being the case, a Mother Tantra is established as followsa it is any Tantra which emphasizes the subject matter of the knowledge of the indissolubil­ ity of Beatitude and Void in the part of 'Insight' on the Void side, while not especially emphasizing (as does a Father Tantra) such things as the method of accomplishin the Illusory Body in the part of
'Means" on the phenomenal fide, or any Anuttara Tar1tra be1onginf.! to its cate ory.

What this means at the least is that all Anuttara Yo a tantras teach the union of "means and insight" identified with "bliss and void," but the Father tantra type empha- sizes the first member of each pair while the Mother tantra type emphasizes the second member. While technical terms here need to be discussed, the position as it stands indi- cates that a lt,ather tantra, emphasizing "means, is more practical than a Mother tantra emphasizing the goal of in­ sight. Not only does this not exactly square with our earlier statement that all tantras are practical, but it seem not to be the case upon reading the texts themselves. Of course, our analysis may be separating a pair united, distinct only in emphasis. Perhaps two additional thoughts can bring them back togethera a Father tantra seems to be more concerned with the particular practice of contemplating light stages; a Mother tantra's concern for the "void" may
be with regard to the particular practice of voidness con­ templation. Whatever the case, our Sa pu a Tantra is an Anuttara Yoga tantra of the Mother type.

1Mkhas grub rje's "Fundamentals," pp. 26), 265.
Interpolation is our own.


Two additional systems of classification concern

the place of a text within the literary cycle. The first is actually manifold but contains two subdivisions of interest to us here a the distinction between Root (mula;
)and Explanatory (vyakhya; b ad rgyud) tantras. The
former distinguishes that particular text which is -- as the name implies -- at the "root" of a literary develop- ment or the chief tantra in a cycle. The latter is nec­ essarily less fundamental and, among other tantras of
like kind, may actually "explain" the meaning of the "root." The final classification, then, is that of Unshared and Shared tantras, texts which are exclusive in some way or
have a more complex relationship to other materials. Again,

we turn to Mkhas grub rje; here he is still discussing the problem of Mother and Father tantras but, in the process, provides information on the character of the Samputaa
Again, in the Fundamental Tantra of Hevajra there is "Thus I have heard;" and in its unshared (with
other Tantras) Explanatory Tantra, the P jara, there is "(I) delight in the Highest of Secrets"; while in the shared Explanatory Tantra, the Samputa, there are both "Thus, I have heard," and "(I) delight in the Highest of Secrets;" consequently it is Non-dual Tantra. That is what is claimed; the authority for it is said to be the explanation by the great magus Nag-po-spyod-pa.

The position is completely untenable; it leads to the absurdity that the Samvara Tantra would in such a case also be a Non-dual Tantra. The Fundamental Tantra
has "(I) delight in the Highest of Secrets;" the un­
shared Explanatory Tantra, the Abhidh ottara has
"Thus I have heard," and the shrred Explanatory Tantra, the Sa pu a has both (phrases).
Clearly, the Samputa Tantra is "Shared Explanatory."
• j
Less clearly, in the thicket of our informant's language, it can be seen as well that it belongs to the "Fundamental" or "Root" Heva.ira Tantra and -- since it is "shared"
also to the mula tantra, the Samvara or Cakrasamvara Tan­ tra. Tsuda confirms our analysis by way of citations from Tson kha pa and Bu ston. The Sa varodaya Tantra, edited and translated in part by Tsuda, is -- along with the Samputa Tantra an Explanatory tantra of the common Root Cakrasamvara,l Without more materials at our disposal and careful analysis, what all this means practically is not
certain. Nothing can be said here with regard to the actual

textual relationship of the Sa pu a and the Cakrasamvara since the mula tantra is not edited. With regard to the Sa varodaya Tantra, superficially the contents do not seem particularly akin to that of the Sa pu a despite their family relationship. The Sa pu a does appear to quote from the mula tantra Heva.ira; more than forty lines in the open­ ing four chapters of our text can be found'in D. Snell- grove's edition of the Heva.ira Tantra and -- if the clas sifications can be trusted at all -- the mula text is not


Mkhas grub rje's "Fundamentals," p. 253. The trans­ lator's actually include catalog numbers of the cited texts in this quotation; we have not included them here.
Sa varodaya Tantra, pp. 27-45.
quoting from the vyakhya text nor is there a more funda­mental common source.l
These quotations, however, are

not documented. In fact, the portion of the Samputa Tan-

tra we have treated never mentions the Hevajra; it names only one tantra on two occasions, the Guhyasamaja Tantra which is considered a Root Anuttarayoga tantra of the Father type. Finally, we note that quotation from the Heva.jra hardly qualifies as "explanation,== and so the actual meaning behind the classification systems needs to be explored.
As noted above, Wayman dates the Guhyasamaja Tantra
in the fourth century. He establishes this by way of a more certain dating of an Explanatory tantra in the fifth century; quite reasonably, the Root tantra is granted a
century's priority over the dependent text.2
There is a

problem here, however, since Wayman wishes to place the composition of all revealed Tantric Buddhist texts, with certain exceptions, within the fourth and fifth centuries. Were this true, all mula tantras would necessarily be dated fourth century to allow for the composi•tion of all their vyakhya tantras a century laterr but this is too simple a solution to individual chronological problems. Besides,
it would deny to the Guhyasamaja itself a certain priority in the tradition, a position that seems happily acknowledged
by other tantras; the Samputa Tantra implies the priority
I •
1The Hevajra Tantra, edited and translated by D. L. Snellgrove, 2 vols. (Londona Oxford University, 1959).
by referring to the scriptures ,generally as "The Sama.ja, etc.," meaning the Guhysama,ja Tantra and others, With this in mind, we can still state a terminus guo for our text as fifth century, but certainly late fifth, The terminus ad quem is provided by Kr acarya, noted above
as seventh or eighth century. Better than that, if Indra­ bhuti, the Elder (late seventh, early eighth century), is
the author of a commentary on the Samputa Tantra, the tantra
I '
should be extant at least a century prior to its being recognized for comment; and this gives us a revised term­ inus ad quem as late sixth century, Our text will then have been composed, "revealed," somewhere between the late fifth century and late sixth century; in recognition of problems discussed above, we sug est early to middle sixth century, at the close of the Gupta period, for the final
2wayman, Yoga, p. 99.
1These matters are treacherous, however, as one can see from comparison with other positions. Snellgrove, Hevajra Tantra, lc 14, 18, refuses to name a date for his text, although he admits it would have to be prior to the eighth century. See The Candamaharosana Tantra, ChaTters I-VIII; edited and translated by Christopher George New Havens American Oriental Society, 1974); there, George provides the date from the seventh to the thirteenth cen­ tury, almost another refusal to give a date, Tsuda, Sam­ varodaya Tantra, p. vii, suggests late eighth century for his text but gives no argument.


Philolo ical considerations

The original text or "autograph" of the Samputa Tantra probably existed in different versions and may have undergone a redaction in its early history; but we
are really not able to do more than surmise such matters. Still, we wish to note that form criticism may bear fruit at the appropriate point in the study of these materi­ als. As we worked our way through four chapters of the
Sa pu a, shifts in style were noticeable, blocks emerged as grammatically more trustworthy, certain sections were clearly derivative as in the case of the quotations from the Hevajra Tantra, the entire third chapter carried the tones of a better poet familiar with the praj ap ramit literature. In brief, seams emerged wmre one did not expect them leading to the impression, if not established fact, that the Sa pu a was put together; to continue the image of Sutra and tantra as "thread," the fabric here
is a "patchwork," of a more or less intentional kind.
The transmission of it would be by way of the handwritten copies of scribes, Buddhist monks who would transcribe with religious care. They may even have been adepts in the tantric practices familiar with the living content of their work; they may even have been scholars. Unfortun-
ately, most scribes were probably only devout. Consider the Mahayana attitude found in the Astasahasrikapraj a­ paramita:
If someone would discipline in Arhatship as many beings
As there are in fields equal to the sands of the river Ganges•
And if someone else, having copied this perfection
of Wisdom,
Would give the book to another being! -- his would be the more distinguished merit.

With religious merit to be gained by merely copying or by merely possessing a copy of a Mah y a text, very many copies of the Sa nu a Tantra must have existed from early times. Since very many scribes must have been involved, it seems likely that only a small proportion of those actually handling the text would have been familiar with the new and esoteric practices. Just as few would have been familiar with the finer points of Sanskrit language which, with the
passing of the Gupta age, was for some becoming esoteric it- self. Add to this the difficulty of transcribing onto a
page what was sometimes only heard in oral transmission and the ravages of time under the best of conditions, and one can readily expect error to creep, if not run headlong, into the original text. Multiply all of this by twelve hundred years, and one arrives at the eighteenth century
1Perfection of Wisdom in Eight Thousand Lines, p. 19.

when the earliest manuscript with which we have worked


was copied. Needless to say, it was corrupt. The reader need only refer to the apparatus of our Sanskrit edition
to experience the extent of the corruption. There appears, then, at the end of the tradition of this text a peculiar phenomenon• every syllable is considered sacred, and in places every syllable is corrupt.
In this context one can see the value of the Tib- etan translations. Our text would have been translated
in the tenth century, a full four hundred years after the date of the autograph, but eight hundred years earlier
than our earliest Sanskrit copy. No doubt different trans-

lations were made, and different versions of the Samputa
underlay these; but a measure of stability would have been provided by Bu ston's redaction of the Tibetan Buddhist Canon in the fourteenth century. The editions, then, would provide even more stability; and with the actual printing
of the Peking edition in the fifteenth century, the Sam--

pu a Tantra in translation would reach a degree of per­ manence that was almost modern.1 What was preserved in this way, furthermore, was an excellent translation. Not
unlike the hieratic art of Tibet, translations for the Canon were formal, literal; the Mah vyutpatti was devised as a translator's tool to show the acceptable range of

1Helmut Hoffmann, The Religions of Tibet (New York• Macmillan Company, 1961), p. 157.
Tibetan renderings of Sanskrit words. Attention was paid not only to the meaning of the Sanskrit but also to the structure of the words in many casesa would render sthita but rnam par gnas would render vyavasthita even though the shift of meaning in the Sanskrit may be minimal.
Such matters are of no little consequence for modern editors

who would attempt to reconstruct an ancient autograph. The work of the Tibetans is truely impressive, and we can for­ give Snellgrove some excess in his enthusiasm when he writesa "Nor would Ifail to mention those Tibetan translators of long ago, without whose labours we should be able to ntake little advance in Buddhist studies. Every one of their
texts is an extraordinary linguistic feat, for no other
translators have ever succeeded in reproducing an original with such painstaking accuracy."!
Our Sanskrit edition of the first four chapters of the Sa pu a Tantra was prepared primarily from three manu­ scripts which we characterize here.
1. Ms. A.
Date copied: given as "modern" and, therefore, 19th or 20th century2
Place copied: Nepal
1Hevajra Tantra, 2a viii.


2The data for Mss. A. and c. are provided by A Cata­ logue of the Sanskrit Manuscripts in the Tokyo University Library, compiled by Seiren Matsunami (Tokyoa Suzuki Research Foundation), p. 152.

Scripta Newari

Size a 89 leaves •. 6 lines per leaf Material: ink on paper
We were provided with a microfilm copy of this manuscript

from the owner, Tokyo University. The condition of the photocopy was excellent and showed a clean careful hand; since it was to this copy that we first turned to deter- mine a character, we call it "A." Its Sanskrit, however, is no less corrupt than that of the other manuscripts.
2. Ms. B.
Date copieda 17431 Place copieda Nepal
Scripta Newari

Size: 69 leaves, 9 lines per leaf Materiala ink on paper
This Ms. is in a Nepalese collection which has been photo­ copied by the Institute for Advanced Studies of World Re­ ligions, Stony Brook, New York; and the Institute provided us with a microfiche copy. While the condition of the Ms. itself appears to be good or excellent and the hand a careful one, our photocopy had only fair to good resolu­ tion. It was interesting for us to have had to work with this Ms. for nearly a half year before receiving the clearer Ms. A. Then we learned what mistakes we were
1nata for this Ms. are provided by the source at the head of the photocopy.
making in our working transcription; and they were pre­ cisely the errors which the scribes themselves had made over the centuries -- loss of subscripts and superscripts, occasional addition of same by "reading" stray marks, con­ fusion of p for y, etc. While "B" is our earliest Ms., this fact proved to be of no consequence for accuracy of reading since only a century's priority is involved and
all Mss. are very late.
3. Ms. C.
DatP copieda 1814 Place copieda Nepal Scripta Newari
Sizea 113 leaves, 6-7 lines per leaf


Materials ink on paper
A microfilm copy of this Ms., owned by Tokyo University, proved to be clear. But the hand is careless, a fact confessed by the scribe who incorporates into the copy de­ vices to indicate that a character should be deleted or adjacent characters reversed; unfortunately, the scribe did not catch all his mistakes. Hardest to work with,
the Ms. still showed a Sanskrit no more or less corrupt

than the others; it was, however, the only copy that was damaged, recorded in Notes to the edition as "missing."
Secondary materials for our Sanskrit edition were Tibetan translati:ons which we characterize as follows.
1. Translation of the Sam.put.a Tantra, Yan dag par sbyor ba "zes bya ba'i rgyud chen po.

Sources Tibetan Buddhist Canon, bka' 'gyur section Editions Japanese Photographic Edition of the Peking Edition (15th century), 20th century
Sizes 35 pages, 84 leaves, 8 lines per leaf

Materials photocopy of woodblock print on paper Our mentor, Professor Alex Wayman, provided us with a xerographic copy which was clear and which, despite var­ ious photographic processes intervening, revealed the careful print of the original. The Tibetan was classi­ cal and relatively free of error; the material is des- ignated "Tib." in the edition.
2. Translation of the Ratnamal , Rin chen phren ba,
a commentary on the Samputa Tantra by one S"'uravajra who
I •

was introduced above.

Sources Tibetan Buddhist Canon, bstan 'gyur section Editions Japanese Photographic Edition of the Peking Edition (15th c.), 20th c.
Sizes 47 pages, 235 leaves, 8 lines per leaf

Materials photocopy of woodblock print on paper The Institute for Advanced Studies of World Religions, Stony Brook, New Yor provided a xerographic copy which had only fair to good resolution; we have the impression that this condition was in part due to the Japanese
Edition itself which seems to have prepared the bstan 'gyur
section of the Canon with less care. Whatever the case, it meant that parts of this source were unreadable. It may be noted that we have made use of the smallest of the three commentaries on the Sa pu aa by comparison, the commentary by Indrabhuti is 104 pa es, that by Abhayakar­ a upta is 146 pp. While we do not doubt that the lar er
works will someday prove the most valuable for understand­
ing the Sama puta a, the smaller and more direct commentary was sufficiently challenging to the present level of our
skills. The source is designated "Comm." in the edition.
As indicated by the data, all Sanskrit copies were written in the Newari script. Since this was not the script in which we were trained, our initial task upon procuring the manuscripts was to establish the syllabus. For this, the "Table of Scripts" found in George's edition of the Candamaharosana Tantra was most useful; and we were able to determine that our individual characters fell within the ran e of possibilities found in different Newari
manuscripts correlated there.1 The same was not true for


the conjunct consonants. Not only did our own conjuncts differ from those presented by Geor e, they differed among our manuscripts and even within any one copy; the variation was such that one could find different formations within
1Capdamaharosaoa Tantra, pp. 88-95. There, one will find as well a bibliography on the Newari script; unfor­ tunately, the important entries were Indian and not avail­ able to us.
the same line. There was also a variety of devices em­ ployed by the scribes -- especially the scribe or scribes
of Ms. c. -- to correct mistakes without erasurea to

indicate deletion, insertion, reversal of characters.

The scribe of Ms. A. may have employed the Visarga as a separator of phrases within the line, but the use is in­ consistent and could not be distinguished from corrupt grammatical use of the Visarga which was frequent in all manuscripts. Since paleo raphy does have a bearing upon the nature of corruption, we list here those individual characters in our manuscripts which were inherently diffi­ cult to distinguish'
1. Vowels
2. Consonants
-a -e
-a -o
u u (no distinction)
c b v (b, v no distinction)
n- t-
n r
p y
bh h
As for the Tibetan materials, the script was classical and only the following characters were inherently diffi­
cult to distinguish&
n t d
p ph b
Judged by the standard of classical Sanskrit or- thography, grammar, and syntax, our manuscripts were extremely corrupt. Corruptions appeared on the average of )/line of 16 syllables which meant that some lines showed every word to be corrupt; this is comparable to the order of corruption found in Tsuda's edition of the
Sa varodaya Tantra. 1 It is more than the avera e of
1/line recorded in Snellgrove's edition of the Hevajra, but Snellgrove has not recorded all variants. 2 All Mss.
were equally corrupt and significant patterns did not emer e; for much of chapter two, however, the corruptions of Mss. A. and C. tended to agree against the reading of
B. With this much error in our copies, agreement among
all Mss. carried less weight than one usually wishes to give such a reading; but often this meant unanimity in error. We record here those features of deviation from
the standard Sanskrit which were present in high frequency; the list is by no means completes
1. Loss
All possibilities discovered with high frequency
-- loss of subscripts, superscripts, letters, words, etc. The loss of final nasal and Anusvara
1see the Sanskrit edition in the Sa varodaya Tantra. 2see the Sanskrit edition in the Hevajra Tantra.
and loss of final Visarga were especially troublesome.
2. Addition
These corruptions tended to be on the order of repetition of words or phrases; Ms. A. appeared to be more culpable in this regard.
3. Confusion of vowels
0 for
for a
a: for

for e

e for
4. Confusion of consonants

c for v

r for 1
n , t interchange r 'n interchange
p ' y interchange
sibilant interchange

5. Double for single consonant

kk for k
gg for g
tt mm
vv for for
for t m
6. Single for double consonants
1 for 11
t for tt
7. Dental for cerebral consonant
t for t.
d for d.
n for n.
8. Final nasal deviation
tp ' m interchange
m.m for 111
Y!laJ:l for m.
By comparison with this order of corruption, the Tibetan material and especially Tib, was relatively free of error,
But corruptions of all kinds do occur and are discussed in the apparatus, The most frequent were those most ex­
pecteds loss of superscript or subscript; p , b interchange,
Since the condition of our materials challenges the attem t to edit, we wish to enter here some observations on the procedures involved, Snellgrove broaches the sub­ ject in his edition of the Hevajra Tantra where he pre­ sents first of all the usual procedure which he calls
"orthodox."1 At the outset, a criterion for accuracy is
assumed; the original text or autograph is understood to have been written in Sanskrit according to standard rules. Then, the Mss,. are collated to allow for the choice of a
1Hevajra Tantra, 2: vii-x.

correct reading or the restoration of a correct reading in the event that all Mss. are corrupt; the variants are recorded. In this way, the editor constructs an accurate basis for translation into another language. He or she may, of course, refer to secondary materials, such as
translations or commentaries, to aid the choice of reading. In the actual procedure of editing the Hevajra, however, Snellgrove "reversed" what has just been described. He confides• "Indeed it must be confessed that the general method has been to first ascertain the intended sense of the text and then edit the manuscripts accordingly."!
This can mean nothing less than the curious circumstance that an authoritative translation was reached without an accurate base. But there was no sleight of hand involved;
Snellgrove simply translated into English the Tibetan trans- lation -- "the most reliable version"-- with the aid of commentaries. Then, for some reason that escapes us, he returned to the Mss. and edited them even though this edi- tion would serve no purpose. His editing consisted of correcting the readings that were clearly "impossible"
given the translation but leaving the remainder of readings

untouched (remainder of one or more Mss.?), recording some variants but not the "useless scribal blunders." Needless to say, Snellgrove's "unorthodox" procedure has come under fire; but he has made an initial assumption which does lead logically to what he has actually done. Characterizing the
1IbJ.. d •t P• V•J.J•.l.
Hevajra, he statesc
More than a hundred lines are quite irregular, and although they clearly represent lokas o£ a kind, it is impossible to see how many of them can ever have been anything but irregular. Thus where there is a choice of reading, the original one is by no means necessarily the one that would permit correct
scansion, Likewise, in the case of grammar and syn­
tax, there seems to be complete carelessness in the matter of endings and irregularities of a kind that would be insufficintly explained by the ignorance
of scribes,1
He assumes that the autograph is corrupt, and he loses
the criterion for accuracy, Snellgrove can no longer choose one Sanskrit reading over another, and he is forced to turn to secondary materials since an English translation is still desired,
In his remarks on editing the Sa varodaya Tantra, Tsuda criticizes Snellgrove for having circumvented the
editing process and for having relied excessively upon the Tibetan materials. 2 But Tsuda has not lost his criterion
for accuracy. He, too, assumes that the original text of the Sa varodaya or any tantra was "imperfect," yet Tsuda hypothetically presupposes the "intention of the author" to write good Sanskrit. While we think Tsuda would have been hard pressed to assume anything else -- the "author" would not have intended to write bad Sanskrit-- the most

1Ibid., p. ix.
2samvarodaya Tantra, pp. 6-16. In a review article,
A. Wayman'has also been critical of Snellgrove's statement
o£ his procedure if not his actual work. He states a "Thus, an E.di tor of a Buddhist Sanskrit manuscript uses the Tibe­ tan translation for predicting the words and syntax of the
interesting point here is Tsuda's presupposition of an "author," a point that is not discussed. At any rate, this author with all his good intention produced a work which did not follow the classical Sanskrit rules.. It happened, hypothetically, in the following ways
The author intended to write the Samvarodaya Tantra in grammatically correct Sanskrit. But, when he felt difficulty in establishing a correct meter,
he gave priority to the meter, especially to the fifth, sixth and seventh syllables (or more ex­ actly, the fourth, the third and the second sylla- 9les from the last, respectively) of each pada of sloka, and consequently introduced various kinds of
grammatical distortion. 1

Tsuda suggests that the distortion thus produced must have been within the acceptable range of possibilities for Sanskrit literature at the time of composition. With
this criterion in hand, the editor cru1 now collate his Sanskrit Mss., choose the correct reading-- but permit distortions within certain syllables of a line written in verse -- record all variants, and realize an accurate base for translation into another language. The procedure is really quite "orthodox," and the Tibetan materials are given their proper secondary function to aid in the pre-

original; and I believe that is a better description of what Snellgrove actually did. That is to say, he some­ times gave greater weight to the predictability of the original Sanskrit words on the basis of the Tibetan trans­ lation than he gave to the words as he actually found them in the three Mss. of which he made primary use." See the review in Journal of the American Oriental Society 80.2 (1960): 159-162.
1Samvarodaya Ta tra, p. 13. Note that there is no criterion for treating the prose.
diction of Sanskrit words.

We judge these discussions to be really quite

hel ful. While they represent two different points of view on the tasks of the editor, they clarify problems involved and provide guidance for avoiding the pitfalls of working with very corrupt materials. Still, Tsuda is no doubt correct in judging Snellgrove's product to be a
translation from the Tibetan and not really an edition of the Sanskrit; Snellgrove, on the other hand, would prob­ ably be right in finding Tsuda too reliant upon internal criteria, resting upon a questionable hypothesis concern­ ing the Sanskrit text. Both, however, strike us as sitting a bit back from their desks, at least for the duration of the discussions; they sound like those philosophers of history whom the historians ignore, going about their
tasks with the requisite skills but also with a lot of

common sense. The roblems encountered by Snellgrove and Tsuda may certainly have been different from our own, but we found it unnecessary to bring into the editing proce­ dure assumptions and hypotheses external to the materials at hand. On the contrary, it seemed to us that the cri­ teria for accuracy emerged from these materials. It was clear that the Tibetan translation was trustworthy; its own orthography, grammar, and meter were so regular that one could assume from that quality observed that care had been taken in the Tibetan translation of the
Sanskrit text. One could then, like Snellgrove, rightly begin with the secondary material in order to anticipate the Sanskrit line. Anticipating and finding, of course, are two different things; and so in case of difficulty
in reconstructing the Sanskrit, one could return to a
different sort of secondary material, the commentary, to solve the dispute. Again, the quality of the Tibetan in the commentary made it highly likely that it reflected the Sanskrit. But its trustworthiness was also of another order; namely, that of the who is explaining what is supposed to be the meaning of a passage. This is, in our opinion, really the key to tantric studies; no matter how corrupt the Sanskrit or Tibetan texts may be, the range of possibilities for a genuine reading is given by this com­ mentary. The dispute resolved, the editor corrects the Sanskrit or, less likely, the Tibetan. In this way, one arrives at the outline of a defensible Sanskrit line and coincidence among the primary and secondary materials.
The problems remaining are critical refinements,

and the Tibetan materials cannot really be brought to bear upon them1 chos in Tib. and Comm. may establish Dharma in Sanskrit and even the use of this word as dharmesu. But
the Tibetan cannot determine the merit of dharmmesu. For this determination, it is tempting to bring in the external criteria of Buddhist Hybrid Sanskrit worked out so pains­
takingly by F. Edgerton.1 It will be noted that our most
frequent corruptions were precisely those items which Edgerton lists as features of a very early dialect; to edit them out of existence just might be tantamount to losing the original text. Again, we felt obliged to let the text decide. Among the several Mss. and within any one Ms., there was neither consistency of classical San­ skrit forms nor consistency of what has been offered as
BHS formation. This, we feel, favors the conservative po­ sition of accepting the standard Sanskrit reading.1 In
addition, we think that our account of the literary his­ tory of the text allows for a high order of corruption, corruption from faulty hearing as well as eyesight which
could produce -mm- for -m- and -esu• for -es.u. Lest we
overcorrect, we did not feel obliged to regularize all gender formations nor did we insist upon the terminal -m when Anusv ra appeared in its stead; finally, we did not
manipulate the lines of verse to create a better poetry
than we appeared to have.2
It should perhaps be noted

1see Franklin Edgerton, Buddhist Hybrid Sanskrit Grammar and Dictionary, 2 vols. (Delhia Motilal Banar­ sidass, 1970s reprint of 1953 edition).
1Edgerton questions our thinking here by suggesting that BHS was "modified in the direction of standard San­ skrit." While this is certainly a possibility, it intro­ duces so many variables that one is unable to edit; indeed, it might even mean that it is not possible to determine
the features of BHS, as Edgerton claims to have done. We
should say, however, that the BHS question is open and potentially very important for our own work. See Edgerton, Buddhist Hybrid Sanskrit, la4.
2We must, like Snellgrove, make our own confession here of leaving very much open in our edition the solution
here that vajra, tattva, and sattva always appeared in our Mss. as vaja, tatva, and satva; we have regularized them without confidence since the consistency of their
formations deserves special study. However our procedures
are judged, the reader will find all variants and some dis- cussion in the apparatus. We are keenly aware that our work is merely in progress.
to problems of meter. Essentially, where it occurred, it was• loka; but frequently the Pada reached nine or ten syllables, and occasionally there appears an independent Pada joined only in sense to a regular line. We think it is most likely that mediocre poetical skill is involved
in many cases; but we also think that a thorough study of
metrical patterns in the Samputa is necessary in the fu- ture. ' •


(Chapter 1)
(o )l namo2 vajra-3 gakaya4 //
eva 5 maya rutam 6 ekasmin samaye7 bhagavan sarva-8
1All Mss. open the Text with the mantra syllable o . Tib. omits o . Comm. shows no awareness of o . We judge it more likely that om was added as a late gesture of rev­ erence than that an early translator would lose a signifi­ cant opening mantra. Accordingly, we qualify its inclu­ sion. Note that Tib. of CMT omits o and George follows All Mss. transliterating the mantra without qualification; the Tib. of HVT also omits o 1n opposition to All Mss.
while Snellgrove fails to treat the line at all in his trans­
2A. vajra akaya nama 3All Mss. vaja
4c. -dakaya
5A, B. evam 6A. ruta.J!l 7c. Samaya sA. sarvva

tathagatakayavakcitta(hrdaya) 1 vajra-2 yo id-3 bhage u vijahara //
tatra khalu bhagavan asiti-4 ko iyogi varamadhye5 vajra- 6 garbham avaloka 7 smitam akar it8 //
1All Mss, omit h daya. Tib, includes s in po, Comm, omits s in po, HVT 1ncludes h daya and s in po.
GST includes h daya, Important witnesses to the contrary,
hrdaya enjoys some early testimony and the testimony of other Anuttarayoga class tantras which traditionally are said to begin in the same way. The fourth chapter of the Text commments on the opening sentence (nidanaJ of all tantras and, significantly, comments on hrdaya. Finally, the symbolic interpretation of the opening forty syllables of the Nidana requires hrdaya,
2All Mss. vaja

JA. yo i 4A. siti 5B. madhya
6All Mss. vaja

?All Mss. avalokkasmitam. Tib, gzigs te 'dzum pa mdzad, Comm. gzigs in 'dzum pa mdzad and also gzigs nas 'dzum pa mdzad. While the Mss, suggest a Dvandva compound, awkward for translation, the Tibetan witnesses indicate an initial inflected participle. A final Visarga is easily lost in oral transcription.

8A. akar it
samanantar1 smite •smin2 vajra-3 garbha4 utthaya­
sanad ekam.sam5 uttarasangam.6 kr.tva
pratis.t.hapya7 kr.tanjali-8 put.ottatva bhagavantam9

1All Mss. samanantara. Tib. ma thag tu. Comm. also reads rna thag tu and explains it as "afterwards" (rjes thags la). While the appearance of a locative absolute here suggests the loss of -e, the line exhibits the tendency to lose Anusv ra; an indeclinable form would be more likely to require commentary.
2A, c. smitesmin; B. smitesmitesmin. Tib. for the whole phrase reads 'dzum pa mdzad ma thag tu de'i tshe. While B. obviously repeats, All Mss. raise the possibility of a BHS locative of smita-; in fact, Edgerton criticizes editors who make our correction "as if the pronoun asmin were present, rather than a case ending." (Edg., 8.70)
Tib. makes it clear, however, that a demonstrative pronoun is present. The pronoun may be part of a locative absolute, but Tib. reads it as a locative of time for what follows.
JAll Mss. vaja 4c. -gabha
5A. ekasam
6All Mss. uttar sanga. The loss of Anusvara is frequent in transcription.

?All Mss. pratisth pya 8c. kr.tam.jali
9c. bhagavanta metad
etad avocat //
srotum icchami jnanendra1 sarva-2 tantranidanaka 1
rahasy s pu odbhuta J lak am4 // bhagavan aha//
aho vajra-5 garbha s dhu s dhu mah!krpa 6 sadhu sadhu maha-7 bodhisattva8 sadhu sadhu gu akara9 //
lA. j !hendra 2A, c. sarvva
JB. sampu odbhuta a c. s putodbhuta
4All Mss. lakqalak asa pu . Tibwmtshan flid. Comm. mtshan ffid. The extent of correction might suggest excessive reliance on Tibetan materialsa but the weight of these materials here is actually due to internal consider­ ations. When the Great Bodhisattvas repeat Vajragarbha's request word for word below, only lak apa is useda when the
Bhagavat explains word for word, again only lak a is used.
Tib, in both cases below reads mtshan flid. While the form
-udbhutam in the original request is repeated below as -ud­ Bhava, we judge the change of an order different from that required to support All Mss. against lak apa.
Finally, it remains for us to account for the corrup­
tion. The original laks a must have been subject to rep­ etition in some waya a mere lengthening of an internal vowel produces a compound and the addition of sa pu a (or possibly sa yuta at some stage) explains the compound, There might even have been an attempt at meter involved since the un­ corrected line displays twenty-four syllables,

5All Mss, vaja
6A. mahakrpa
7A, B. omit mah -. Tib, chen po. 8All Mss. -bodhisatva
9B, c. g kara
yad rahasya sarva tantre u tat sarva p cchate-1 cchaya2 //
atha te vajra-3 garbhapramukh 4 mahabodhisattva 5
prahar otphullalocana 6 prcchantiha svasamdehan? prani­ patya8muhurmuhu 9 //
sarvatantra lO kim ucyate1 1 Nidana tu kath bhavet /
lA, B. p:rcchya
2A, c. tetvaya; B. tasyaya. Tib. 'dod pas. JAll Mss. vaja-
4B. pramukha; c. pramukho
c. -bodhisatvah.; A. -bodhisatvam.ah.
prahar optphullalocana ; c. prahar otphulavada

?A. sa dehat; B. sadehana; c. sa hon. Tib. rah gi the tshom. Lack of a case ending in Tib. points to accus­ ative case in Skt.
8A, c. pranipatya 9A. muhumuhul;l

1°B. -tantra. There is the possibility that uninflect­ ed forms should be accepted here given the context of ex­ plaining terms. Any one Ms., however, shows variety in this regard and the weight of this section is on inflection. The singular number is acceptable here in the sense of "every tantra."
A. ucyanteh
rahasyo •tra1 kim2 ucyeta3 sampu odbhav 4 kathan5 nama6 //
lak a 7 tatra kath bhavet // bhagavan aha //
sarva8 ca te tantr 9 ca sarva-10 tantra 11
1A. rahasyetha; B. rahasye 'tha; c. rahasya 'tha. Tib. 'di la gsan ba. Comrn. gan la gsan ba. While the Mss. suggest atha as part of the solution here, it lacks the
locative nuance suggested by Tibetan materials. The weak­
.ness of these materials -- Tib. points to atra while Cornrn. points to yatra -- is offset by the appearance in the same line of the locative particle tatra translated by Tib. de la. A similar yet different word may then be signaled here by
Tib. 'di la, which Comrn. corrupts. Internal consideration
also allows rahasya in the nominative singular masculine, an inflection which appears below. The corruption then amounts to the easy confusion of -o and -e and frequent loss of sub­ script -r; the gain of aspiration is possible in oral trans­ scription.

20 •
ko m
0 s m 0 ss 0 ng.
JA. ucyetaQ; c. (u-missing)cyate. Tib. gsuns. A form of the same verb is translated immediately above as brjod; the change in translation supports a change in form.
4All Mss. samputodbhava. Tib, yah dag sbyor byun. Note that Vajragarbha's original request readssa putodbhuta translated yah dag par sbyor ba las byun ba.
5c. kath8.J!l
6c. nam
7A. lak aQr c. lak ana
8A, c. sarvva. The uninflected form is acceptable given the context of explaining the first member of a com­ pound.
9B, c. t tras 10A, c. sarvva-
11A. -tantra; c. t tra
sarva1-tantra6abdena2 samajadayas3
nidanarn.4 bhfita-
ni citarn ityartha I hariharahira ya-5 garbha-6 Sravaka-7
pratyekabuddhanarn agocaratvad rahasya I sampu am prajrto­ payatmaka 8 tantrodbhava 9 sampu a10 samapattir11 ityar­ tha I udbhava utpatti 12 /eva bhfita sthira-13 calatmaka-14
1A. sarvva; c. sa(missing) 2A. sabdena

3B. -adaya s; c. -aday 4c. Nidana
5c. hira:rpnya-
6A. -garbha
?A. 6ravakal;l
8A. praj apayatmaka ; C. praj opayanrnaka 9A. tantrodbhava; c. tantrobhava
10c. sarnputal:l
11B, c. sarnadhipattir 12c. utpatil).
13A. sthila-; c. thira
14A. -calatrnaka ; c. -caratmaka
(sarvabhava)-1svabhav tmakam.2 I laks.anam ityanena laks.yate31/ athava sarva-4 tantranidana-5 s pu a-6 sabdena7
vajrasattvo8 'bhidh!yate9 I rahasya ityanena lO upadesika­ mantroddharo11 aleya- 12 devatady-13 abhi ekalak ana ca II
praj opayatmaka tantr tan14 me15 nigadita 16 €r u I

1A, c. sarvvabhava; B. sarvabhava. Tib. omits. Comm. omits. Lacking Tibetan witnesses, the compound probably represents a repetition of Svabhava and subse.quent altera­ tion of sva- to sarva -; accordingly, its presence is qual­ ified.
B. svabhatmaka
JA. lak yete II ityanena lak am; B. lak yate ity­ anena lakean ; c. lak ate I ityanena lak ana . Tib. mtshan
nid ces bya ba ni 'dis mtson par byed pa. The major corrup­ tion or word order here is due to confusion of similar words and due to corrupt punctuation.
A, c. sarvva 5c. nidanam 6c. samputa 7c. sabdena
A11 Mss. vajasatvo
9B. bhidh!yato; c. bhidh!yete

10A, B. rahasyetyanena; c. rahasyatenena. Double sa - dhi is the corruption here: rahasya iti to rahasyeti.
11c. mantroddhara
12A. m ale-; B, mandalaya-; C. m areya, 13B. devatay-
14A. B. tat 15A. rna
16A, B. nigaditah
pratham 1 unyata vicintya 2 mal J prak lya dehinalJl //
rupadhatviti4 unyanam5 abda6 tathaiva karayet 1
gandha-7 dnatviti8 Uhyana 9 rasa1° tathaiva11 karayet //
spar6a-12 dh tviti13 Uhyan 14 manas15 tathaiva16
karayl3t //
1B. pratham unyata
2A. vicintya ; c. vicitya JA. mala; c. Mara
4All Mss. in this section read dhatu without inflec­ tion before iti.

5A. sunyan : B. sunya; C. sunya . Tib. ston pa yi. Regular meter is served by genitive plural.

6A. sabdanJ C, sabdam. Lack of inflection here par­ allels -dhatviti in the same line.

7A. gandha
aB. dhatvi
9A. sunyana'f!lal); c. siinyanam lOA, c. rasam.
tas./ca .va
12c. sparsa 13B. -dhati
14A. siinyana ; c. sunyanalJl
15c• manal!l
16All Mss. tasyaiva. Tib. de nid. This is obviously
a corruption of tathaiva translated de id consistently in
this section.

vajragar.bha uvacal II
cak ur-2 atmeti3 unyana 4 madhye kath 5 vi­ ji{anal!l bhavet I
srotatmeti Unyana 6 madhye7 Vijnana 8 kathal!l
ghraoatmeti finyan 9 madhye vijfianal!l 10 katham bhavet I
jihvatmeti unyana 11 madhye viji{ana katham
bhavet II
kayatmeti sunyanam12 madhye 13 Vijnana 14 kathal!l bhavet I
1All Mss. vajagarbha uvaca. After the connecting phrase, A. repeats the entire preceding section beginning with rupadhatviti.

2All Mss. cak ur- is being treated as an uninflected form in compound parallel to grammatical structures above. This serves as the grammatical pattern for what follows.

3c. -anmeti
B. linyana ; C • siinyana
5B. katha
A. siinyan!I!l; B. iirnyana; C. suny8.naf9
7B. madhya 8
B. vij?tana

c. siinyal'\am


sunyan ; c. sunya(missing) 12A, C. siinyanaii).

l3B. madhya 14A. Vijnana

manobahi1'].1 tinya:haJ!l 2 madhye vi jnanam3 kathall) bhavet //
tasma-n na-st•4 r-upan5 na rotaro7 I
dra .ta-
ca nas/abd o6 na
na gandho8 napi9 ghrata na rase napi rasakary1°;; na sparsa11 napi spra a ca12 na citta13 n pi
cintaka.Ip 14 //
1A. -cahil)
A, c• siinyanam
3c. missing
4c. tasmanasti. The passage appears in HVT (1.5.l)r nasti rupa na draeta ca na sabdo napi srota ca // na gan­ dho napi ghr t ca na rase napi rasaka // na spar o n pi spra ta ca na Citta napi caittika //
5c. r'ilpaJ!l
B. sabda1'].1 C. 6abda
7B. sotaro. Tib thos JB po. HVT (1.5.1) 6rota ca. The inflection accepted here and witnessed by All Mss. ap­ pears in Skt. as nominative plural of stem. This may represent a confusion of number acceptable to BHS but is
more likely a BHS nominative singular of -f stem specifical­
ly noted by Edgerton as derived from the Skt. accusative singular -aram. (Edg., 13.19) The acceptability of our form seems as likely as the corruption of ca into -ro as HVT suggests.
c. gandha
9A. napi
1°A. rasaka; c. rasakal)
11c. na spar6a na- is missing. 12
A. cal)
13c. citra 14
All Mss. citrikam. Tib. for the Pada sems med sems
las byun ba'an med. HVT caittika , translated sems las
bhagavan aha//
u1 tattva2 yatha margamJ advaya dvayavarjitam4 II apratarkam5 avij ey6 durbodhy tarkikais 7 tatha II rahasya sarva-8 buddnanam9 ak asadrs tatha I Sravaka na prajananti 10 ajnanatamasa vrta ll II vasananta- 12 nay j an13 pratyekabuddhanirmita 14 I

byun ba. The necessity of a word for "thought" and a word for "thinker" is the stable aspect of an otherwise problem­ atic P da. We have chosen cintaka as an acceptable word for "thinker" and as a form easily corrupted to citrika, What this apparently means, however, is that the Tib. translation
is wrong: it reflects caittika which in fact we find in Snell­ grove's edition of HVT, But caittika cannot be translated "thinker," a fact which Snellgrove happily ignores.
1B. sllnya
2A, tatva; B, c. tatv
3All Mss. marggam
4A. dvayavarjita; B,dvayarjita ; c. dvayavarjita
5A, B, apratarkkam; c. missing
6A, B. avij ayarn; C.(missing)vijl{ayam. The vowels e and a are easily confused,
7B. tarkkikais 8A, C, sarvva
9A. buddh ; B. buddha a; c. budhanam
1°B. prajanati; c. praja(missing) 11A. v;-ta
B. casana-
14A, c. nirmmit ; B. nirmmita

tepi sarve1 na jananti rahasya buddha- 2 gocara 3 II sa sarar.Qava-4 ghor 5 utti a-6 gatacetas 7 I madhye vartita- 8 dehanam9 utti a- 10 lak a ucyate II lak alak alak 11 vij ana- 12 j ana!3 cetas 14 I j anena j eyam15 aloke16 jneyana gatim Ik ayet17 II
lA, B. sarvve
. bu(missing)
3A, gacara
4B. sasara ave; c. sasara ava
5B. ghoran ; c. ghor a m 6A, c. urttirQa
7c. cetasa
8A11 Mss. varttita
9A. dehana a m; B, c. dehana 10A. urttirn.a
11A. lak alak aoa lak ana ; B. l k alak anaksanam;
c. lak lak analak ana . T b. mtshan nid mtshan g i mtshon
bar byed, Comm. explains three wordsa mtshon bya, mtshon
byed, mtshan id. The form lak alak a was encountered above where it was judged to be a corruption by way of repe­ tition of the single word lak • Here, the corrupt form must contain two words to provide a phrase with three words
signaled by Tibetan materials. The corruption, then, amow1ts to the lengthening of an internal vowel.

12A l Mss. vijfiana . Given the exceptional corruption
of the l ne, the agreement of Mss. here carries little weight.

13B, omits. c. gata. 14c. cetasa
15All Mss. j ay2 . See Note 12 above, 16A. alokka; B, a lokka

gat! cetas tu yatha2 svecchadi-3 gamyat h.4/ sahasranekadha5 yogr6 purva-7 janma-8 gati-9 gata //
(lak ana )1° Unye11 sa cintya12 purva-13 lak a14 uktita_tl15 /
unyak etradi- 16 dehasya bij vapayati17 buddhiman 18//

17c. ik aye.
1A. gatik aJ c. gati k a 2
c. jatha
A. gamyat ; B. gamyata; c. gammyata .
for -a is not a frequent corruption, it does
for - is frequent.

5B. -anekadha
While -a
occur; and -a
6All Mss. yogi. Tib omits. The line is doubtfulr and Tib may also be corrupt. The translation for yogi which is omitted here appears transliterated yo gis in the line fol­ lowing. Since both lines in Skt. and Tib. show regular meter and can be translated, more materials are needed to deter­ mine a strong reading.

7A, B. purvva 8B, c. jatma
9B. gati
10All Mss. samatam. Tib. mtshan id. The correction boldly relies upon the'Tibetan but is supported by the sense of the remainder of the linea two forms of lak ana are im­ plied. The nominative singular here accommodates•a nomina­ tive absolute structure also implied by the line. See, however, Note 6 above.
1lc. Unya

12A. sa cintyal:)r c. s citya 13c. purvva
14c. ralak ana 15A. uktita

nagi-1 su irad2 vinirgat viraja tat prakirtita 3/ bodhicitt 4 prabhasvara 5 6uddha-6 spha ika-7
sa ibha8 //
piXca-9 jnanamaya tattva10 sar apa-11 phala-12 mat aka I

16A11 Mss. ksatradi- Edgerton lists the uncorrected form as a BHS alternative; we judge it just as likely that our Mss. have confused vowels a and e, a frequent corruption.

17A. bij pa yatu; B. b!japayantu; c. b!j payatu. Tib. sa bon gdab. Comm. sa bon gdab, A form of the verb vap-
is anticipated by Tib, but pat- seems more likely with ac­ cusative plural feminine of b!ja. Our understanding of the meaning in this context, however, requires a singular form
of Bija and suggests in turn that b!ja- reflects the loss of initial v- of a causative form of vap-, the expected verb, This simple loss of a consonant may then have led to the other corruptions here,
18B. character undetermined 1A. nabhi; B. nadi1 C. nadi
A, • su9 ra ; • su ra • • m zes spy .

3All Mss. prakirtita
A. Bodhicitta a r B, c. Bodhicitta 5A. prabhasvara ah
6c. suddha
7A, c. sphatika
B. sam.nibhah.
9A. paca

1°A. tatvaJ B, C, tatva 11B, c. sar apas
12c. phara
tasya madhye 1 sthita devam2 avyakta-3 vyakta-4 rupi5 //
ardha-6 matram? para sUk ma 8 bindu-9 rupa manomayalJl I
hrd-10 madhye vasate nitya jyoti-11 var o maha­

dyutil;l12 //
dvada antan navanta ca-13 pada-14 talamastaka 1

lB. madhya; c. mad he 2A. devah m-
JA. avyakta

4c. vakta

5A. rup1 ; B, C. rup1n . Tib. ran bzin can. A
form of rupin is required and agreement with devam points
to accusative neuter. Note, however, that meter is thereby made irregular. Edg. does not list rupinam as a possible form in BHS. •
6A. arddha; B, c. addha
7A. matra 8
A, C. 6uk ma ; B. 6uk ya. Tib. phra.

9A. bi du 10All Mss. h;-t- 11A. jyo
12B. dyuti1 c. dyuti

lJA. navanta6cah; B. navantatva; C. navantasya. Note that our correction still permits two genders for -anta in the same line, an acceptable level of irregularity.

14All Mss. apada. Tib. rkah•••med par. The obvious short vowel required here has been corrupted by way of
samdhi with a final short a preceding, e.g. c. -tasya apada-. See Note above.

tantureko vinirgatya 1 nabhimadhye vyavasthita II
p camr-2 kala-3 matras4 tu nagendrak tivistara I
sa kalapi5ca hitva tu6 vajra-7 madhye8 vinirgata9 II
yoni-1° madhye11 sthita bij12 dharmadhatu- 13
. 14
dravik tal)'l
kramas calanal)'l 15 tasya16 navadvare u17 sarvada18 II

B. vinirgattya panIllcama-
kara 4c. matral)'l 5c. karapi 6A. tuh
7All Mss. vaja 8B. madhya
9B, c. vinirggata
10c. yon! 11B. madhya 12A. bijal)'l
c. dharmmadh tu
-kr.tam.ah. sam.caranam.
16c. t(missing)
17A, B. dvale!:]u 18A, c. sarvvada

agnibrahma(randhra ) 1 tath 2 II pfthividvayabijasya cak us3 tasyaiva yojayet I vayu unyadi- 4 bijasya5 n saka adicetasa II
amrtambusub jasya jihvendriya- 6 prabhur i6vara 7 I
griva mohabijana 8 bahubhya kle atas tath911 calita10 hrdayell de e u12 pr ibhya 13 n bhimul-
ayol').14 I

1All Mss. brahmamidam. Tib, tshans pa'i dnos. The reading -idam finds no parallel in Tib. where also we would expect tsh ;s pa'i bu ga for our qualified correction, We suggest a major corruption but one within the probability for this Text• -ida from -chidra is possible; -id from -ran­ dhram seems as likely for oral transcription and may repre­ sent a gradual process of corruption not easily traced. The reading dnos in Tib. points to Skt, rupa or maya but also to mula; below we find nibhimula at the "navel" orifice. Our solution is tentative.
A. omits. Note that the line contains only one Pada.

3c, cak u
4All Mss, 6unyadi, Tib. ston so s. The -adi here
and elsewhere appears as an unusual sign of the plural. See
also ka adi in this line,
5c. missing
6A, B. jihvendrir C, jihvendri. Tib. lee II dban por. Since ya is easily confused with pa in 0 1r script, it is likely that ya was absorbed by the following prabhur. Tib, dban po is the standard translation for indriya.
7 _,
B. l.svaratpa.J:l
aA. bijana al) 9c. missing. 100• missing.


12A. e uJ:l

anga 1akena bijanam 1 ang 1a2 dharayed3 vidhin 411
vy pi5 vyapta jagat sarv6 sthavarady sajangama I
adhara 7 bhavate 8 tasya brahma sasurasurai II
Bhavana bhagam ityahur9 yatra sa calate prabhuQ1° I karm karma11 bhavet12 tasya yavad13 dehe vyavasthitaQII kurute sarva-14 karm i15 ubhani16 va6ubhani17 va I

13All Mss• p ibhyaJtl. Tib. peyugs bdag corrected to phvu dag. Comm. ph...vugs. Loss of subscript -r is frequent.
14c. murayo
1A. bijana ; c. bijana m- 2c. aga ta
3c. dharaye 4A. vidhi
5c. the following thirty lines, from vyapi to and in­ cluding bhedanai , are missing.
6B. sarvva 7A. adhalal'!l BE. bhavete 9A. -ahu lOA. prabhu
11A, B. karmmakarmma

12B. bhavat lJB. yanad 14B. sarvva
15A. karmmani; B. karmm'ir:li
16Be /sub


17B. omits va.

yogas tu samata prokta pujan 1 Bhavana bhavet // Karma-2 deha yada3 bhagna tadrso devata bhavet I

sva aktis tu tada tasya yena vyapta sthira cala // va a tasya vijaniyad akasasadr a4 mata I
nirv e tu sthito vira nirlepa-5 malavarjita iti // matara bhaginr 6 caiva7 duhitaram 8 bandhavrn9 tatha brahm i 10 k atriy i 11 caiva12 vaisyaka lJ
udr i 14 tatha //

1B, punjanam. Note that Tib. translates rab du sbyor ba which anticipates pray,,ga; perhaps this is a translator• s gloss on the meaning of p\ija.

2A, B, karmma- JB, yado
5A, nirllepa; B. nillepa 6B, bhagini
7A, caivaJ:l

8A, B. duhitam. Edg,{lJ.lO) lists this as a BHS in­ flection for -r stemJ m tr in the same line, however, receives a standard Skt. inflection as do other terms in the series. Thus, no internal criterion emerges in support of the Mss.
as the earlier reading while it can be argued that -am re­ flects a late attempt to regularize meter. At best, the grammar of the uncorrected line suggests a bread range of inflection possibilities, including standard Skt, and BHS, but does not suggest a solution to the question of priority.

9B, ba'f!ldhavin lOB, brahma.r;ti
llA, k atrini; B. kf?atril').!J!I

12A, caival].

13A, caisikiDp; B. vaiscikal!l 14A, S'ildri in; B. s'lldri.nim.

rajak1 na ika 1 caiva qombi2 CaQ alininJ tatha I praj opaya- 4 vidhanena5 pujayet tattva-6 vatsalah711 sevitavy8 prayatnena9 yatha bhedo na jayate I
agupte kriyate du kh vya4a-10 cauradi-11 bhucarai /1
mudra panca- 12 vidha prokti kulabhedena bheditah I
brahm t13 dvijakulaja sa ca tathagata14 mata II

1A. na ika
2A. dombi ah; B, dombi JA, c alinin
4Twenty-nine of the following forty-two lines, the
remainder of the chapter, are found in HVT 1.5; three addi­ tional lines are related to that reference, Significant agreements or disagreements with Snellgrove's edition will
be noted for purposes of comparison. See Notes to Tib. edi­

5A, vidhinena
6A,B, tatva-, HVT, tattva-

?All Mss, vatsalah. Tib, uncorrected rig pas. Comm. rigs pas, HVT, vatsala ; rig pas. Tib, and HVT Tibetan point to some form of vid-, corrupted to a form beginning vat-, Comm,, however, supports the Mss, and indicates a corruption of Tib, which we correct to rigs pas. Snellgrove shows no awareness of the problem,
8B. sevisya

9A, prayetvena 10A, B, vyada
11HVT cauragni; ,,,me yis
12B, paca 13B, brahmani
14A. Tathagata

k atriy Ilrajagotri 2 mor -3 cchandradi-4 kulaja5j amrta-6 vajreti7 kathyate //
vai y B gopalik caiva sa Karma-9 kulik mata I

udri r1° vr rll caiva12 mahavairocani mata // nati1 admakuli caiva rajaki Karma-14 kuli15 tatha16j

1A, B. ks.atrin. ! 2A. gotra

JA. omits second Pada here but includes it five lines below. Tib. supports the reading of B. here and below; the meaning of the passage also supports B.

4B. -cchidra dia A below (See Note 3 aboveh -cchidradi. Tib. zla sogs.

5A. below (See Note 3 above),kaulija 6A. sa 'mitaa B. amita
7A, B. vajeti
8A. vai6a; B. vaisiki 9B. karmma
10A. sudri i; B. sudr1
11B. vr adhi 12
A. caivatt

lJA. na"t:i
14A, B. karmma l5A. kaulika 16A. omits

gombi1vajra-2 kult3 khy ta ratna4 c 9alini5 j eya // paflcamudra tu ni6cit 6 //
tathagatan 7 kula8 caitat9 s k epe a-1° bhidhi-

yate /
tathataya gata 6riman11 agata ca tathaiva ca // anaya prajffaya yukty 1 2 tathagato 'bhidhiyatelJ // kula p cavidha proktam ananta 6atadha14 kula / pascat15 trividhata y ti kayavakcittabhedanai //

lA, B. dombi 2A, B. vaja- JA. -kuli
4A. ratna. A. adds the Pada missing five lines abovea see Note 3 on preceding page.

5A. vari:lalini
6HvT vini6citah. Snellgrove edits for the plural form of the meaning. We accept the singular here and following
in a plural sense.
7A. tathagatya 8 ..,
A, kulalJffi

9A. caital1 B. caitamt. The corrected singular is translated plural by Tib. idi rnams1 this is strong evidence that the singular forms in this section actually do carry
a plural sense. See Translation,
10B. saks.epen.a- 11A. 6rima
B. yukka

B. omit Avagraha, easily lost in transcription.

B. satatha pas/ ca-n

ku1a- n-a paNnca- 1 bhu-ta- n-
a pa,n.J caskandha-2
svarup•1-a I
vajra-3 ratnapadmam idharana (prasffti)4 kulani II

nasti bhavako5 na bhavyo6 'sti7 nasti mantro na ca devata I
ti thato 8 mantradevau ca ni prapa ca-9 svabhavata II
ak obhya-10 vairocana- 11 mogharatnarolikasattvikai 121 brahma vi u13 6iva14 sarvo vibuddhas tattvam15
ucyate II

1B. paca-
A. -ska'f!ldha

3All Mss. vaja

4B. prasphUtiJ c. prasbhuti. Tib. rim pa. Comm. suggests lam. The acceptance of Ms. A. is highly tentative given the range of disagreement in our materials.
5c. Bhava(missing) 6A. bhavya
7All Mss. omit Avagraha. Edg, lists the omission as a feature of BHS; we judge it just as likely a corruption during oral transmission.
8 . . . .
A, t1 thetau; B. t1 thatau. T1b. yan dag gnas.
HVT sa stitau. Tib. suggests the form edited by Snellgrove. We judge a similar correction for our Mss. to be too ex­ tensive and correct for a form easily corrupted by the addi­ tion of a superscript. The purpose of the addition probably was a false attempt for agreement.

9B, c. ni prapa'f!lca 10A. ak obhye
A. vairocana ; c. vairocano
12A, C. satvikai ; B. sasikai . HVT satvika . 13
A. vif:?nu
14A. omits
SiVaQ2sada sukaly at3 sarva4 sarva-5 tmani sthita // satsukhatvena tattva 6 (tad)7 vibuddho bodhanad
rate}J.8 /

dehe s bhavatr-9 tyasmad devateti nigadyate //

15All Mss. tatvam. HVT tattvam.

1A. visan8daB, c. visan . The confusion of sibilants is frequent and unpredictable.
2c. 6iva

3B. sukalyat
B, c. sarvval].
5B. sarvva-

6All Mss. tatva

7A. jffoa B, c. jna. Tib. de. HVT ca; de. Tibetan materials point to a demonstrative, but the corruption is not easily traced. We, therefore, qualify the correction.
MW indicates tattvaj a to be an epithet of a Brahmana if
such a form were to come easily to mind upon hearing tattva-, the corruption might enter in that way. It is not possible
to determine the status of ca in Snellgrove's edition; but it is not likely to have been translated de.
All Mss. -ataQ. Tib. dga' ba.

9B, c. sambhavanti
bhaga sJ tu a -4 vidhanyahur ai varyadi5 -gu khilanll athava kle adi-6 k bhagnavan7 iti bhagavan II
bhaJ;lYate praJ.nNa-
J.anayar'"l t•J.8
J.analj'l yatah. I
bhaginiti tatha9 praj a vibhag dar ayed yataq10 II

1All Mss. syastiti

2A. yoyoja; B. buddhasya yoja; c. yo buddhasya ja. Tib. mna ba'J. phyir. HVT for the Pada, bhago 'syastiti buddhasya; sans rgyas 'di la bha ga mna'. Comm. discusses briefly skal ba drug dan ldan pa'i. Our correction is highly tentative. It is based upon witness in All Mss., Tib., and Comm. for a word expressing possession. The Skt. asti may serve this purpose and be indicated by Tib. rona ba; but this would be an unusual translation of asti and would leave the line metrically short. Buddhasya would then reg­ ularize the meter but is witnessed by only two Mss. and not at all by Tib.; but HVT materials provide some justification for its inclusion. Our own solution is metrically regular.

JB. bhagas
All Mss.
5A. ais"'caryya;

c • aJ....s, carya
A11 Mss. kre adi-

7A. bhakambhagnavan 8A. janayayttti
9A. bhaginitatha 10A. vatal)

rajaki bh yate 1 praj a sattvan 2 raxjanad3 yata 1 duhita bha yate praj a gupan4 duhanad yata II nartaki5 bh yate prajrta c calatvat mah krp 6 I aspar6a bhagavatr 7 yasmad 8 gombi9 tasmat10 pra-
kathyate11 II
jap jalpanam12 akhyatam 13 alikaliprajalpanat14 I

lB. bhavyate: c. bhanyente 2All Mss. satvana

3skt. is translated 'tshod pa carrying the meaning of "cook" but also of "boil" related to the activity of dying. Nevertheless, we might have expected tshos, "dye." We note this because Tib. adds here the quarter-versea tso blag (correction of bslag) rna zes de bzin brjod, as if the meaning of "dyer" were not already covered. HVT Tibetan also addss gtso blag rna es de bfin brjod, a quarter-verse corrupt as
it stands certainly but apparently attempting to make the same point. While Snellgrove does not edit correctly nor qualify its place in the Tibetan edition, he is no doubt
correct in excluding the passage from the translation. The Skt. materials are regular as they stand; and the addition of Tib. may be a translator's gloss, a feature not absent in our Tib. version.

4B. gw;tanan a
c • gurnana-n
5A11 Mss. narttaki
6A. mahak a
7A. bhavati
8A. yasyada B, C. jasmad 9All Mss. dombi
1°B, c. tasman
11c. -kathyateh 12c. yalpal)am
13All Mss. khy ta
14A. prajapanat

m al padalekha 1 syan2 malanad3 maQ alam4 ucyate5 //
karaspho o bhaved6 mudra angulya7 mo ana tatha I
tad dhyeya cintit8 yac ca dhyeya yasmad9 vicintanal]l //
pitari prapta yat1° saukhy11 tat sukhal]l bhujyate

maranam yena sukheneha tat Sukha Dhyana ucyate // ityabhi-13 dhanabhidheyabodhicittot- 14 padadi bhav­
ana-15tattva-16 prakar a prathamal]l //

1A. -lekha
2A, syat

3All Mss. malanat 4A. man.d.alarn.
5A, vadetJ c. ucyato
Al1 Mss. bhavet
7A, sungulya1 B, ailgufljalya; C, angunjalyal!l 8
A, ciJ!ltitalp
A, yasmataa B, smad1 C. yasyad

10All Mss, yatha. Tib, de yi. Regular meter is also provided by our correction.

11Be saikhYa.J!l 12All Mss. svayam

13B, c. iti abhi- 14c, Bodhi(missing)
15c, bhavaJ).a
16All Mss, tatva

(Chapter 2)

(bhagav n aha)l //
atha saptatri 6ad-2 bodhipak ik n3 -dharman4 kathayi yami5 //
adhyatmakaye bahirdha-6 kaye? adhyatmabahirdha- 8

1All Mss. omit. Tib. omits. We are not claiming
the phrase to be a part of an original text since the sup­
port is lacking. It is added here -- and elsewhere -- in a qualified way to aid the translation. It is also added in the spirit of what The Buddha is said to have announced in the Dharmasailgitisutra as the style of a siitraa "Men­ dicants, my Sacred Words (pravacana) must be introduced with the phrase, 'Thus I have heard on a certain occasion
(evam rna a: srutam ekasmin sama e),'" and so on; and, "At
intervals appropriate connecting phrases should be in­ serted." Mkhas grub rje's "Fundamentals," p. 55.
2A, c. -tri ad 3A. -sayak ikaTQ
A. dharmrna ; B. dharrnma; c. dharmman 5A. kathayayi yami
6All Mss. bahirddha. Unanimity of reading may point to a BHS form although this particular word is not listed in the Edgerton lexicon. Below, however, the form is reg­ ular in Ms. C, a fact which weakens a BHS argument.
7A. adds bahirddhakaye 8A, B. bahirddha

kaye kayanudars11 viharatityapi 2 sa prajanan3 smrtiman viniya loke avidy -4 daurmanasye I adhyatm 5 vedana6 ba­ hirdha7 vedana8 adhyatmabahirdha-9 vedana1° Vedana-1 1

nudarst viharat tyapi12 sa prajanan sm timan vintya13 loke14 Avidya-15 daurmanasye 16 I adhyatmadharme u 17

2A, viharatity JA. prasatpjanan
4A. avidhya; B, C. abhidhya

5B, asma 6A. vedanah
7All Mss, bahirddha
8A. vedanal].

9All Mss. bahirddha 10A. vedanal)
11A, B. omit
12A. vihatnyapir c. viharattsyapi 13A, viniya
14A, lokel;t
15A, avidhya; B, c. abhidhya 16A, daurmanasyel)
17A. dharmme u.l').; B, c. dharmmes.u

bahirdhadharme v- 1 adhy tmabahirdha- 2 dharme u3 Dharma-4 nudar i viharatityapi5 sa prajanan smrtim n6 vin!ya loke Avidya-7 daurmanasye8 I adhyatma-9 citte bahirdha-10 citte adhyatmabahirdha-11 citte citt nudarsi viharat tyapi12
sa praj nan13 smrtiman vin!ya14 loke Avidya- 15 daurman­
asye16 I im ni catv ri smrtyupasthanani //
anutpannana 17 papakanam 18 aku alana 19 dharm m20

1A. bahirddhadharmmesuh; B. omits; c. bahirddha- dhammai u • •
2All Mss. bahirddha

JA, dharmme uQ; B, c. dharmme u 4All Mss. dharmma-
5A. viharatityapi; B. vihatityapi 6A. sm:r;timana.t;t

?All Mss. abhidhya 8c. daumanasye
9A. omits line to imani catvari. 100• bhidya
11B. bahirddha; c. bahirdhya
120• viharepati

14c. viniye

15B, c. abhidhya 160• daumanasye 17A. anutpannana 18A. papakarnal'_l

anutpannaya-1 cchand janayati 2 vyayacchati3 viryarn4 arabhate5 I citta6 prag h ati7 sarnyakpr idadhati 8 I
papakanarn akusalanam.9 dharrnan.arn.lO prahan.aya-11
cchand janayati 12 vyayacchati viryarn13 arabhate I citt
prag h ati 14 sarnyak-15 pr idadhati16 I anutpannan l7

19A. aku alanarn.ah.r c. akusaranarn. 20A, B. dharrnrnan.arn.; c. dharrnmanarn 1A. anutpannaJ:l
2A. janayatil;t JA. vyayacchatih
4A, c. viryya

5A. arabhate.t) 6B. Citta
7A. pragrhnatil;tr B. pratig hnati; c. pragrhnati 8B, c. -pranidadh ti
9A. aku6alanal!l
10A, B. dharrnrnana : c. omits 11A. prahanayalf: c. prahanaya 12A. janayatil)
13A. viryyarn; c. viryyarn
14A. prag:rhnatiJ:t: B. pratigrhnati: c. prag hnati 1.5c= sarprnyak-

16A. -pra idadhati.t) 17A. anutpannanam.ah.

ku alana dharm am1utpadanaya- 2 cchand -3 janayati4 vyayacchate viryam5 arabhate I Citta 6 pratig ati7 sam­ praryi-8 dadhati I evam9 utpannana ku alana lO dharmana ll sthitaye121aprameyopayabhavanaparipuraye13 punar Bhavana­ vrddhi-14 vaipulye- 15 cchanda16 janayati17 vyayacchati 18

1A. dharmman.amah.; B, c. 2A. utpadanaya ; C. utpad aya

3c. cchaytldan- 4A. -janayatil} 5All Mss. viryya
61\. cittar.nah.
?A. pratigr.hnatih.; B, c. pratigrhnati
8A. sarnpani-; c. sarnprani-

9c. eva

kusala-na -al;l

llA, B. dharrnrnan.arn.; c. dharrnrnanal'!l 12A. sthitaye.t:
13A. paripuraya 14A. vrddhira
15A. vaipulyal).; B, c. vaipulya

16B, c. cchada 17A. janayatil). 18A. vyayacchatil}

viryarnl arabhati I Citta pratigrhryati2 sampranidadhati I

imani catvariJ samyakprah ani //
Chanda-4 sarnadhiprah a-5 sa skara-6 samanvagata 7 fddhipada bhavayati 8 I vivekaniscita 9 viraganiscita 10 nirodhani6cita 11 vyavasargapari ata 12 m rnecchando 'tilino bhavisyati 1J natipragrhita iti I Virya-1 4 Samadhi-

1A. viryya; c. viryya
2A. pratigrhnati ; B, c. pratigrhn ti JA. catvari
4A, c. ccha da; B. cchanda. The consonant doubling seems to have derived fran an initial corruption of punctu­
ation; Mss. B, c. fail to close the line preceding.
5A. prah am; c. prahana 6A. skara; c. saskara 7A. samanvagata a
A. bhavayatiq

9All Mss. nisrtam. Tib. nes par 'byun ba. Tib. def­ initely points to our correction which could have been cor­ rupted easily in oral transcription. The repeated reading of nisrta in lines following should not be taken as support for the reading but probably as a kind of regularization based upon the initial corruption.
10A. nisrta a ; B, c. nis ta 11A. nisrta a ; B, c. nisrta 12Ao pari ata a ; Co parinata lJA. bhavis.yatih.
14A, C • v1-ryya

praha a- 1 sa skara- 2 samanvagatam3 :rddhipadam4 bhavayati5/ vivekaniscita 6 viragani cita 7 nirodhani cita B vyava­ sarga-9 pari ata 10 ma me viryam1 1 atil!na bhavi yati 12
natip a rhitam iti13 I mima sa-14 samadhiprah a-15 sa ­
sk rasamanvagatam16 ddhipada bhavayatil? I vivekaniscita lB

lA, prahanal} 2c. saskara
3A, samanvagatal)m 4A, -pada
5A, bhavayatil;l
6A, nis;:-tayt1al;lr B, c. nis:rtaf!l
?A. nis;-ta!l)al); B, c. nisr.tam.
sA. nis!'tall}a r B, c. nis;-taiT} 9B, C, vyavasargga
1°A. pari atayt1al); pali ata , c. 11A, viryya ; B. Virya; C, vtryya 12A. bhavi yati ; B, bhav! yati
13A, pragrhita itir B. natipratigrhita iti; C, nata- g;ohi"tal]l iti
14A, C. mimaiT}Sa 15A, prahane
16A. samanvagata r C, sam vagatam
17A. bhavayatil}
18A. nis;-tamaJ:l; B, c. nis;-ta

viraganiscita 1 nirodhaniscita 2 vyavasarga-3 pari ata4
rna me mima satilina5 bhavi yati6 natipragrhiteti7 I Citta-8 samadhiprah a-9 s skara-10 samanvagatam11 ddhipada 12 bhavayati 13 I vivekani6cita 14 viraganiscita 15 Nirodha­
niscita 16 vyavasarga-17 pari at 18 rna me cittam linam
bhavi yati19 natipragrhitam 20 iti I ime catvara21 rddhi-

1All Mss. omit viraga ni6citam. Tib. 'dod chags dan bral bas nes par 'byun ba dan. The"pattern of the section favors Tib.
A. n1sr.tam.ah.; , • n1sr.tam. 3All Mss. vyavasargga
A. pariryata a

5A, C. mima sa 'tilina; B. mimamsa 'tina 6A. bhavi yati
7A. natinatipragrhfta itir B. npratig h!ta itir
c. natipragrhita iti .
A. cirtta
9A. praha a 10c. saskara
11A. samanvagata m; C. samanvagata
12All Mss. rddhipada 13A. bhavayati
14A. nisrta a ; B, c. nisrta
l5A. omits viragani6cita ; B, c. viraganisrta
A. nisrta a ; B, c. nisrta 17A, c. vyavasargga
A. pari ata a

pada sakamavacara l II
lokikY2 samyakd tim 6raddadhati3 svakarma-4
Vipaka-5 pratisar o bhavati I yad eva karma6 kari yami7 kalya a 8 va papaka 9 valO tasya Karma oll Vipaka prati­ vedayiqyami12 I sa jivita-13 hetor api14 papaka 1Rarma16
na sa skaroti17 I idarn ucyate18 raddhendriya II

19A. bhavi yatil;l
20A. pragrhitam; B. natipratigrh!tam
21c. catvari
1A. -avacala ah; B. avacala ; C. avacara 2A. lokiki
JA. addhadhati
4All Mss. svakarmma- 5A. vipakal;l
6A, c. karrnma

?A. kari yami ; c. karyasyami 8B. kalyiD;ta
9All Mss. papakam 10A. va
11A, B. karmmaQo; c. karmma
12A. prativedayi yami ; B. prati vedayi yami 13A. jivita
14A. apif:l 15c. papaka
16A. karmrna a ; B, c. karrnm a
17A. sa skaroti; c. saskaroti 18B. ucyata

yanl dharm 2 sraddhendriye a sraddadhati3 tan4 dharman5 v!ryendriye a6 samudanayati I idam7 ucyate viryendriya 8 //
yan dharman9viryendriye a 10 samudanayati11 tan12
dharmanlJ sm t!ndriye a14 vipraQatayati15 I idam16 ucyate smrtindriya //
yan dharman17 smrtindriye a1 8 viprapasayati 19 tan

1All Mss. ya. The conjunct nasal is easily lost in transcription.
2A, c. dharmman; B. dharmana JA. raddadhati
4B. dhan

5B. dharmmana ; c. dhamman 6
A. viryyandriye ar c. viryyandri a
7A, c. ida
8A. viryyendriya ; C. viryyandriya 9c. dharmman
10A. viryyendriye ar c. viryyandriyena
11A. samudanaryati 12c. ta
13A. dharmmat; B, c. dharmman
14A, C. smttindriyenar B. smrta driyana

15A. viprana aryatir B. vipranasayati; c. vipranasayati 16c. ida
17A, B. dharmmanr c. dhasma 18c. smrtyandriyena
19A. viprarya6aya ti ; B. vipr asayatir c. vipranasayati

dharman1 samadhindriye a 2 ekagran3 karoti I idam4 ucyate5 samadhindriya 6 II
yan7 dharman8 samadhindriye a9 ekagran1° karoti11
tan12 dharman13 praj endriye a 14 pratividhyati I
sa te9u te9u dharme9u15 pratyavek9aQajatTyo bhavati I

idam16 ucyate praj endriya 17 //

1B, c. dharmman
2A. samadhindriye ah; C. samadh!ndriyena 3All Mss. ekagr!
4A. idal]l

5B. ucyal]lte 6
A. samaddhindriyal]l; C. samadhindriya
7A. omits
A. dharmman; B. dharmmanr c. dharmma 9A. -Indriye al;l
lOA, c. ekagri; B. ek gr'i
11A. taroti
12A. kan; c. ta 13B, c. dharmman

14A. praj andriye ar B. praj edriye a 15All Mss. dharmrne9u
l6A, C. idaJJl

17c. praj andriya

ekani pa cendriya i1 pa ca-2 balani bhavanti3 II
tadyatha I raddhabala 4 Virya-5 bala 6 sm tibala 7 samadhibalam 8 praj abala 9 I imani p ca-10 balani II
tatra katamani saptabodhyangani11 I tadyatha I smrtis bodhyang 12 Dharma-13 pravicaya-14 sa bodhyanga 15 viryasa bodhyanga 16 pritisa bodhyangam 17 pra rabdhi-18
sa bodhyanga 19 samadhisa bodhyanga 20 I upek sa bodhyan-












1oc. 11B.






pancandriyani ; B, pa candriy i; c. pancandri-

pa ca-

C, bhavati

-balam.ah. C, viryya­

-bala al;l

-bala al;l

-bala al;l

-bala a ; B. -balah pa ca-
bodhya gani; c. bodha gani

bodhyanga ; B. sambodhyajrta; C, sa bodhya ga

c. dharmma pravica
sabodhyaj a ; c. sa bodhya j jna

viryyasa bodhyangam; B. viryasa bodhyajna; C,

-sa bodhyaj a; C, -sa bodhyaga


prasraddhi; B, c. prasradhi

ga l viveka-2 ni cita J viragani6cita 4 nirodhaniscita 5 vyavasargapari ata 6 Dharma-? pravicaya 8 saptabodhyanga­ din9 bhavayet II
Arya tango1° marga 11 I tatra katama samyag-12
dr9tir13 1 ya iokottar14 natmadr9 isamutthita15 na jivo16 na sattvo17 na pudgalo 18 na po9o19 na manujo 20 na karako21

19B, c. sa bodhyaj a

B. sa bodhyaj a ; c. bodhya jna
1B. sa bodhyajfla ; c. sa bodhya j a 2B. vivaka
JA. nis ta a ; B, c. nis ta
4A. nis ta a ; B. nisrtani; c. nisrta 5A. nirs.tam.ah.; B, c. nisr.tam.
A. vyavasarganisrta a pariryata ; B, c. vyavasarg­ gapa£i ata . Tib. adds apparently by way of a gloss• kun nas non mons pa dan bral ba'i kun nas.
7All Mss. dharmma

BAll Mss. pravicayadi. Tib. rab tu 'byed pa. The corruption -adi which lacks inflection seems to be a repe­ tition of the -adi following.
9B. bodhyagadin; c. bodhya gadin 10A. aryyastango; C. aryya9ta gom

11B. margga ; C. rnargga 12All Mss. sarnyak-
13c. -dr.s.t.i
14All Mss. 'n trna- • Our correction reflects Tib. which treats the negative as applying to the verb; at the close of the section, the negative is removed from the verb as ex­ pected. The pattern in Tib. follows the pattern of posi­
tive and negative explanations found in Skt. following.

na vedako1na sa6vato 2 nocchedo na bhavo3 na vibhavo4 1 kula-5 vyakrto 6 dr tir7 I yavan8 na sa s ro9 na nirv a 10 d i-11 samutthita12 (yavan1rla sa sara na parinirv rya
dr ti-14 samntthita15) iyam ucyate samyagdr ti}} 16 II
- --
15A. samut"'.. d. tat'tlal;l

16A. j!vo

17A. satvof;l; B,C. satvo

18A. pUl1galo ; B, c. pungalo 19A. posoJ:l
20A. manujoJ;l
21A. karakol;l: c. kalako 1vedakoJ;l, A.
2A. s"-a-"s-catol;l
3A. bhavo ; B. vo; c. bhavova 4
A. vibhavob; B. vibhava

5All Mss. nakula-. Tib omits the negative. The sense of the passage supports Tib. See Translation.
6A. vyakrtel;l; B. vyakrta

7Tib. adds here kun nas bslan ba dan, which may simply be. a glossQ
8c. yava a 9A. s8.!f1sarol) 10c. nirvanal!l
11A. dista; B, c. disti
12A. samuthito; B. samusthita

13B, c. yavat. The phrase from yavan to samutthita is not supported by Tib. and may be a repetition withal­ teration of the preceding phrase.

yai s kalpai1 Raga-2 dve Amoha-3 kle6a4 samut­
tis.t.anti5 tan sam.kalpan6 na sam.kalpayati I yaih.

sa kalpaih- 7
6ilasamadhi-8 praj avimukti-9 vimukti-1° j anadar,ana-11
Skandha 12 samutti anti13 tan s kalpan sa kalpayati14 1
ayam15 ucyate samyaksa 1n.e.l p3_ t1 II

14All Mss. di "t;i 15B. samusthita
16A. samyakdi i; B, c. samyakdistil]. 1A. sam. kalpeh.
2c. Raga

3A. mohal',l 4c. Klesa
5A.; B. -tis.t.ati
6A. samkalpa; c. sakalpan 7c. sakalpail',l
BA. Samadhi}].

9A. vimukti}J. 1oc. vimukta 11A. dar anav
12All Mss. skam.dh-ah.
13A. -ti ta antil',l
14B. asamkalpayatir c. asakalpayati 15c. ayaf!l

ya vagl atmana paran2 va3 na tapayati4 na kle,ay­ ati5 nopahasati taya sa6 samahita-7 yuktaya vaca8 saman­ vagato bhavati I yaya vaca9 samyagaryanlO avatarayati iyam ucyate samyag-11 vak II
yat Karma-12 kr a kr avipaka 13 tat karma14
nabhi-15 sa skaroti16 I yat17 Karma-18 ukla 19 ukla- 20 Vipaka tat karma21 nabhisa skaroti 22 1 yat karma23 ukla-

1B, C. vak 2A. para

3B. unreadable
4A. taQpayati; B. tapayani 5c. omits na kle6ayati
6 .
All Mss. sa
7A. samahitah
BA. vacaJ:l; c. 9A. vacal).

lOA. -yam; B. -tayam; c. 11 A. samyavagvak
12All Mss. karmma
13A. vipakaii}aQ. 14All Mas. karmma

15All Mss. abhi-. The corruption derives from the confusion of inflection of Karma preceding and a negative particle following; Karma is being treated erroneously as instrumental, with subsequent loss of the negative,

16c. saskaroti

17B, omits the next two lines• from yet to and includ­ in -samskaroti.

kr avipaka tat karma1 nabhisa skaroti2 I yat3 karma4 sukla-5 kr avipakak ayaya vartate6 tat Karma-7 bhisa skar­
oti8 I sat karma9 pratisarana samyakkarmanta 10 I ayam
ucyate11samyak- 12 karmantaQ13 II

18A, c. karmma 19c. sukra!Jl
20A. uklarmma; C. sukra
21A, c. karmma
22A. abhisa skaroti; c. abhisatkaroti 23A, c. karmma

1A, c. karmma
2A. abhisa skaroti; c. abhisaskaroti 3All Mss. ye
4All Mss. karmma

5c. ukra
6All Mss. varttate 7All Mss. karmma-
8c. saskaroti

9All Mss. karmma
10B. samyakrmmanta ; c. samyakkarmmanta 11A. ucyatel)
12A. sal]lyak

13All Mss. -karmmantaq

yadarya(vat sa)1 dhutagu a 2 salina-3 nusak anata4 na kuhanata na lapanata5 na kathasurata svacara-6 ilata7 paralokesvari- 8 yuktata2 tmalabhatmasa tu iniravadyat­
Arya-10 nuj atajivanata I ayam ucyate11 samyagajivaQ //

yo vyayamo mithya- ryair12 ananuj ato13 ragadve a­ mohakle6an na14 sevitavyan15 icchati I yo16 vyayamary 17


1A. vamsa; B.

c. vasa. T1b.

lta bur.
points to -vat which could easily be corrupted to va ; a may then represent a confusion of sibilants. Yet,
vamsa is meaningful in the line. Without additional ma­ terials, we qualify the passage.

2All Mss. gur;1a

3A, B. sa le6a; C. sale a. Tib. zum preceded by the addition y dag pa'i chos, apparently as a gloss. The correction is major and only slightly traceable from the corruption; but the reading of the Mss. is almost meaningless.

4B, c. anusark a ata. Tib. dul ba does not readily anticipate the Skt. form which is, nevertheless, attested by MW as sak9ana.

5A. yanata; c. lpa
6All Mss. svaca. Tib. ran gi spyod pa'i 7A. srilata; c. silata
Be. lokescar'i

9A. yuktata 1°A. aryya- 11A. ucyate})
12A. aryyar; B. arye; C. ayair
13A. anantajnatah; B. anantaj ata; c. anantaj ata 14A. omits; B. a-; c. a-. Tib. mi 'dod pa. Tib sig-
nals a negative associated with the verb; na may have been
absorbed by the preceding nasal termination.

samyagarya-1 marga-2 satyavataro nirva a-3 gam! prati­ padam aryapyati4 ta vyayama 5 samanucarati 6 I ayam ucyate samyagvyayama II
yasupasthitaprakampa-7 rjukaya- kutila sa sara-8
do adin avadar ika9 nirva apathaprarTta-10 smrti 11 I

15A. sevitavyat 16A, B. so
17A. vyamai:t

1A. Arya

2B, c. sarga-

JA. nirvot;1a
4B, c. arppayati

5A. vyayamaJ!l

6All Mss. samanucchati. Tib. yah dag par rjes su 'bran ba. There seems to be confusion with icchati appear­ ing in the preceding line.

?All Mss. -aprakamppa 8A. Sal]lsala

9Tib. for the phrase, dban du byed pa ston. The addition may be a .gloss but also may reflect an c ••1mision in Skt.

10All Mss. pra eti. Tib. 'dod pa'i. The Tib. reading is questionable, but the Skt. may be also; the translator has felt it necessary to add the gloss, mi brjed pa.

11A. sm.rtir

ayam1 Arya-2 marga-3 sa prayoga4 I iyam ucyate samyak­ smrtit;t5 II
yasmin samadhau sthitaQ sarva-6 sattva-7 vipra­ mok aya samyaksthiratva 8 nirvaryam avakramati9 I ayam ucyate samyaksam dhi II
eta eva samakhyata10 muditadyas11 tu bhumaya 1 2 II
alayo13 bodhisattv am14 indriyadisvarupi am15 1

1A. ayaJ!laJ:l; B, c. ayaJ!l 2A, c. aryya
3c. maga

4A. satpprayoga

5A. -smrti 6c. sarvva
7All Mss. satva
8c. sthirala 9B. nirvfu:lamati lOA. samakhyataJ:l 11A. muditayas 12A. bhiimaya 1Jc. alayo
14A, -satvanam.ah.J B, c. -satvanam
15B. svarupirya

Alaya sarva-1 buddhan 2 skandhadina 3vi e ata 4 // buddhanam.5 bodhisattvanam.6 buddhatvabhik 7 parah8/ dvatri ad-9 bodhicittan ya ihayati10 pu ata11 I

indriyaskandhabhutan 12 Buddhatva tad anantar // svadehe caiva13 Buddhatva 14 sthita15 nanyatra
kutracit16 /
dehad anyatra buddhatvam 17 aj anenavrta mata //

1A, c. sarvva- 2B. -buddhana
3A. sk dhadina ; B. ska dhadina ; c. ska dhadini 4A. vises.atah.
5c. budhana
6A. bodhisa.tvanal!lal;l; B, c. bodhisatvana 7B, c. buddhatvavabhika
8A. paraey; B. pana

9B. -trif!ldva

lOAll Mss. ihayanti 11All Mss. pur atal!l 12A. bhutanaif
13A. caiva}f; B. ceva 14c. baddhatval!l
15A. sthital!lah 16B, c. kutracita 17c. buddhatvam

svadehastha1mahaj ana 2 sarvasa kalpa-3 varjita / vyapaka sarvavastuna 4 dehastho5 'pi6 na dehajaq //
vajragarbha7 uvaca // dehe8 katama n qya 9 //
bhagavan aha //

atarn eka lO dvida adhika ll catu cakra-12 prabhe­ datal)13 /
bodhicittasvarupe a1 4 nagr-15 dvatri at-16 tama mata17;;

lB. -stha 2A. j an a 3B. kalpa
4A, B. vasthunaytl; c. vasunam

5A. sto; B. stha 6All Mss. -pi 7All Mss. vaja- sA. deha
9All Mss. nadya})
1°B. satakam.; c. llA. dvi6adhikam.ah.

120• catu-

13A. prabhedata; c. pradabheta 14A, c. svarupena
15A. nadi; B, C • nadi 160• dvatrisat
17B, c. mata

tadyatha //
abhedyat mata suk rna- 2 rupadi divya varna tu vamani3 I kurrnaj 4 bhavaki seka do a vi tha ca tu mata5 II arvarr6 !ta-7 do ma ca lalana rasanavadhnt!8 I
prav a h a9 ca va a1° samanyall hetu-12 d yika // viyoga prerna i siddha pavaki sumana tatha I
t varta13 karnini geha14 c gika15 maradarika 16 //

1The thirty-two Channel and following short dialogue are found in HVT. 1.1.16-20a. Significant differences and agreements can be found in the Notes.
2All Mss. suk ma

3All Mss. damani. Tib. thun nu rna. HVT. thun nu rna. 4All Mss. kurmmaja
5A. ca tu m ra; B, c. ca matura 6All Mss. sarvari
7All Ms s. S'!ta-

8A. rasana 'vadhuti

9Tib. rahs rna dan. HVT k a; rens rna dan. Snellgrove fails to discuss the discrepancy in his edition.

lOA, c. va a; B. var-unreadable. Tib. mdog rna dan. HVT va a; sin tu gzugs can rna.

11A. sarnanya; B. samanya 12A. hetu ; Co hetu
13A. trvartta; B. n vartta; c. travartte
14A. gehal) 15c. candika
16All Mss. dayika

(vajragarbha uvaca1)//
eta naqyo2 bhagavan kidr as3 tribhavapari ata // (bhagavan ha4)//

sarvagrahyagrahakavarjita 5 // iti(catusmrtyupasth acatusamyakpraha acatv ddhi­
pada-6) pance-7 ndriya-8 panca-9 balasaptabodhyahgarya­
anga-10 marga-11 paryantadi-1 2 vivarana Bodhicitta-13 vataro nama dvitiya-14 prakar a 15 //

1The connecting phrase is added by the Editor to facilitate translation.
2All Mss. nadyo

JA. kidrsa s; c. kidrsa

4The connecting phrase is added by the Fditor to facilitate translation.

5All Mss. varjita

6All Mss. omit from catu- to and including rddhipada. Tib, dran pane bar bzag (corrected) pa bzi dah /'yah dag
par sbon ba bzi dan I rdsu 'phral gyi rkan pa bzi dah I

7B, c. paS'ca 8
A. -endriya}f

9B, c. pa!Jlca-

10A. bodhyangaryya9t ga; B. bodh garya ga; c. bo- dhyal!lgaya ?tal!tga
11Bo C. margga
12A. paryyantadi ; c. paryyatadi 13c. bodhicita
14B. dvitiyalJl 15c. prakaralJa

(Chapter J)

atha bhagavanta sarvatathagata puja krtva pr i­ patyaivam ahu l //
bha asva 2 bhagavan sara rahasya 3 jnanam uttama //
atha bhagavan sarva-4 tathagatadhye ana viditva sarvatantra-5 hrdayaj ana- 6 pradipannama-7 Samadhi sama­

padyeda 8 sarvatantrarahasyam udajahara9 //
rahasye1° parame 11 ramye12 sarvatmani13 sada sthita14;

1A. aha

2B, c. bhaf?aya 3B. ram.hasyam. 4A. sarva
5A. tam.tra 6c. j anal).
7A. pradipanama. Tib. here reads mar me rdo rje fes bya ba'i; we insert the addition of Tib. in a qualified way in the Translation.
8B, c. samapadyada 9c. udajaharal'].
lOB. rahasya 11A. paramyef:t
12A. ra'ftlmye; c. ramye

13A. sarvanmani; B, c. sarvvatmani 14A. sthita}f

sarvabuddhamaya sattvo1 vajrasattvaq 2 pra sukh 3 II
asau hi bhagavan yoga sthitaparamasasvata-4 para I manmatha-5 pratyutpannas tu6 svabhavo dur-7 atikrama ll vicitrakarma- 8 yogena9 vicitravidhika k in m1° I
buddhavajra- 11 dharadyas tu12 krtakaritaya smrtaQ II
sarva-13 buddhadisthiracalasarvabhavo bhavatyasau14 I
sarvabuddhasamayoga 15 kini-16 jalasa vara I
anena maya-17 yogena sarvatovi vam 18 uttama //

1A, B. satvo; c. sato 2All Mss. vajasatva

3c. -su a 4A. S'asvatah
5B, c. matmatha 6A, c. tuh
7A. due-; c. dul- 8All Mss, karmma

9c. yogenal) 10A. kak ina 11All Mss. vaja 12c. tu
13B. sarvva
14A. asauh.

15A. omits from samayoga to and including siddham sarva, two lines below. B. samayoga
16B, c. dakin'i 17c. maya
18c. visya


sarvaeattva-1 rtham
sarvastrim yaya siddha2 svarupaparivartanai II
vicitramayamudreya 3 akini-4 ti ca5 mlecchaya6 I dyau-7 vihayasagamane 8 dhatur atra vikalpita 9 II sarvaka acara- 10 siddhir gakin!ti11 prasidhyati I sarvatovisvamudra tu12 sarvatovi vas varair iti II vajra 13 vajra- 14 dhara caiva15 padm 6 padma-

1All Mss. sattva 2A, c. siddhamal').
JA. mudreya a ; B. mudraya; c. mudreya 4All Mss. Dakini
5B, c. omit 6B. ji{occhaya
7All Mss. dau. Tib. transliterates imprecisely,
tai bai ha ya sar, and then translates the remainder of the phrase, 'gro bar.
8All Mss. gam e

9A. vikampita; B. vikalpeita 10A. cara ; C. rana
11A. dakinitil').; B. Dakini; c. dakiniti
12A. tul').
13A, B. vaja; c. vaja 14All Mss. vajra
15A. caiva 16
C • paral!l

l?A. dhal!lras

maQir m idhara caiva bhavantye arn 1 kulani ca II
atha sarva-2 tathagatabhibhavanavirajapadan3 nama Samadhi samapadyeda4 Bodhicitta udajahara5 //
na unya napi6 casunya madhyama nopalabhyate7 I
praj aparamitayoga(d hy-8) upaya 9 karu atmana10 1 tata 11 sukaru o-12 paya praj aparamitasphu a 13 // avikalpe u dharme u14 na bhavo na ca Bhavana I
avikalpadhi- 15 mok e api16 kalpayet sarvakalpan //

1A. bhaval)'ltyefllarp; c. bhavansyeqal}'l 2c. sarvva
3c. vadarp. 4A. -eda al;t
5B, c. udrajahara

6A. vapi

7B. papalabhyate; c. parabhyate

8A. yogah-indecipherable; B. yo a-indecipherable; c. joga-indecipherable. Tib. sbyor ba. The indecipherable conjunct character is written differently in each Ms; the the consonant -h, however, seems to be present. Since Tib. translates only one word, we consider a tentative solution by way of a particle a likely possibility.
9B. upaya
10A. -atmanarpa 11A. tata
12A. sukuru o-

13B. spha am; c. sphuta 14All Mss. dharmmefjlU
15Tib. reads rtag indicating Skt. sada here. 16A • ap1".h

avikalpe u dharme u1 sattva-2 rthaparikalpana I tathagatatmaka 3 dharm4 na dharmi5 na ca dharmata6 I prati rutkaravasamat 7 tato8 'sau9 Dharma-1° vag bhavet II
tato Mahayana- 11 sambhavabhavanagu avistarai trai­ yadhvika-12 sarva13 Tathagata anena14 stotra-15 rajena

sa stuvanti16 //

1All Mss. dharmme u 2All Mss. satvartha-

3c. -atmaka 4
A. Dharma a ; B, c. dharmma
5A. dharmi; B, c. dharmmt 6All Mss. dharrnmata
?A. sanat
8A. tamo; C, tatau 9A. sai; B, c. sau 10A, c. dharmma
11c. maha1Jlsna
A. -yadhvika ; B. -yadhika
13c. sarvva 14A. aneka
l5c. strotra
16A. sastuvantisma; B. sa stuvantisya; c. sastuvan­

namas1 tu yogadhipasattva- 2 mocakaya3 namas4 tu sarvatma-5 jaika-6 bhavakaya I namas7 tu sa sarar ava-8
mohacchedakaya 9 namas tu sarvatattva- 10 j anaikadar6akaya 111 punar api puja 12 krtva pr ipatyaivam 13 ahu 14 II
bha asva15 bhagavan sara vigrahal]l /1
bhagavan ha II
sarva-16 dharmaika- 1 7

yadyad indriyamarga-18 tval]l ya yat tattva-19 svabhavata l

1A. namo; B. nama 2All Mss, sattva

3A, mocatah; B, C, mocaka. Tib, shows the dative case as expected here.and below,

4All Mss, namo, This corruption, repeated below, seems to derive from the frequency of the corrupt form in
Skt, texts and infrequency of the correct form for this con­

5B. sarva mar c. savanmi
6All Mss. ja eka 7All Mss, namo
All Mss, -ar ava
9A. -cchedaka ; B. -ccheda a ; c. -cchedaka 1°A, satva; B, c. tatva
11All Mss, dar aka 12c. puja



B. -padyevamJ c. -patyavam


bha aya sarvva

asamahitayogena 1 nityam2 eva samahitav //
yasmat3 sarva-4 tmanatmani5 vij ana 6 skandham a{­

rita 7 /
na kecit8 pratibudhyante9 mughacitta1° nara-11 dhamah// Citta caitta12 ca13 cidrupa vij ana j eya-14 sva­
rupakaVl I

17A. dharmmeka; B, c. dharmmaika 18B, C. margga
19All Mss. tatva 1A. yogena.t).
2A. nityem

4c. sava
5c. amani


• .If,)-.,.

7A. a-

_, .
B, c. asr ta

sB. kacit

9A. budhya te 10c. Citta
11B. nana-
All Mss. Citta . Tib. sems byun
13A. ci
14A, C. jt{aya

karye u1 karanatva ca dharm a2 Dharma-3 Bhavana II yasmad4 Dharma-5 bahutvena6 nadi6rotam7 ivodadhi8 I mok o9 hi1° eka eva11 syad12 bahutva 13 nopalabhyate141

guruvaktrat15 tal16 labhyate1 7 yathadhyatmarp18 vyava­ sthitatl9 II
iti tattvodd e6a-20 prakara:r)a trtiyarp21 /I

1All Mss. kaye u. Tib. 'bras bu rnams la 2All Mss. dharmmana

3All Mss. dharmma
4A. yasmat; c. yasma 5All Mss. d.harmma 6B. bahubhyena
7A. srotar; C. 6ratam 8c. ivodadhi
9c. mok a
10B. sti; C. di 11A. va
12c. sya

13A. batva ; B. dhahutv 14c. noparatyate
15A. bhaktat
16c. -t ll- 17B, c. labhete
18B. -adhatmam; c. unreadable
19c. vevasthitaJ!l
20All Mss. tatvoddesa 21B. trtiya


(bhagavan aha)1 //
(Chapter 4)
athata 2 sa pravak yami3 sarva-4 tantraikanir aya 5 / samatvat sarva-6 sattvana 7 tryasram8 udararupi a 9 // adharaslO tu bhavet te9a1l brahmadina 12 ca sahita13 I

saiva14 praj aparamita sa v takararupini //

1The connectin phrase is added by the Editor to facilitate translation,

2B, c. athata

3A, sa pravakf'jlarni}) 4c. sarvva
5A, nirJ:laYaJ!lal:l 6B, sarvva
7A, B, satvanaJ11; c. satvanal!l sA. tisram
9B, c. rupi!)a})
10c. adhara llc. bhavetesam
12A. brakradinaJ11; B, c. brahmadina

l3All Mss. sarvatah., Tib. bcas 14B, c. sa eva

sa1 eva vi ayatita 2 sarva-3 pra ihrdi sthita4 I
ki 5 te vistarato 'ta
tval!l? II
sa k epat tad eva buddha­
buddhatvam nanyair yat prapta 8 kalpasa khyeya­ kotibhir yavat I
asminnapi janmani9 tvam10 prapnoti 11 sat-12 sukhe­

naiva II

0ne expects the feminine gender in agreement with the
antecedent in the preceding line. This line, however, is
cast in the masculine; and there are metrical considerations.
2c.!tah. 3A, sarvva
4A. sthita 5c. kin 6B, tal].
?B, buddha tva; c. butva

8A. nanyairyatprapha a ; B. buddhabhyal!lyadaprapa ­ nanyairyatpraptam; c. Buddhatva dayaprappal!lnanyairyan­ prapta . Tib. ji ltar sans rgyas gah thob pa. The Mss. show corruptions of omission and repetition. We follow the Tibetan as well as the consistent elements of the Mss. to correct for a Pada with regular meter,

9A, B, jatmani
10All Mss, ty . Tib, khyod 11B, prapnobhi


athava1 vajra-2 dharatvam athava3 cakravartitvarn I
astamahasiddhim va4 anya manas!psita vapi II
moho rago5 dve o6 manas tvrr ya ca p casa kle a 7 I
sattvas8 tu pratibaddha9 yena10 jayanti11 svanga- kenaiva12 II
ebhir baddha sattva13 a gatisa saravartino jata I
kurvanty-1 4 anekapapa kle air15 vimohitap santa //

1All Mss. athava
2All Mss. vaJ.a

3All Mss. athavanehnaiva. Tib, van na. The corrup­ tion by addition has simply been copied by All Mss., the u­ nanimity here not giving weight to their reading. The cor­ rected line is metrically regular.

4A. va 5A, rago 6A. dve a
7c. sam.kre ah.
8All Mss. satvas

9A. baddhavai; c. baddh dyam 10c. omits
11A. janati; B. jananti; c. omits. Tib, 'joms
12A, svangena; B. vaisvangakenaiva 13All Mss. satv
14B, kurvvanty-

l5A. klesai; c. klesai

atha caitan vihantu- 1 vidagdha 2 buddhena3 nir­ mita-4 yuktiJ:l I
pratyayabhuta klesah8 praj opayas tu nirmita-9
hetul').10 /
phalabhuto hyasau lak yas tridhatu u nirmalaloka 11 II ye yena vidhina sthapya bhutas te tasya rupi al'). I indhana(vidagdhaya)12 yena ramya-13 jinanat.akam.
divyal]l 1/

1c. viharttu!!l
2All Mss. vidagdha JB. ddhena
4A. nirpita; B, c. nirmmita 5c. dul').-
6A, viyoga

7B, patitana; c. pattitana 8c. kresaJ:l
9A, c. nirrnrnita; B, nirmmita lOA, B, hantu
11A, nirmmalaloka; B, c. nirmmaHU okaf:l

12A. for the Pada, indhanavadninyayena; B, c. indha­ navadhniny yena. Tib, bud sin bsregs pa'i rigs(corrected) pas ni. The intended meaning seems clear enough from Tib, although rigs is a very strong translation of yena which is, therefore, doubtful. The chief difficulty is correcting
for the form which lies behind bsregs pa. Our solution accounts for most of the consonants and vowels in the read­ ings of the Mss., but the correction is tentative.

ayam1 eva 6uddhalak o (p ca-)2 Skandha-3 yatana­ dhatavatl I
ta hanti samartho4 bhutva satru5 samartho6 yatha atrurp? II
Moha 8 Moha-9 vi,uddhya dvesa-1° visuddhya tatha dve am11 I
Raga 12 ragavisuddhya manavisuddhya mahamana //

13All Mss. ramya . Tib. for the Pada, dgyes pa'i rgyal ba gar mkhan mchog. Were we to accept the reading
of the Mss., the sense of Tib. would not be possible. While Tib. may be providing a particular interpretation among
others, it is just as likely that the Skt. has disrupted a compound.
1A. aye

2All Mss. p ca. Tib. omits. The reading of the Mss. produces an irregular line metrically and is a likely addition.

3A, B. ska dha-; c. skadha- 4B. sarnarth
5A. sarnantu; B, c. santu
6A, B. samartha
7A. anturp; B. santu; C. saturp 8B. soha
9B. omits 10A, B. dhe a
11A. dhe9a; dhe arp
12A. Raga

tr yam ir yavi uddhya sarvavi uddhas1 tu vajra- 2 dhrk3 1
tadrupavi uddhya p cakle a sama yanti II
etani pancakul ni pa ca-4 S"Xanani p cabuddha I

ebhyo jata sattva5 vajra-6 garbhary? krodhas tri- dhatuka ca II
ayam evadhyatmabhedo guruvaktral labhyate 8 I
ye guruhina sattva na labhante te mantramudram api II

1c. visudhis 2All Mss. vaja

3A, dhrktrvah; B, C, dhrjflavaey, Tib. 'dzin mgon. The presence'of'mgon in Tib, suggests a line longer than our correction; and the readings of the Mss, point to the
same, This, indeed, may still be the case. Nevertheless, mgon need not reflect natha or a similar word in Skt. but may carry the same nuance as -vat which can be implied in Skt. without actually appearing. The Mss. have had trouble with the unusual -t stem inflection which could lead to a variety of formations at the end of a line, Our solution, finally, is metrically regular.
B, pa ca

5All Mss, satva
All Mss. vaja-

7A. -garbha

A. labhyate; c. labhete. All Mss, add 'ti sya ena. Tib. omits the addition. This phrase is required two lines below where it is omitted; it is, therefore, misplaced,

(yo 'tispas.t.ena)l jambudvipe 2 tato 'smin3 I buddha(dhyu ite)4 trikopa-5 ke uddhe I ekarakrtimadhye ramyasyaivam 6 yatha bhavati II
triko e maQ9ale ramy vajra-7 (rali-)8 vinismrta 9 I

lA11 Mss. om.t the phrase.
T"b• gan.
z'•g s'"'•n. du
gsal bas ni. The phrase 'ti spa teraappears two lines above without support from Tib.r we judge it misplaced and orig­ inally here. The Editor adds yo in reconstruction of Tib. gan zig, the choice of the inflection determined by the Avagraha in the Mss, It must be noted, however, that a
line of three Padas which are metrically irregular has been
produced. These three Padas are translated as two in Tib. The line continues to be doubtful.
2B. ja-unreadable-dvipa 3c. smin
4All Mss. dhyU ite. Tib, bsnagse In a questionable line, this reading is highly questionable. While Tib, points clearly to a form meaning "praise," the closest we come to the Skt. is a form of root -stu or -sa s as found in MW. Without additional witnesses at this point of our analysis,
we simply record the unanimous reading of the Mss. in a
tentative WiY.
5A. trikona
All Mss. madhyeva yasyaivam. Tib. dgyes pa'i e ba .
Tib. points precisely to our correction which also alters
the Skt, reading in a minimal way; and both Skt. and Tib.
are metrically regular. While we thereby produce what seems
to be the authentic reading, it remains highly doubtful from the standpoint of meaning. A more meaningful correction of Skt. would be• madhye va ramyasya, which is not supported
by Tib. content or order and alters the Mss, more thoroughly,

7All Mss. vaja

8A. lili; B. rasi. Tib. ra li. This seems to be a technical term, the meaning of which has not been determined. It is, therefore, doubtful.
9c. smrta

dharmodayetil vikhyata yo ita bhaga ityapi II
tasya madhye 2 gata padmam.3 a apattra4 sakar.Qika 5 I tatrali-6 kalisa mi6ra7 a tau varga8 vyavasthita II kurvanti Karma-9 sa ghata lO mantrarupe a dehina I
pa casadak aranyeva1 1 vedas tu sa sthita12 ime II
tantr nam caiva1.3 mantrana astra a bahyarupina 14 I

etani vajra-15 garbhasvarup i16 nanyad17 ebhyc 'sti ki cit18 II

1All Mss. dharmmo- 2A. madhe; B. madha
.3A, c. pada m
All Mss. padr . Tib. 'dab rna

5A. sakka ika ; B. sakarD it ; c. saka ika 6A. tatrali
7A. sanmi6 ; B, c. sanmi6ra 8All vargga
9All M ss.

lOB. sagh t8.J!l; c. saiJlghata 11B. ak aradhyeva
12c. sasthita 1.3A. caival).
14c. rupinalllaJ:l

l5A1J. Mss. vaja
16A, c. sarvarup i; B. sarvarup i. Tib. rah bzin 17B. nanyad
18B. kin'ci

akaca atapaya avarga l ca tani p ca6ad api I
samani nirmitani2 1
vajra-3 ralau padmagatani pratidalam 4 as.t.a(dig. - vidik u- )5 viditani II

e a madhye tu ki jalke vidyate parame vara /
a abhir varga-6 kai caiva? ve thitaparamak ara II akara8 sarvavar agro9 mahartho varganayaka I tatraiva10 samudbhuta sarva-11 mantras 12 tu dehina 131/
khang jana- 14 padalepa- 15 gulikapatalayaks.anganah. I

1B, C, varga
2A, B. nimitani; c, nimittani JA. vaktra; B, c. vaja
4c. -daram

SA, c. dik u; B. dik a. Tib. phyogs rnams dan mtshams rnams su. While Skt. dik u is possible and Tib, may be a gloss, the reconstruction from Tib. produces nearly regular meter. Besides, the full formula could easily have been corrupted by absorption of similar forms.
6c. vargga

7All Mss. ceva. Tib. omits. The form is present for metrical purposes alone; and iva implied by the uncorrected reading does not serve that purpose. Also, iva is meaning­ less here.

B. akara a r c. akara
9A. sarvarl)l1a-; c. sarvava a-
10All Mss. tatva 11c. sarvva
12c. mantra
13A. dehina 14c. ailgajana

l5A. p dale pa-; c. padarepa

trailokodara-1 varti-2 sutaka-3 gati(karrna)nuga 4 (siddhayal))5 //
ya sarvas tu rnaharddhaya svabhavane6 (satya ca)? kamanvitat) I
ta varga9taka-8 madhyagat svaravarat niryanti9 vargatmakal;l //
ya 10 kascit prasaro vaca janana pratipadyate I

1A. lokkadara; B, c. lokkodara 2A. vartta; B. vartti; c. cartti

3A. sutak; B. sthataka; C. s(rnissing)taka

4A. karmrnanugat); B. karmanuga; c. karmmanga. Tib. rjes su 'bran. The absence of translation in Tib. together with the especially corrupt character of the passage qual­ ifies the original presence of Karma; its absence along
with the absence of siddhayat) allows regular meter.

5All Mss. siddhayah. Tib. omits. The absence of translation in Tib. together with extensive corruption of words and meter here make suspect the original presence
of siddhaya . The word could have been added to explain the "members."
6A. svab


7B. satya-unreadable. Tib. for the Pada, ran gi khyim du lna bcu 'dod pa dah ldan par 'grub bo. Virtual
unanimity on this reading without a viable alternative makes
it difficult to eliminate this phrase. It is highly doubt­ ful, however, according to Tib. -- which itself is doubtful with an irregular line.
8A, B. vargga taka; c. vargga take 9c. niryyahti
10A • yel)

sa sarvo mantrarupo hi tasmad eva praj yate II
nado hi mantram ityukta l sarve a2 tu 'ariri a 3 I

vi vagranthi-4 mahasthanat5 Dharma-6 ralir7 vini - caret I
na kat
6cin niyato mantrah.
siddhyr.ddhyas tu8 vyavasthi-

anutpannasvabhavo9 hi mantro varge varap 10 para I ata para pravakqyami sa putodbhava-1 1 lak a a II ekara prthivi j eya Karma-12 mudra tu locana13 I

1B. ityukta 2c. sarvve aJ!1

3c. Sarira'{la
4A. vi6vargrathi; B. vi cagrathi; c. vi cagra thi 5A. -sthanata ; B. -sthanata
6A, c. dharmma
7A, B. ralla; c. ralle. Tib. ra li
8A. siddhirddhyastu; B. siddhirvy stu; c. siddhirddhya 9c. svabha
1°A, B. vargge 6vara}J.; c. varge s•vara

11B, c. sampu odbhava 12All Mss. karmma.

13c. locanaJ!1

mahakrpa mahopayal svarupa2 visva-3 gocara I
sthita nirrnan.a-4 cakre5 vai nabhau II
va -6 kara tu7 jala j eya8 Dharma-9 mudra tu ma-
maki' I
maitripranidhi- 10 rupa11 tu12 devi vajra-13 kulod­ bhava14 I
sthita sa Dharma-15 cakre tu hrdaye sva adalambuje 1611 makaro 17 vahnir uddi to18 mahamudra tu p[Qqara19 I

1All Mss. mahopaya

2A. visvarupa; B, C. viscarupa. Tib. ran b in. e corruption apparently has derived from confusion with v1sva later in the same line.

4A. nirmmaf:ta; c. nirmmana 5c. cakra
6B. va-

7A. karaip.stu; B. karastu

J•n"'eyam.ah. ; B,

9A, c. dharrnma

lOB. pral)it;liJ c. pranidhi

12A. tutl
13All Mss. vaja 14c. kulobhava
15A. dharrnmacakretu; c. s.t.adares.u 16A. dalal]lbuje; C, darambuje
17A. makare

muditabalayogenal devi padmakulodbhava I
sthita sa bhoga-2 cakre tu3 k he4 dvya a-5 da- lambuje6 II
yakaro vayurupas tu sarva-7 kle aprabh jaka 8 I
mahasamayamudra vai9 dev•i Karma-10 kulodbhava II
upek aj anayogena1 1 Tara sa saratari I12 I
sthita mahasukhe sira -13 cakre dvatri saddalapan­ kaje14 II


19A, B. pal'}gala; c. padara lA. yogenal)
2B, c. sambhoga 3A, tu.Q
4B, c. kan.d.e
5A. tvya a; B. vya "!:a 6c. daHibuje
7c. sarvva

8A, c. prabhal)'ljakal} 9B, c. cai
lOA, c. karmma

11A. yogena ; c. yogel)a 12A. tari1fit).

14c. PaJ.Ilkaje

ekarerya1 smrta Prajna va kara 2 capyupayakap I
va karabhusita -3 casav ek ra4 sobh te5 dhruva II adhorddhva- 6 samapanno7 prajnopayasvabhavata 8 I evam iti nipatena9 avocat10 paryanta 11 tantra-
samudayao12 sucyate13 va (vak yati)14 //
maya rutam eva hi mahar go 15 •tra sada sthita I sru

1c. ekarena
2A. rvaka-ras; c. va ka-ras

3All Mss. bhusita . The sense of the passage renders the plural impossibie here.

4A. amkarah; B, c. akara. Tib. e. The sense of the passage sup orts'Tib.; it would not be difficult for a single vowel to be lost or transformed in transcription.

5All Mss. sobhate
6All Mss. adhorddha
7A. samapanna; B, C, samapannya 8c. svabhavata
A. nipatenao
10B. avocata

11A. paryyanta ary; B. aparyanta 12A. samudaya
l3A. sucyato; B. sucyatai

14All Mss. vak yati. Tib. omits This readin at the end of a line creatin irre lar meter is highly doubtful; it has the nature of a gloss perpetuated by the scribes.

l5All Mss. rajo

ravarya 1 nirdi ta tam2 ya ca mahasukho3 natha4 /

maya5 6rutam iti rotena j anena ruta na6 tvadhigata 7// ajra arbha uvaca)8 //
nadhigate9tu bha avan ko dos.ah.10 //
bhagavan aha//
sa itikarakade akayor abheda syat I athavadhiga­ tam11 eva vaineya-1 2 janava at13 de aka eva sa gitikara­ ka 14 syat15 //

1 / /
A, C. srava a; B. sava a

2All Mss. omit. Tib. ta corrected to tam. It is clear from the context that the syllable tam must be foand in the line, and Tib. supports this. It happarently has been absorbed by the preceding word •&irdi tam whose final letters are similar to the omission, likely corruption in
oral transcription.
- suso
4A. natha

5A, B. mayeva; c. mayava . Tib. thos pa. This cor­ rection is facilitated by the fact that this word, accord­ in to the context, must appear in the opening line of the Text.
6A. ca
7All Mss. -gata

8The connecting phrase has been added by the Fditor to facilitate translation.

9All Mss. adhigate. Tib. rtogs pa ni rna yin no. Several factors argue for the negative heres the clear in­ dication from Tib, the relatively easy loss of an initial letter of a line; the absurdity of Vajragarbha's question here if the negative is omitted.
A. de a
11B, c. athabodhigatam 12A, c. vainaya

de ako 'ham aha dharma1 rotaha 2 svag air3 yutal;l II
(vajra arbha uvaca)4 II
tat katha j ayate5 II
(bhagavan ha)6 II

n tyati yatra mahasukha-7 natho nrtyaty ekam anekarasa­
nam8 iti vacanat I yat kim.cid 9 akhyatam.

bhagavata kula-
putras1° tat11 maya ekasminn eva12 kale ekasminn eva13 k!?ane14 /1

13A. vas t 14c. karakal) 15A. syatal;l
1A. hammarnrldharmma; B. hammahaddharma; c. ammaha­ ddharmma. Tlb. for the passage 'chad pa bo na chos kyan na.
B. otaha1J1
3B. svagudhairt C. svagunair

4The connecting phrase has been added to facilitate translation.
5A. jrfayatel)
6The connecting phrase has been added to facilitate translation.
7All Mss. -suha

8All Mss. -rasaneti. We correct for the inflection in agreement with ekam in the line.

9A, ki cid; B. ka cid

10A, B. putral); C, putra-unreadable. Here c. adds a repetition of the previous thirty-four syllables.
11c. tanyaya

6rutam ity-1 adhigata 2 1 anenacintya-3 (vimok a-)4 samadhilabhatva 5 tac citt 6 II
samaye7 kala ityukta kalo8 hi trividho mata 9 I sukalas caiva du kalo10 acintya-11 kala12 eva ca II k Iradhar vat13 nasapu arandhre a- 14 ata sukala15
ucyate 16 I

12c ekasrnineva

13A. omits kale ekasminneva t4c. k adhe
lA. ruta iti 2B. edhigatal!l JB. -acittya
4All Mss. vimok a. Tib. omits. The reading is doubt-

5A, B. labhitva s; C. rabhitva s 6A. tacittam.; B. taccitarn.
7A. sarnaya ; B. Samaya ; c. sarnaya. The correct form
should repeat the form in the opening line of the Text.
8c. karo 9A. gata
lOA. kalo ; c. karo

llB. avintya};l 12c. kara
13A. -va; B, c. vata 14B. randhedha-
15B, c. sukara 16c. ucyato

gatas caiva bahnirupe a du kalas1 tayor ekasya k!rtitah.2 /
asahayo bhaved eka kalas tvacintyata J gatat II
na r go na viragas ca madhyama nopalabhyate 4 I

tatra Raga a aktilak a o5 virago nirodho mata // abhya6 rahita madhyama na pratiyate7 etat trayan8
nopalabhyate9 I
Raga caiva viragas ca mi6ribhutam anavila 10 II
tatha ragaviragabhyam 11 eka samarasa k ana I
sarvabhavanam.12 (Bhavana kalaika mata)1J II

1A. dukalas
2A. kirttita; B, c. kirttita
JA. cintyata ; B. vintyatam: c. cintyata 4A. nopalabhyete; c. noparabhyate.

5A, B. lak a ; c. rak ana
6c. abhya
7A. pratieka ; c. pratiiyate 8A. etatuyan; B. etatatrayan
9A. -oparabhete; c. -opalasyate 10A. anavilama ; C. anavira
11A. -abhya a ; B, c. -abhya 12A. Bhavana

l3All Mss. omit this phrase. Tib. bsgom pa'i dus ni gcig tu gsuns. We judge it highly likely that Tib. is au­ thentica not only because its meaning fits the context, but also because Tib. omits the next passage found in Skt. Skt. seems to have confused similar initial elements of each pas-

bhagavan sa ucyate I tad yatha I ai varyasya1 sam­ agrasya rupasya ya asa 6riya j anasyartha- 2 prayatnas­ ya I saoDa bhaga3 iti smrtio4 I so 'syastiti5 bhagavan I athava yo6 vipak ikan7 dharman8 tan9 bhagnavan itilO bha­ gavan II
sarva-11 tathagatakayavakcittanyeva1 2 hrdaya 13 tad

sage, bh van and bhagavan; Tib. has confused parallel elements, bsgom and bcom. As a result of this kind of con­ fusion, Skt. and Tib. each loses a passage, but each loses a different one, The Editor has reconstructed the lost passage in a highly tentative way,
1A , C , a1svaryyasya; 2c. j anasyall}s

B • esvaryasya

JA. sarryryabhaga; B, C, sa abhaga

4A, smrte, A, adds here a lengthy passa e of fifteen lines which appear to reflect non-tantric Mah yana inter­ pretation. There is no support for the interpolation from the other Mss. or Tib,

5A. • sya •st'iti
All Mss, ye
7A, B. vaipak dka; c. s.aipaks.ika sA, c. dharrnmas; B, dharrnas

9c, tana lOB, imi
11A, c. sarvva-

12B, cintya 13B, h:rdaya

eva vajra 1 saiva yo9it2 tasya3 bhagas tatra vijahara II

(vajragarbha uvaca)4 //
he bhagavan vijahareti kim ukta 5 bhavati II
bhagavan aha II
samarasasarvatathagata- 6 vicitropayena vineyata-7 manasi8hrtavan9ityarthaIsarvatathagatakayavakcitta­ h dayavajra-10 svabhavay11 dharmodayaya 12 yada bhaga­ van vijahara13 tada maya rutam iti I tatha14 I kle6o

1All Mss. vaja
2A. saivayo i; c. sa vajo it JB, c. tasma
4The connecting phrase has been added by the Editor to facilitate translation.

5B. aktal]l

6A. tathagatal:t
7B. vinayata; c. cireyata
sc. manasi 9A. dvatavan
10All Mss. vaja
11A. svabhavaya av 12A, c. dharmmodayaya
13B. bhagava-unreadable-ijahara 14c. gatha

vinihanyate 1 praj ayadhikasca te2 kle as3 I tasmar14 praj a bhaga ucyate I tasmin5 bhage sakalatathagato6 yo itsahita 7 sthita 8 /1
(vajragarbha uvaca)9 II
aho hi sarva-10 buddhana 11 sarva-12 j ajnanam anuttam8J!l13 II
(bhagavan aha)14 II
yat tathagatasaukhyartha 15 bharyatvam 16 upas­ k:r;tam17 II

1A. vinihanyate J B, c. pinihanyate 2
N A. praj adhin {cata: B. praj adhTnascate; c. pra-
jnadhinasca te

3B, c. klesr h 4A, B. tasmat

5A. tasmi
All Mss. Tathagata
?B. sahita; c. sahitata
8A. sthita
9The connecting phrase has been added by the Editor to facilitate translation.

11A. buddhana a ; c. buddhana 12A, c. sarvva
13All Mss. uttam . Tib. bla na med

14The connec t• ng phrase to facilitate translation.
15A. -artha 16
A. bharyyatvam

has been added by the Editor

iti kulaputra1 evam maya rutam.
varn.ah.2 Dharma-3
dau nitya 4 sa sthita I uddha sa buddha-5 siddhijana-
karya6 II
akathya7 vajra- 8 p ena9 maya kathita I varryair 10 ebhi 11 sattv 12 paran13 tu sa saram atidura 14 / lak yam krtva sak ad 15 sak t ta116 ltnacittena17 prapnotyasmin 18 janmani19 Buddhatva vajra- 20 sattvatva 21 va I yadeva padam 22 acintya na prapta yac23 ca sugata 24 ca buddha

17All Mss. apisv!k ta . Tib. e bar mdzad pa. Our correction remains questionable.
1B. putro
2A. va a ; B, c. va a 3A, c. dharmma-
4All Mss. nitya
5A. 6uddhah.
6A. k-arye;
sam.budha; B. sabuddha
B • kaye

7A, B. kathya; c. 'kathya. Tib. rna b6ad. Tib. re­ ceives support from the Avagraha in c.
8All Mss. vaja
9A, c. p e; B. pase 10All Mss. va er 11A. ebhi
12All Mss. sattvary

l3All Mss, prayan. Tib, pha rol du, 14
A. -dulall}

l5All Mss, sak ada 16A. t

bhavanti I samyag-1 uktasu lak ite lak ye2 sevita-3 matre a4 avicau5 pray ti I sa kli a-6 cittena7 satya-8 k ayitaQ9 I sa kli a bhavabhayabh!tas tu sadhakaij II

17A. la acittena; B. linacittana; C. linacittena 18B. prapnotyasmin; c. prapnotyasmi
19B, c. jatmani
20All Mss. vaja
21A, B. satvatvan; c. satvatvam 22B. avadam
23B, C. cac
24All Mss. sugatai{ca 1A. samyat-
2A, c. lak ya 3A. secita
4B. matrer,1a 5c. avicau 6B. sakli ta
?A. cittenal)
A. s tya. Tib. ro. The terms are doubtful although
they could have a related meaning.
9All Mss. ks.y tah.. Tib. span

6riman1 bhavayati suddhacitta praj opayaprayoge a 2 tenaJ yatha sa prapta4 6uddha sa vedanatmaka 5 lak yam pra­
j opayavikara 6 gaganasam 7 tribhuvanaloka II
tal lak ya dullak ya 8 I sarvagata hetupratyaya-9 vadhir a cintama ir iva loke varthakara I sarvasiddhya­ grarp10 II
iti sarvatantranidanarahasyat 11 rtsa putodbhava-1 2


samaptah.13 II

lc. 6rivam

2A. prayogenal) JB. tenai
4A, B. saT(lprapta

5A. -atyaka 6A. vikaral]lal']. 7c. saya
8A. omits dullaks.yam.; B. dullankf?YaTIJ
9A. pratyeya- 10A, B. atra 11A. rahasyatatj
12B. Sr!sa11JpU Odbhaval'].; C. risamputobhaval'].

lJB. samapt al)


(Chapter 1)

rdo rje rnkha' 'gro la phyag 'teal 1o111
'di skad bdag gis thos pa'i dus gcig na I bcorn ldan 'das de bzin g e s pa thams cad kyi sku dan gsun dan thugs kyi snin po 2rdo rje btsun mo'i bha ga3rnarns la so II
de nas mad bcom ldan 'das kyis rnal 'byor dban phug bye phra brgyad bc 'i nan nas rdo rje s in po la
gzigs te 'dzum pa mdzad do II

'dzum pa mdzad rna thag tu de'i tshe rdo rje sfiin po stan las lans te I bla gos prag pa cig tu byas nas pus rno g'yas pa'i lha na sa la btsugs te II thal mo sbyar ba byas nas bcorn ldan 'das la 'di skad ces gsol
to II

ye 6es kyi dban po rgyud thams cad kyi glen g i'i gsan ba yah dag par sbyor ba las byun ba'i mtshan id
nan par 'tshal lo II

1All Skt. Mss. add om; see Note to Skt. edition. Also note that Cornrn. reads.dpal rdo rje rnkha' 'gro rna la phyag 'tshal lo. The line appears in introductory remarks and may not be a glo s on the Text.

2 .
All Mss. omit h:rdaya. See Note to Skt. edition.

JSkt. bhaga is always transliterated by Tib.

(bcom ldan 'das kyis bka' stsal to 1/)1
e ma'o rdo rje s in po legs so legs so II snin rje chen po legs so legs so II byan chub sems dp chen po yon tan gyi ('byun gnas)2 legs so legs so II gan ig

rgyud thams cad du gsan ba de thams cad 'dod pas fus
§ig II

de nas rdo rje s in po la sogs pa'i byan chub sems dpa' chen po de rnams rab tu 'dzum in mig phye bas phyag
•tshal te II ran gi the tshom yan dan yan du lus pa II
rgyud ni thams cad ci zes brjod II
glen gzi3 yan ni ji ltar lags II
'di la
gsan ba ci zes gsuns II
ji ltar yan dag sbyor byun5 min II

de la mtshan id ji ltar lags // bcom ldan 'das kyis bka' stsal ba/1
rgyud dan thams cad de ni rgyud thams cad do II
rgyud thams cad 6 kyi sgras na 'dus pa la sogs pa'o II
de rnams kyi glen gz
ni byun bar nes pa
zes bya ba'i

1Tib. omits the introduction. Comm. bka' stsal lo. All Mss. bhagavan aha. The Bhagavat's next speech immedi­ ately following is introduced by the same Skt. and the verb is translated exactly as the Comrn here.

l:ara. 'byvn cy of
Tib. phun po. Comm. 'byun gnas. All Mss. gu a-
The appearance of phun is an obvious corruption of
while the loss of gnas is perhaps nne to the frequen­ phun po for Skandha in Buddhist texts.

3Tib. glan gzi
4c omrn. gan• la

5Appeal-: i_n Vajragarbha •s original request as yan dag par sbyor ba las byuh ba.

don to II khyab ;jug dan I dban phyug chen po dan I
tshans pa dan I fian thos dan I ran sans rgyas rnarns kyi
spyod yul rna yin pa'i phyir gsan ba'o //
yan dag par

sbyor bani thabs dan es rab kyi bdag id can no //

de las byun bani yan dag par sbyor ba'i tin ne 'dzin ces bya ba'i don to// 'byun bani skye ba ste //de ltar gnas pa dan g'yo ba'i dhos po'i ran b in gyis bdag nid
can du gyur ba'o // rntshan id ces bya bani 'dis1 rntshon

par byed pa'o //

yan na rgyud thams cad kyi glen gzi yan dag par
sbyor ba zes bya ba'i 2 sgras ni rdo rje sems dpar gsuns so// gsan ba zes bya ba 'dis ni man nag can gyi snags btu ba dan// dkyil 'khor gyi lha la sogs pa (dban bskur ba)'i3
rntshan id do //

thabs dan ses rab bdag nid rgyud // de ni de yis bsad kyis non //
dan por ston id bsarns pas ni // lus can rnams kyi dri rna bkru //
6Tib. thams tsad 1
Tib. 'das

2Tib, yan dag par sbyor ba las byun ba'i bya ba'i sgras. Corrected Skt. sa puta6abdena. Comm. kun du kha sbyor zes bya ba'i sgras, Since Tib, elsewhere uses zes bya ba'i while Comm. is more likely to use bya•ba'i, the
es has probably been corrupted to las with the subsequent formation of the frequent compound sa putodbhava.

3corrected Skt. devatldyabhi eka. Comm, in intro­ ductory analysis, dban bskur id kyis and in commentary proper, dban bskur ba'i. Since devatady- is treated sep­ parately by Comm. as lha la sogs pa, dban bskur ba is not being included within - dy,

gzugs kyi khams ni ston pa yi // sgra la yan ni de yis bya //
dri yi khams ni ston pa yi // ro la de nid kyis ni bya //
reg bya'i khams ni ston pa yi //

yid la de id kyis ni bya // rdo rje snin pos gsol pa //
mig gi ran bzin ston pa yi // nan du rnam es ji ltar lags // rna ba'i ran bzin ston pa yi // nan du rnam ses ji ltar lags // sna yi ran bzin ston pa yi // nan du rnam ses ji ltar lags // lee yi ran b in ston pa yi // nan du rnam 6es ji ltar lags//
lus kyi ran b in ston pa yi //
nan du rnam ses ji ltar lags// yid kyi ran bzin ston pa yi // dbus su rnam ses ji ltar lags//
de phyir gzugs med lta bo1 med //

sgra med thos pa po yan med // dri med snom pa po yan med // ro med myon ba bo 'ah med //

1Tib. lta bu. The passage appears in HVT. (1.5.l)a
no bos gzugs med mthon ba po// sgra med thos pa po yah med//
dri med snom pa po yan med// ro med myon ba po yah med // reg med reg pa po yan med // sems med sems las byun ba'an dfi

reg med reg pa po yan med II
sems med (sems las byun ba)1 'an med II
bcom ldan 'das kyis bka' stsal pa II
is med is ni rnam spans pa'i II
de nid lam ni ji bfin on II rtog ge med cin rnam ses med II de bzin rtog ge pas2mi rtogs II sans rgyas kun gyi gsan bani II nam mkha' lta bur am pa nid II nan thos rnams kyis mi ses te II mi 6es pa'i mun pas bsgribs /1 bag chags san ba'i ye ses gah II
ran rgyal rnams kyi sprul ba'o II
gsan ba sans rgyas spyod yul id II
de rnams kun gyis mi ses so II

'khor ba'i rgya mtsho'i 'jigs pa las// rgal bar 'gro ba'i sems kyis ni II
lus kyi dbus su bskor ba yi rgal ba mtshon bar gsuns pa ste rnam ses ye ses sems kyis ni
mtshan rtid mtshan g i mtshon bar byed /1

1The line is problematic. It is supposed to trans­ late the corrected Skt. na Citta napi cintakam; but sems las byun ba reflects caittika.

2Tib. rtog ge bsam. Corrected Skt. tarkikais. Comm. rtog ges. The instrumental plural of the Skt. and the necessity of a substitute for bsam for reasons of meter point to an agent noun.

ye s,;'es kyis n1.

s,;'es bya blta


es bya rnams kyi 'gro ba brtag // 'gro ba brtag pa'i sems kyis ni // ji ltar ran 'dod la sogs 'gro // 'gro bar snon ni skye bani //
ston phrag du rna dag du 'gro //1
(yo gis)2

gsuns mtshan nid kyi //

mtshan nid ston bar bsarns nas ni // lus kyi ston ba'i fin sogs la //
blo dan ldan pas sa bon gdab //
rtsa yi mdzes spyi las byun ba // de ni rdul bral rab tu grags // dag pa'i sel gyi 'od3 dan rnnam // byan chub sems ni 'od gsal ba // ye es lrla yi dnos de id //

yuns dkar gyis ni sbom phra tsarn // de yi nan du gnas pa'i lha //
mi gsal gsal ba'i ran bzin can // rna rno phyed pa rnchog tu phra //
thig le'i gzugs kyi yid kyi dnos // snin gi dbus su rtag gnas sin //
1The line is doubtful. See Note to Skt. edition. 2skt. omits. The preceding Skt. line reads yogi
not rendered into Tibetan; apparently the word is trans­
literated here. Since both the Skt. and Tib. lines show regular meter and can be translated with the same gener­ al sense, the stronger reading is not apparent givenour materials. We simply note the line is doubtful.
3skt. omits.

'od zer ldan fin1 gzi brjid che II bcu is mtha' dan dgu yi mthar II rkari mthil mgo bo med par ni II thag ba gcig ni rnam phyun nas II

lte ba'i dbus su rnam par gnas // cha ni lna pa tsam id ni II klu2dban dbyibs su rgyas pa la II
de yi cha yan phye nas ni II
rdo rje'i dbus su rnam phyun ste II skyes gnas dbus su sa bon gnas II chos dbyins khu bar byas nas ni II sgo ni dgu po thams cad du II
de ni rim pas 'pho bar 'gyur II
de bfin me ni tshans pa'i (dnos 3 //
sa yi sa bon is kyis ni II mig la de id sbyar bar bya II rlun gi ston sogs sa bon gyi II
sems kyis sna dan rna ba sogs II
bdud rtsi chu yi sa bon lee /I
dban por gtso bo dban phyug rtid II
mgrin par gti mug sa bon te II
lag par on mons te bfin no II

1skt. shows no equivalent; its presence may be al­ lowed in a general sense of "sphere."
2Tib. klu:h

3we expect tshails pavi bu ga for Skt. brahmarandhra. Tib. reads dnos which points to rupa, maya, and also mrrla; but -randhra is tentative. See Note to Skt. edition.

g'yo ba s in gi gnas rnams su II

dag lte
ba'i rtsa bar ni II
yan lag brgyad kyi sa bon gyis /1 yan lag brgyad kyi cho gas 'dzin II brtan sogs g'yo bar bcas pa yi //
'gro kun khyab par byed pas khyab II
tshans sogs lha dan lha min rnams II
de yi gdan du gyur pa yin II
gan la gtso bo spyod gyur pa'i II
khyim ni bha ga es su gsuns II ji srid lus la rnam gnas ba /1 de yi las dan las min 'gyur II dge dan mi dge dan cun zad II
byas ba'i las ni thams cad byed II
rnal 'byor mflam pa id du gsuns II rab du sbyor ba2 bsgom pa yin II gan tshe las kyi lus bcom ba II
de ltar lha ru 'gyur ba yin II
gan gis brtan dan g'yo khyab pa II
des ni de tshe ran gis nus II nam mkha' lta bur mnam nid ni II de yi kha dog es bar bya II

1Tib phyugs bdag. Corrected Skt. prat)ibhy lp. Comm. phyugs. According to Das, dag often points to Skt. dual inflection found here in All Mss, The corruption of dag to bdag may have occurred by way of -ba dag.

2Tib. anticipates prayoga, yet Skt. reads puja, Perhaps the Tibetan is a translator's gloss on the meaning of puja here.

mya nan 'das nas dpa' po ni II gos pa med cin dri rna spans II rna dan sri11 mo nid dan yan II bu mo de b in en mo dah II
bram ze mo dan rgal rigs dan II
rje'u mo de bzin dmans mo dan II
gar rna tshos rna g'yun mo dan II
de bzin du ni gdol pa mo II
thabs1 dan ses rab cho ga yis II
de nid rigs pas2 mchod par bya II
ji ltar bye bar mi 'gyur bar I/
rab tu 'bad bas bsten par bya /I

rna gsan sbrul dan chom rkun sogs // sa la spyod pa3 sdug bsnal byed II rigs kyi dbye bas phye nas ni II
phyag rgya4 rnam pa

1n" ar n.

gsun.s //
bram ze skyes is rigs skyes rna II

de bzin gsegs pa id du gsuns // rgyal rigs rna ni rgyal5 po'i rus II

rma bya zla sogs rigs las skyes //

Most of the remaining lines in the chapter are
found in HVT (1.5.2-J, 7-21). We show significant agreements and disagreements with Snellgrove's edition
for purposes of comparison. See Notes to Skt. edition.
Tib. rig pas. All Mss. vatsala . Comm. rigs pas.
HVT vatsala ; rig pas. Tib. points to some form of vid­ supported by HVT Tibetan; Comrn, however, points to cor­
ruption of Tib. supported by All Mss. and HVT Sanskrit. Snellgrove shows no awareness of the problem.

rdo rje bdud rtsi es byar grags II
rje'u mo ba glan skon ba ste II
las kyi rigs las skyes par gsuns II
dmans mo yan ni bran mo ste II
rnam par snail mdzad chen por gsuns II
gar rna pad ma'i rigs nid de II
thos ma de bfin las kyi rigs II g'yun mo rdo rje'i rigs su bsad II rin chen gdol pa mor es bya II phyag rgya lnar ni rnam par nes II
'di rnams1 de b in gsegs pa'i rigs II mdor bsdus pas ni brjod par bya II dpal ldan de bzin id gsegs sin II
de bfin slar yah gsegs ba nid II
'd1' s n1•
ses rab
r1• gs pas2n1• //
de bzin
gsegs par mnon
par brjod II
rigs ni rnam pa lna ru gsuns II rigs ni brgya dah mtha' yas dan II sku gsun thugs kyi dbye ba yis II phyi nas rnam pa gsum du 'gyur II

3HVT sa spyod me yis1 •••'agni
4Tib. phyag brgya 5Tib. unreadable
1The plural translates the Skt. singular etat which does carry a'-plural sense here. See Translation.
2T'1b • r1.g. Skt. yuktya. Comm. rigs

rigs da.:il 'byun ba lria rnams dan II
phun bo lna yi ran bzin gyis II
rdo rje rin chen pad nor bu II
'dzin pa'i rim1 pa'i rigs rnams so II
sgom pa po med bsgom bya med II snags med lha yan yod rna yin II spros pa rned pa'i ran bzin id II
snags dan lhar ni yan dag gnas II rnam snail mi bskyod don yod dan II rin chen dpag med serns dpa' 'o II tshans pa khyab 'jug i ba dah II thams cad sans rgyas de id brjod II

tshans pa sgrib bral sails rgyas id // 'jug phyir khyab 'jug brjod par bya II i ba rtag du dge ba ste II
tharns cad kun gyi bdag nid gnas // dam pa'i bde bas de id de //
dga' ba rtogs phyir rnam sans rgyas II
gail phyir lus las byun ba id II lha zes rnnon par brjod par bya II 'di la bha ga rona' ba'i phyir II
bcom ldan 'das •fes brjod par bya II
bha ga rnam pa drug du brjod II
dbah phyug la sogs yon tan kun I/

Ms. A. prasuti accepted tentatively; B. prasphttti;
c. prasbhuti. Comm. suggests lam. The passage is highly

yan na on mons la sogs pa II
'joms pa'i phyir na bcom ldan 'das II gan phyir skye 'gro skyed pa'i phyir II ses rab rna ses brjod par bya II
gan phyir skal pa ston pa'i phyir II
ses rab de nid srin mo nid II
ses rab tshos mar brjod bani II

sems can thams cad 'tshod pa'i phyir // (btso blagl rna es de b in brjod) II
yon tan 'jo bar gsuns pa'i phyir II
ses rab bu mor brjod par bya II
s in rje chen pos g'yo pa'i phyir II
ses rab gar mar brjod par bya II
gan phyir bcom ldan mi reg pa II
de phyir g'yun mor brjod par bya II
a li ka li rab smras ba II
bzlas pa smra bar yan dag bsad II
rkan pa'i rjes ni dkyil 'khor yin II ned phyir dkyil 'khor brjod par bya II lag pa s yur ba phya rgya yin II
sor mo ed pa'an de b in no II
gan phyir rnam sems bsam gtan ni //
gail sems de ni bsam gtan no /I

1Tib. bslag. The inclusion of this quarter-verse is strongly qualified since it lacks any Skt. witness, HVT Tibetan reads -- incorrectly -- gtso blag rna zes de b in brjod, Since we may have expected tshos instead of 'tshod in the passage preceding, the addition may represent a translator's gloss, See Skt. edition for fuller discussion.

pha las bde ba gan thob ba // de yi bde ba ran gis bza' //
bde ba gan gis 'chi1 ba 'dir //

de yi bde ba bsam gtan brjod //

brjod par bya ba dan rjod par byed ba'i byan chub kyi sems bskyed pa la sogs pa'i bsgom pa'i de kho na'i rab tu byed ba ste dan po 'o//


(Chapter 2)

(bcom ldan 'das kyis bka' stsal ba

1 II
de nas byan chub kyi phyogs kyi chos sum bcu rtsa2 bdun brjod par bya bo II nan gi lus dan II phi'i lus
dan II phi nan l?{is ka'i lus la lus kyi rjes su blta
fin gnas pa'i nhyir yan dag par rab tu ses te II rna Rig pa la sogs pas yid3 is su yur pa'i 'dul ba'i 'jig rten la dran na dah ldan paI0 II nan {2:i s tshor ba dah II
phyi'i tshor ba dan// phyi nan is tshor ba la tshor ba'i rjes su blta fin gnas pa'i phyir yan dae par rab tu
ses te II rna rie pa la so s pas yid is su eyur pa'i
'dul ba'i 'jig rten la dran pa dah ldan pa'o II nan {2:i chos rnams dan I phyi 'i chos rnams dan II phyi nan is ka'i chos rnams la chos kyi rjes su blta in gnas pa'i phyir yan dag par rab tu ses te II rna Rig pa la sogs pas yid is su gyur pa'i'dul ba'i 'jig rten la dran pa dah ldan pa'o II nan gi sems4 dan II phyi'i sems dan phyi nan
is ka'i sems la sems kyis rjes su blta in gnas pa'i

1Tib. omits. Skt. omits. We include this phrase to aid the translation and qualify its presence here and elsewhere. For a literary, if not a philological, ra­ tionale for our procedure, see Note to Skt. edition.

2TJ.. h""0


3Tib, yad 4Tib. sams

phyir yan dag par rab tu s"'es te rna Rig pa la sogs pas

yid is su gyur pa'i 'dul pa'i 'jig rten la dran pa dan ldan pa'o II 'di rnarns ni dran pa e bar bfag pa1 bzi'o II
sdig pa rni d e ba'i chos rna skyes pa rnams rni bskyed pa'i phyir 'dun pa skyed do II 'bad do II brtson 'grus rtsorn rno II sems rab tu 'jog go II smon lam yan dag par
'debs so II sdig pa rni dge ba'i chos skyes pa rnarns span
pa'i phyir 'dun pa skyed do II 'bad do II brtson 'grus rtsorn rno /I serns rab tu •jog go II srnon lam yan dag par 'debs so II dge ba'i chos rna skyes pa rnams ba skyed pa'i
phyir 'dun pa skyed do II 'bad do II brtson 'grus rtsom mo II
sems rab tu 'j01 go II smon lam yan dag par 'debs so II

de ltar dge ba'i chos skyes ba rnams as bar. bya ba'i phyir thabs tshad rned pa bsgorns pas yons su gan ba'i phyir
yons su gan bar gyur pas bsgorn pa darn pa 'phel bar bya ba'i phyir 'dun pa skyed do II 'bad do II brtson • rus rtsom
mo II serns rab tu 'jog go II srnon lam yah dag par 'debs
so II 'di rnams ni yan dag par spon bzi'o II

'dun pa'i tin ne 'dzin spon ba'i 'du byed dan ldan pa'i rdzu 'phrul gyi rkan pa bsgom par bya ste II bye brag phyed pas nes par 'byun ba dan II 'dod chags dan
bral bas nes par 'byun ba dan 'gog pas nes par 'byun ba
dan II rnam par byan bar gyur bas2

bdag gi 'dun pa nub

Tib. {!!,zag
2Tib. rnarn par byan bar gyur pa translates vyavasarga­ pari ata. The Skt. is unusual, a fact which has resulted
in a rather awkward Tibetan rendering. The awkwardness is witnessed by corruptiom in lines that follow and also in an

par 'gyur zin 6in tu rab tu 'jog pa rna yin pa'o II brtson 'grus kyi tin ne 'dzin spon ba'i u byed dan ldan pa'i rdzu 'phrul gyi rkan pa bsgorn par bya ste II bye brag phyed pas nes par 'byun ba dan // 'dcu chags dari bral bas nes par 'byuri ba dan I 'gog pas 1'lo:S par
yun ba dan II rnarn par byan parl gyur pas bdag gi brtson
'grus nub par 'gyur zin in tu 'jog ba rna yin ba'o II dpyod pa'i tin ne 'dzin spon ba'i 'du byed dan ldan pa'i rdzu 'phrul gyi rkan pa bsgom par bya ste II bye brag
phyed pas nes par 'byun ba dan II 'dod chags dan bral bas
lies par •byun 'Pa dan II
'gog pas nes par 'byun ba dan II rnarn par byan 2par gyur pas bdag gi dpyod pa nub par 'gyur zi:n S'in tu rab tu 'jog pa rna yin pa'o II sems kyi tin ne 'dzin spon ba'i u byed dan ldan pa'i rdzu 'phrul gyi rkan ba bsgorn par bya
ste II bye brag phyed bas nes par 'byun ba dan II 'dod chags dan bral bas ries par 'byun ba dan II 'gog pas lies par 'byun ba dan II rnarn par byan bas ries par 'byuh bas bdag gi sems nub par 'gyur zin in tu rab tu 'jog pa rna
yin pa'o II 'di rnarns ni rdzu 'phrul gyi rkari pa bzi stell
'dod par spyod pa dan bcas pa'o II 'jig rten pa'i yan dag pa •i 1ta ba la dad par byed de II ran gi las kyi rnarn par smin pa la so sor yid ches par gyur pa gan {ig

attempt at explanation by the translator below.

1Tib. byed 2Tib. byed

de ltar las byed 1 bar 'gyur badge ba 'am sdig pa de dan de'i las kyi rnam par srnin pa so sor arns su myon par 'gyur pa'i phyir de ni 'tsho ba'i rgyur yan sdig pa'i
las mnon par 'du mi byed pa'o II 'di ni dad pa'i dban
por e:su-ns so II

gan fig chos rnarns la dad pa'i dban pos dad par byed pa'i chos rnams la brtson 'grus kyi dban pos 6es
par byed do II 'di ni brtson 'grus kyi dban por gsuns so II

gan {ig chos rnams la brtson 'grus kyi dban pos

6es pa de'i chos rnams ni dran pa'i dbari pos rab tu 'jig parmi byed pa'o // 'di ni dran pa'i dban por gsuns so II
gan zig chos rnams la dran pas rab tu 'jig parmi byed pa de'i chos rnams ni tin ne 'dzin gyi dban pos
rtse gcig tu byed pa'o II 'di ni tin ne 'dzin gyi dban por
gsuns so II

gan {ig chos rnams la tin rle 'dzin gyi dban pos rtse gcig tu byed pa de'i chos rnarns ni tes rab kyi dban pos so sor Rig pa ste //de ni chos de dan de rnams la
so sor rtor pa skye bar 'gyur ro II di ni ses rab kyi
dban por gsuns so II

dban po lna po 'di rnams 'phel ba ni stobs lna ru 'gyur te 1/
dad pa'i stobs dan II brtson 'grus kyi stobs dan II

dran pa'i stobs dan tin ne 'dzin gyi stobs dan// ses rab kyi



stobs te // 'di rnams ni stobs lna•o //
de la byan chub kyi yan lag bdun gail Z'e na II

'di lta ste // dran pa yarl dag byan chub kyi yan lag dan// chos rab tu rnam par 'byed pa yan dag byan chub kyi yan
lag dan // brtson 'grus yan dag byan chub kyi yan lag dan//

d a• pa yan dag byan chub kyi yan lag dan// in tu spyans ua yan dag byan chub kyi yan lag dan // tin ne 'dzin yan

dag byan chub kyi yan lag dan // btan s oms yan dag byan
chub kyi yan lag ste //bye brag phyed pas nes par 'byun pa dan // 'dod chags dan bral bas nes par 'byun ba dan // 'gog pas rles par 'byun ba dan // kun nas on mons pa dan

bral ba'i kun nas rnam par byan bar gyur pa dan // chos rab tu 'byed pa ni byan chub kyi yan lag bdun la sogs pa sgom pa'o //
de la 'phags pa'i lam yan lag brgyad pa gail fe na //

yan dag pa'i lta bani gan zig 'jig rten las 'das paste// bdag tu lta bas kun nas bslan pa rna yin pa dan// sems
can med pa dan // •tsho ba med pa dan // gso pa med pa dari // (skyes bu med pa dah)1 //gail zag med pa dan //
sed las skyes med pa dan // (sed bu med pa dan)2 // byed pa

po med pa dan// tshor ba po med pa'i lta bas kun nas

bslan pa chad pa dan.3 // rtag par lta bas kun nas bslan pa

1There is no support for this phrase in Skt.
2There is no support for this phrase in Skt; and it may be a gloss on the preceding phrase.
3Tib. chad pa rna yin pa dan. The negative verbal
form here is eccentric for the section while precisely this form as a verbal termination is omitted below. We judge the phrase to be misplaced.

rna yin pa dan// (srid pa dan)1 srid pa bral ba'i lta bas kun nas bslari pa (rna yin pa) dan2 // rigs su lun

bstan pa'i lta bas(kun nas bslan pa dan3)// ji srid 'khor ba rna yin zin rnya rian las 'das pa'i yan rna yin pa'i lta bas kun nas bslan paste// 'dina yah dag pa'i lta bar
gsuns so // gan {ig yan dag pa'i rtog pa ni 'dod chags dan//

{e ;dan dan// ti mug dan on mons pas kun nas bslan pa'i rto pa de ni yan dag pa'i rtog par rni 'gyur ro // yan
dag pa'i rtog pagan zig tshul khrirns dan// tin ne 'dzin

dan// es rab dan // rnam par 'grol ba dan // rna.rn par
, . . . . . .
'grol ba'i ye ses mthon ba' phun pos kun slon ba de' yan

dag pa'i rtog 'a de ni yan dag pa'i rtog par 'gyur te // 'di ni yah dag pa'i rtog par gsuns so II gah zig tshig gis bdag la gdun par mi bya in gfan la gdun bar mi bya ba
dan // bdag la mi bya in gzan dag on mons par mi bya ba dan // bdag la rni bya (in gfan dag la yan dbyas4 par rni bya ba ste II de mr{am par bzag pa dan ldan pas tshig dan

ldan par 'gyur tin // gan gi tshig gis yan dag pa'i 'phags par 'jug par 'gyur ba 'di ni yan dag pa'i rlag ces gsuns so// gan zig las na po dan// nag po'i rnarn par smin pa de'i
las mnon par 'du mi byed pa dan // gan {ig las dkar ba dan

dkar po'i rnam par smin pa de'i las mnon par 'du mi byed pa c!an II gail zig las dkar po dan II nag poI i rnam par

1Tib, omits, Skt. bhavo. The sense of the passage supports Skt.

2Tib. omits. Precisely this verbal form is added by Tib. above; we judge it to have been misplaced, See Note above.

smin pa de'i las mnon par 'du mi byed pa dan II gan
ig las (dkar po dan)1 II nag po dan II dkar po'i rnam ar smin pa dan II (nag po)2 zad pa'i bsam pas gnas ba de'i las mnon par 'du byed de II dam pa'i las lta bur 'gyur bani yan dag pa'i las kyi mtha'o II gan gi tshe 'phags pa lta bur dul ba dan II yon tan(yan dag pa'i
chos3)zum {in dul ba dan II kha gsag med ba dan II man du
smra ba rna yin ba dan II gtam la mi dga' ba dan II ran gi spyod pa'i khrims dan II 'jig rten pha rol gyi dban phyug dan ldan pa dan II bdag gi rned ba kha na rna tho ba med pas bdag chog par 'dzin ba dan II 'phags pa'i bka'i gnan
bas 'tsho ba ste II 'di ni yan dag pa'i 'tsho bar gsuns so II
rtsol ba ni gail Z'ig brdzun pa dan II 'phags pas ma gnan ba dan II 'dod chags dan I te sdan dan II gti mug dan II
non moris pa rjes su sten ar mi 'dod ba ni rtsol ba ste II

rtsol pa gan {ig yan dag pa'i 'phags pa'i lam gyi bden par 'jug cin mya rian las 'das par 'gro ba dan II so sor4 gnas su skyed par byed a de ni rtsol ba yan dag par rjes su 'bran ba ste 'di ni yan dag pa'i rtsol par gsuns so II
gan lane bar gnas pa de ni rabm mi g'yo ba dan II lus
dran pos kyog pa rna yin pa dan II 'khor ba'i skyon sna tshogs5 la sogs pa (dbah du byed pa)6 ston cin II mya rian

Jskt. omits the phrase, but the sense seems the same.
Tib. dbya

1There is no support for this phrase in Skt., but the sense is not disturbed.
2Skt. does not support this word although the sense is not disturbed; its place, however, is awkward.

las 'das pa'i gnas '(•dod pa'i1) dran pa ni (mi br jed pa2)
ste II 'di ni 'phags pa'i lam la3yan dag pa'i sbyar ba'oll 'di ni yan dag pa'i dran par gsuns so II yan dag pa'i tin

ne 'dzin ni tin rle 'dzin gan la gnas nas sems can thams cad rnam par thar par bya ba'i phyir yan dag par gnas te II
mya nan las 'das pa mnan pa'o II 'dini yan dag pa'i tin
ne 'dzin Zes gsuns so II de rtid yan dag b'ad ba ni II dga' ba la sogs pa rnams te4 II dban po la sogs ran bzin gyill byan chub sems dpa' rnams kyi gnas II phun bo la sogs
khyad bar gyi II sans rgyas thams cad kyi ni gnas II
sans rgyas byan chub sems dpa' yi II sans rgyas Xid ni mchog du 'ban byan sems sum cu rtsa gXis kyis II gan tshe 'dir ni gail gyur pa II dban po phun po 'byuri ba yi II
sans rgyas de yi de rna thag II ran lus id la sahs rgyas flid II g{an gan5 du 'an mi gnas so lllus las gzan du sans rgyas id II mi ses pa yis bsgribs par brjod II ye 6es chen po ran lus gnas II rtog pa thams cad yah dag
spans II dnos po kun Ia khyab pa po II lus gnas lus la rna skyes ba'o II rdo rje sBin pos gsol ba II lus la rtsa du lags II bcom ldan 'das kyis bka' stsal ba II 'khor lo

3This passage finds no support in Skt, and may be a gloss.
4 . .
T1b, so so'l gnas su
5Tib, sna chogs
6The phrase finds no support in Skt.
Corrected Skt. pr ita suggests a weakness in Tib.
here. The weakness may reflect, however, a doubtful Skt. line. See Notes above and below.

bfi yi rab dbye bas II brgya phrag gcig dan ni {u lhag II
byan chub sems kyi ran bzin gyis II rtsa rnams sum cu rtsa gffis mchog II gsuns pa II 'di lta ste111 mi phyed rna
zes brjod pa dan II phra gzugs rna dan II brtse ba rna dan2;1 g'yon pa rna dan II thun nu ma dan II rus sbal skyes rna
dan II bsgom pa ma dari II dban rna dan II skyon rna dari II 'jug rna dan II rna mo dan II mtshan mo dan II bsil sbyin rna dan II tsha ba rna dan II brkyari rna dan II ro rna darl II
kun dar rna dan II 'bab rna dan II raris rna dan II mdog rna dari II (gzugs can rna dan)3 II spyi rna dan II rgyu sbyi rna
dari II sbyor bral ma da.ri II sdug gu ma dan II grub rna dan II 'tshod rna dan II de b in du yid bzan rna dah II sum skor rna dan II 'dod ma dan II khyim ma dan II gtum mo dan II bdud

This phrase finds no support in Skt. and may be a
gloss to explain an ambiguous line.
3Tib. 'phags pa'i yan dag pa'i lam la sbyar ba'o. Skt. Arya marga sa prayoga . Mixed syntax is not uncommon for Tibetan translation.
Corrected Skt. muditadyas tu bhumaya .
Tib. ngan

1The thirty-two Ohannels and closing brief dialogue are found in HVT 1.1.16-20a.
Tib. translates divy . HVT rtse ba
3This reading finds no support in Skt. and, if admit­ ted, names a thirty-third channel which is either wrong or problematic. Tib., however, does parallel HVT. sin tu gzugs can rna which Snellgrove's edition claims to translate var,o . Despite Snellgrove's discrepancy, his lists total the expected thirty-two.

'dral ma'o II
(rdo rje s irl pos gsol ba)1 II

bcom ldan 'das rtsa 'di dag lta bu lags srid gsum yons su gyur ba ste II
(bcom ldan 'das kyis bka' stsal ba)2 II
pa dari 'dzin pa thams cad yons su spans pa 'oil


pa ne bar bzag pa ZJ.
dan II yan dag par sbon
ba bzi dan II rdsu 'phral gyi rkan pa bZ""J•.
dan 11 dban po
lna dan II stobs lna dan II byari chub kyi yan lag bdun dan II 'phags pa'i lam yan lag brgyad pa'i mthar thug pa
la sogs pa'i rnam par 'grol ba II byan chub kyi sems 'jug pa zes bya ba ste rab tu byed pa is pa'o II

1The connecting phrase has been added by the Editor.
See Note above.

(Chapter J)

de nas bcom ldan 'das la de bfin 6egs pa thams cad kyis mchod ba byas nas phyag 'tshal te 'di skad ces gsol to I/

gsan ba'i ye ses bla med pa //
s in po bcom ldan bsad du gsol II

de nas bcom ldan 'das kyis de bfin g6e s pa thams cad kyis gsol ba btab par mkhyen n2.::; rgyud thams cad kyi
s in po ye es mar me rdo rje1fes bya ba'i tin rle 'dzin
la snoms bar zugs te rgyud thams cad kyi gsan ba 'di bka' stsal ba 1/
gsan ba'i mchog gi dgyes pa na II thams cad bdag nid rtag tu b ugs II sans r yas kun dnos sems dpa' ni II rdo rje sems dpa' bde ba'i mchog II 'di ni bcom ldanrnal 'byor te /1 brtan pa2 rtag pa mchog tu yan II yid srab so sor skyes ba id 1/
rtag tu ran b in gnan dka' ba /1

1skt. omits. See translation. 2Tib. bstan pa. Skt. sthita

sna tshogs las kyi rnal 'byor gyis // sna tshogs cho ga 'dod pa yi //
sans rgyas rdo rje 'dzin sogs rnarns //

serns can rnams ni 'dul bar brjod // sans rgyas kun sogs brtan dan gyo // tharns cad dnos o srid pa 'di // sans rgvas kun dan mflarn sbyor ba //
mkha' • ro rna yi dra ba'i sdorn //

'dis ni s£yu ma'i rnal 'byor yis // tharns cad las ni sna tshogs rnchog // sans r yas sogs pas 'dul ba'i phyir // dnos grub serns can kun don rnchog // ran gi no bos yons bsgyur bas //
bud med sgyu rna tharns cad 'grub //

sna tshogs sgyu rna'i phyag rgya 'di // rnkha' 'gro rna fes kla klo'i skad //
ai bai ha ya sar 'gro bar //
skad 1 dbyins 'di ni rnarn par rtog2 // narn rnkha' kun tu sbyod 'grub bas // rnkha' 'gro rna fes rab tu grags //
kun nas sna tshogs phyag r£ya ste // kun las sna tshogs sdorn pa'o //
rdo rje rdo rje 'dzin pa nid //

pan rna de bzin pad rna 'dzin //

lskt. omits 2Tib. rtag

nor bu nor bu 'dzin pa nid II
'di rnams kyi ni ri s rnams yin II
de nas de bz"in s,e s 1 pa thams cad zil yis gnon pa rdul dan bral ba'i enas es bya ba'i tin ne 'dzin la
s oms par fu s te II byan chub kyi sems 'di bka' stsal
pa II
stori min storl ba min pa'an min II dbus mar dmi s par mi 'gyur te II es rab pha rol phyin sbyor ba II thabs ni s iri rje'i bdag nid de II
de nas mdzes pa'i s in rje thabs II es rab pha rol phyin sal ba II rnam rtog med pa'i cho :; rnams la II dnos po med cin bsgom pa 'an med II rnam rtog med pas rtag mos pas II brta pa kun yis brtag par bya II rnam rtog med pa'i chos rnams la II sems can don ni yons su brtag II
de bzin §e s pa kun bda chos II chos can med cin chos id med II sgra br an lta bur m am pa id II
de naS I di ni ChOS nag 1b,YLl: //

de nas theg pa chen po las byun ba'i bsgom pas yon tan rgyas pa rnams kyis dus gsum gyi de bZ'in g6e s pa


thams cad kyis bstod pa'i rgyal po 'dis yan• bstoct pa II

dag par
rnal 'byor gyi bdag po sems can grol bar mdzad pa la bstod cin phyag 'tshal lo II thams cad 1 bdag ffid skyes pa gcig par mdzad pa la bstod cili phyag 'tshal
lo II 'khor ba'i rgya mtsho'i gti mue: gcod par mdzad pa la bstod cin phyag 'tshal lo II de kno na thams cad kyi

ve es gcig du ston par mdzad pa la bstod cin uhyag
'tshal te II bdag rtae: tu phyag 'tshal lo II yan mchod pa byas nas phyag 'tshal te 'di skad ces gsol to II

chos rnams thams cad gcig pa'i sku'ill snin po bcom ldan b6ad du gsol II
bcom ldan 'das kyis bka' stsal pa II
dbail po gail dan gan lam id II
de dan de yi rail bzin 'gyur II
mnam par rna bzag
mal 'byor gyis II
rtag bar 6in du mfiam bzag id I
gan phyir thams cad bda nid kyi II rnam ses phun bor gnas bani II skyes bu tha mal sems rmons pa II la las so sor rtoe: mi nus II
sems dan sems byun sems kyis dnos II
ye ses ses bya
rail bzin can II
'bras bu rnams la byed pa i{id II

1Tib. sems can. Skt. sarva-

chos dan chos nan1bs om ba id II
gan phyir chos ni du mas ni II
chu bo'i rgyun ni rgya mtsho ltar II
thar pa id ni gcig pu yin II mail po dmigs par mi 'gyur ro II ji 1tar nail du rnam .mas pa II
bla ma'i z/al las rned par 'gyur II

de kho na fiid mdor bstan pa ste rab tu byed pa trsum pa'o II

1Tib. chos can, Skt, dharman m

(Chapter 4)

(bcom ldan 'das kyis bka' stsal ba)1//
de nas rgyud rnams thams cad du /1 gtan la phab pa rab bsad bya II sems can kun la m am pa'i nhyir II
sku sum rgya che'i2 gzugs can flid 1/
tshans so,gs lha dan lha min bcas II
de rnams kyi ni gnas su 'gyur II
de id ses rab pha rol phyin II
kun rdzob rnam pa'i gzugs can rna II
de id yul las rnarn par 'das I/

sroe- cha s kun ,gyi
s"n"' 1.n. la gnas II
de phyir sa rgya3 ci
bya II
mdor bsdus pas ni sans rgyas id II
bskal pa bye ba rans med nas II ji ltar sans rgyas gan thob pa II gan gis dam pa'i bde bas khyod II skye ba 'dir ni thob par 'gyur II yan na-rdo rje 'dzin pa id /1
yah na 'khor los sgyur ba 'am II

The connecting phrase is added by the Fditor to
facilitate translation.
Tib. cha'i

3na rgyal. Skt. vistarato

'dis ni dnos grub chen po brgyact II
g an yan 'di la 'dod pa 'an II
gti rnu chags sdan ria rgyal dan II phra do dah ni on mons lnas II sems can so sor bcins pa an II
ran i yan la flid kyis 'jams II
'dis ni • ro dru serns can bci'ns II

'khor bar 'khor (in skye ba fiid //

nNon mon•s pas n1•

rrnon• s 1

yur uas //
sdi pa dum byed par • yur II
de phyir 'di beam cin bsre id 2 II sans rgyas kyis sprul rigs pas ni II 'khor ba'i r ya rntshor lhun ba yi II sbyor bral sdug bsnal rnthon nas ni II thabs dan ges rab sprul pa'i r yus II

nNon mon• s rnams n1•
yid ches 'gyur II
mtshon u 'di ni 'bras bur gyur II khams sum du ni dri med snan II an dan gan gis as gyur ba II
de ni de yi no bo can II
bud in bsre s pa'i rigs3 pas ni II
dgyes pa'i r yal ba gar rnkhan mchog II
phun po khams dan skye mched rnams II

Tib. mons

mkhyen n.1d •

Skt • v1'dagdha

3Tib. rags. Skt. yena. While the corrected form appears elsewhere in the Text in a similar context, the Skt parallel is weak here.

•di flid rnarn dag mtshon pa 'di II
ji ltar d ra po nus pa la II
d ra bo nus pas de bsnun pa II
ti mu.2" da.2" pas gti mug ste II

ze sdan


pas ze

sdann""'• d II
'dod chags dag pas 'dod chags te II na rgyal rnam da na rgyal che II phrag do dag pa.s phrag dog T{id II
tharns cad rnarn dag rdo 'dzin mgon 1/
de yi no bo rnarn dag pa II
on mons lna ni am par 'gyur II
de ltar rigs ni lna rnams dan II
ye ses ln.a dan. . 1
ln.a II
'di las rdo rje s in po dan II
khro bo khams .2"SUm sems can skye II
B:i dbye ba t din1\1. d
_, N
n1. II
bla ma'i
zal las rned par
'.2"yur II

judge Tib,
bla rna dman ba'i sems can rnams II snae:s dan phyae: rg:va med mi •e;:vur II gail Z'ig S'in du gsal bas ni I12
'dzam3 glin de 'dir sans rgyas bsna s II
gru gsum dag pa e yi dbyibs II

1Tib. das r yas 2
There is no equivalent ?ada in the Mss. here. We
a misplacement has occurred and decide in favor of
See Note to the Skt. edition.
Tib. mdzam

dbus su dgyes pa'i (e ba )1yin//

ru gsum dkyil 'khor ams dga' bar// rdo rje ra li2 (rnam par byuri)J //

btsun mo'i pha ga fes kyan bya // chos kyi 'byun gnas £es byar bsad // de yi nan son pad rna //
'dab rna brgyad ba s in por bcas //

de la a li ka li 'dres //

sde tshan brgyad ni rnam par gnas // snags kyi gzugs kyi lus can gyi // las kyi tshogs ni byed pa • vur // yi ge rnams ni lna bcu id //
rig byed rnams la e bar enas //

snags rnarns id dan rgyud rnarns dan // phyi rol tshul gyi bstan bcos kyi //
rdo rje s iri po'i ran bfin 'di // gzan du (a las)4 can zad rned //

a ka ca a ta ba ya sa//

sde tshan (dga' ba)5 lna bcu yan // am pa dag du sprul pa rnarns //

corrected Skt, parallels Tib,; but the presence of
evam here seems meaningless. See Note to Skt. edition,
This is a transliteration of what appears to be a
technical term in Skt., the meaning of which has not been determined.

3The phrase parallels Skt. vinismrta in a very loose way and is, therefore, doubtful.
skt, ebhyo appears to lie behind a las, a question­
able but not entirely inaccurate parallel,

rdo rje ra li pad mar son II
'dab rna brgyad po phyoe:s rnams dan II mtshams rnams su ni 'di rnams rigs II de rnams dbus su ze 'bru la II
dam pa'i dban phyug rna yod do II
sde tshan brgyad po rnams kyis ni II
yi ge dam pa yons su bskor II ani yig 'bru kun gyi mchog II don chen sde tshan gtso bo ni II
lus can rnams kyi snags kun ni II 'di id las ni yan da 'byun II gan Z'ie: ral gri mig sman dan II rkan byug ri 1u1 sa 'o2 dan II e:nod sbyin mo'i yan lag rna II
'jig rten gsum gyi khons su rgyu II
dnul chu'i 'gro bas rjes su 'bran II

z"l'•.g thams

cad mthu che ba II

rarl gi khyam du (lna bcu)3 'dod pa dan ldan par ('grub bo)4 1/

5There is no support in Skt.
This is a contraction for ril bu. See Wayman,
Yoe:a, p. 291.
2 • .
TJ.b. rkan 'og

3see Note below.
Skt. fvr the Pada, svabhavane satyam(corrected) ca k m vitas. The lack of parallel here and irree:ular meter in both Skt. and Tib. leaves both lines doubtful.

de ni sde tshan brgyad nan son //

dbyans rnchog sde tshan bdag id byin // skye bo rnarns kyi tshig rgyas na //
gan ig cun zad rtogs bya ba //

de kun snags kyi gzugs kyis ni // de phyir rab tu skye bar 'gyur // lus can rnarns ni tharns cad kyi // sgra ni snags es brjod par bya //
sna tshogs rndud pa'i gnas chen las //

chos kyi ra li rnarn par 'byun //

dnos grub rdzu 'phrul rnam gnas pa'i // snags dan nes pa an yan rned //
ran bfin gyis ni rna skyes pa'i //

snags kyis yig 'bru dban phyug rnchog // yan dag sbyor 'byun rntshan id ni //
de nas rncho du rab b ad bya // e ni sa ru (es par bya //
las kyi phyag rgya spyan rna nid //

s in rje chen po thabs chen po //
bz"'1•n sna tshogs spyod yul rna II
lte bar sna tshogs1 'darn skyes la // sprul pa'i 'khor lor rnam par gnas // ba ni chu ru es par bya //
chos kyi phyag rgya rna rna ki //

1This very loosely translates Skt.

byams pa dan ni smon lam tshul1 II
lha mo rdo rje'i rigs byun ma 2 II
s in khar chu skyes 'dab bryad lall chos kyi 'khor lor yan dag gnas II rna ni me ru3 brjod pa ste II
phyag rgya chen po gos dkar mo II
dea' dan stobs kyi rnam 'byor gyis II
lha mo pad ma'i ri s byun rna II
lkol.t mar chu skyes 'dab brgyad is II lons spyod 'khor lor yan dag gnas II ya4 ni rlun gi ran b;in te II
on mons thams cad rab 'joms byed II
dam tshig phyag rgya chen po nes II
lha mo las kyi rigs byun ma5 II

sNnoms ye ses mal 'byor gyis II
sgrol rna 'khor ba sgrol ma'o II 'dab skyes 'dam rna sum bcu g is II bde chen 'khor lo chen por gnas II e ni ses rab brjod uar bya II
ba ni thabs dan ldan pa ffid II


This is an unusual translation of rupa.
Tib. gtso mo. Skt. udbhava, Since byun rna
for the same Skt,, we have corrected in that

3Tib, rna ru 4
Tib. ya. Skt. ya

appears way.

5Tib. gtso mo. Skt. udbhava, We have corrected for
a form appearing elsewhere in this section, See Note above,

ba gyis rnam par brgyan pa id II e ni nes par mdzes par 'gyur II thabs dan tes rab ran bZ'in gyis II

sten dan 'og du snoms par
'jug II
e 9{id dan ni ba'Ql n'id nag II

( gnNl•.S
g¥fis su ni brjod pa n'id)1 II

'di skad ces bya ba Il•.
tshig gi II
phrad ni de skad ces gsuris nas II

bya ba'i mthar thug pa'i rgyud II


tshogs la bsnegs so
es brjod do II
bdag gis thos pa zes bya ba ni de skad du 'dod chags chen po 'di la rtag du gnas pa'o II sru {es bya rna ni
'dzag par gsuns so II tam2 zes bya bani gah yan bde ba
chen po'i mgon po'o II bdag gis thos pa zes bya ba ni (bdag id kyis thos pa'i phyir)3 rna ba'i ye te kyis thos pa yin gyi II
(rdo rje s irl pos gsol ba)4 II
rtogs pa ni rna yin no II bcom ldan 'das mi rto?s pa'i skyon gan lags II
bcom ldan 'da kyis bka' stsal pa II
sdud par byed pa dan 'chad pa po II
dag dbyer med par 'gyur ro II

1skt. does not show an equivalent Pada. The Tib. addition is probably a translator's gloss.

2Tib. ta. Uncorrected Skt. omits. We correct for tam in both editions since the context of the line requires it. The superscript of Tib. is easily losta for more dis­ cussion, see Note to the Skt. edition.

yan na rto s pa id ni gdul II
bya'i skye bo'i dban du byas na II

'chad ba po id dan sdud par byed pa yin te II
had pa po na chos kyan ria II
ran gi tshogs ldan ffan pa na II
(rdo rje s in pos gsol ba)1 II
de dag ji ltar 'tshal bar b yi II
(bcom ldan 'das kyis bka' stsal ba)2 II

gan du bde ba chen po'i mgon po gar mdzad pa'i (tshe)3 II gcig id gcig dan du ma'i ro fes bya ba'i tshig gis (rol par mdzad pa'o)4 II rigs kyi bu bcom ldan 'das
kyis gan cun zad gsuns pa de bdag gis gcig na es bya ba
ni dus gcig id la'o II gcig na (es bya bani yud cig
la'o II thos pa zes bya bani lhag par son pa ste II 'dis bsam gyis mi khyab pari tin ne 'dzin rfled do zes bya bar
bsflegs 5so II

4The connecting phrase has been added by the Editor to facilitate translation.

3This phrase does not appear in Skt. and seems to be in the nature of a gloss.
1The connecting phrase has been added by the Fditor. 2
see Note preceding.
3skt. does not reflect this word but reads nrtyaty at this point in the line. The discrepancy does not ser­ iously affect t:::•anslation.
This phrase does not appear in Skt.

5skt. taccitta. The meaning of "aspiration" carried by bs egs makes the translation loose but possible.

dus z"'es bya ba ni dus tshad ces gsuns so //
dus ni rnam pa gsum du brjod // bde ba'i dus dan sdug bsnal dus II bsam gyis mi khyab dus fiid do II 'o ma'i rgyun ltar sna bug du /I 'on ba bde ba'i dus su gsuns II me'i gzugs kyis • ro ba id 1/
sdug bsnal dus dan de gcig grags II
cig ni lhan ciB; med pa yin I/
dus ni bsam gyis mi khyab 'gyur II 'don chags med cin chags bral med II dbu mar drnigs par mi 'gyur ro /I
de la 'dod chags nus mtshan id /1 'dod chags bral ba 'gog par gsuns II 'di bral dbu mar mi rtog pas /i
'di sum dmigs par mi 'gyur ro /1

'dod chags T{id dan chags bral nid II
1\1 II
par f!..yur pa rnog pa med
de b.z...1. n ro mnam

'dod chap.:s chags bral las II
gcig bu'i skad cig rna II
dnos po kun gyi ro rn?tam pa II
bsgom pa'i dus ni gcig tu gsuns /1

(bcom ldan 'das de gsuns)1 II
'di lta ste II dban ph,yugdan II gzugs dan II
rrags pa dan II dpal dan I ye ses dan I brtson •grus (phun sum tshogs pa ste)2 II bha ga drug ces gsuns so II de 'di la mna' ba'i phyir bcom ldan 'das so II yari na

gari fig mi mthun pa'i phyo s kyis chos de rnams bcom pa'i phyir bcom ldan 'das RO II de bzin g egs pa thams cad
kyi sku dan gsun dan thugs rnams dan de rnams n d kyi
s irl po dan II de id rdo rje dan de nid kyi btsun 0 dan II
de'i bha ga de la bzu rs so II
(rdo rje s in pos gsol ba)3 II

lags II

bcom ldan 'das bzugs so


bya bas ci
bcom ldan 'das kyi bka'stsal lo II

de bzin gsegs pa thams cad kyi ro ci pa thabs sna tshogs kyis 'dul ba'i skye bo'i yid 'phrog pa dan ldan pa
zes bya ba'i don to II de bzin gsegs pa thams cad kyi Rku

dan gsun dan thugs kyi s in po rdo rje'i ran bzin gan gi chos kyi 'byun as la II an gi tshe bcom ldan 'das bzugs pa cte'i tshe bdag gis thos pa'o II 'di lta ste II ses rab
aan gis on mons pa dan ne ba'i on mons pa 'joms pa
de'i phyir es rab bha gar gsuns te II bha ga der de b{in

1Tib. omits this passage. Skt. bhagavan sa uc rate. We judge that Skt. is authentic with the loss of a passage in Tib. due to confusion of initial elements in adjacent passages, i.e. confusion of bsgom and bcom. The Editor has reconstructed in a tentative way. See Note to Skt.
The phrase does not appear in Skt. and is doubtful.
3The connecting phrase has been added by the Editor.

gsegs pa thams cad btsun mo lhan cig du bzugs so II
(rdo rje s in pos sol ba)l II
e rna sans rgyas thams cad kyi II kun mkhyen ye {es bla named /1
(bcom ldan 'das kyis bka' stsal ba)2 II
de bfin 6egs pa'i bde don gan II btsun mo e bar mdzad pa'o II
rigs kyi bu 'di skad bdag gis thos pa z""es bya ba'i yig 'bru ni chos kyi thog mar rtog tu yan dag par gnas
pa ste II dag pa dan yan dag par rdzogs pa'i sans rgyas
kyi dnos grub skyed par byed pa'o II lag na rdo rje rna
bsad pa de nas (khyod la)3 bsad do II tshig 'bru 'di rnams
sems can rnams
'khor ba las pha rol du sin tu
('gro ba mnon du)4 mtshon par mdzad pa dan II mnon du


pa de nub pa'i


kyis skye ba 'dir sans rgyas

'am5 rdo rje sems dpa' thob par 'gyur te II gail dag bsam
gyis mi khyab pa'i gnas rna thob pa ni
bde bar segs pa
ste sans rgyas yin no II mtshan gzi mtshon ar byed pa ni
(sems par)
yan dag par gsuns te (sdan bas)7 bsten pa can

1The connecting phrase has been added by the Editor.
2See Note preceding. 3skt. shows no parallel.
4 . .
There l.S no support 1.n Skt. for what may be a gloss.

5Tib. dam, Skt, v[
Skt. does not reflect the term,
7see Note preceding.

gyis mnar rned par rab tu ltun par 'gyur bas na II kun nas on mons pa can gyi sems kyis (ro)l span par bya'ol/
kun nas non mons pa can gyi srid pa'i 'jigs pa las 2
('das pa'i)
sgrub pa po II dpal dan ldan pas dag pa'i
sems sgorn ar byed de thabs dan ses rab kyi sbyor ba 'dis

ji ltar yan dag pa thob pa dag pa'i yan dag par rig pa'i bdag ffid can kyi mtshan gzi thabs dan (es rab kyis nus
pas nam rnkha' don mfiam zin gnas gsum du s:nan ba ste II

de'i mtshon gzi ni mtshon par dka' ba dan thams cad du son zin rgyu d n rkyen spans pas yid bfin gyi nor bu lta
bur 'jig rten gyi(ran dan g;an gytPdon byed de II dnos
grub thams cad kyi mchog go II

rgyud thams cad kyi glen gzi gsan ba dpal yan dag

par sbyor ba las byun ba'i brtag pa'i rgyal po ste dah po rdzogs so 411

1Skt. satya. The terms are doubtful although related in meaning.
Skt. does not reflect the term which appears to be
a .Q;loss.

3see Note immediately above. 4Tib. sho


(Chapter 1)

(Om.) Homa e to VajraQaka.

Thus I heard at one time1 The Bha avat was dwelling in the bhagas of the Vajra-ladies, (the heart) of the body, speech, and mind of all Tathagatas.
Now then, seein Vajragarbha in the midst of ei ht

hundred million lords of yo a, the Bhagavat smiled.

No sooner did he smile than Vajra arbha rose from his seat, threw his upper garment over one shoulder, and put his right knee on the earth. Stretching forth joined palms, he petitioned the Bhagavat with these words•
0 Lord of Knowled e, I wish to hear the character­ istic of "arising from samnu a," the secret of the nidi!na of all tantras.
The Bhagavat exclaimed:

Oh, Vajra arbhal Excellent, excellent, Great Com­ passionate Onel Excellent, excellent, Great Bodhisattva! Excellent, excellent, Source of Virtuesl Whatever is secret in all the tantras, ask for all of it just as you wish.

Then, with Vajra arbha at their head, their eyes open wide, thrilled, bowing again and again, those Great Bodhisattvas voiced their doubts on this occasiona
What does "all tantras" mean? And in what sense is there a "nid na"?
What is "secret" there? What does the name "arising

from samputa" mean?

What is the "characteristic" there?

The Bha avat answereda

"All tantras" is "all" and those "tantras." The phrase "all tantras" means the §ama,ja, etc. The "nidana" of them means the inevitable rule of their arising. It is "secret" because it is not within the range of Hari,
Hara, and Hira yagarbha or of Sravakas and Pratyekabuddhas. "Samuuta" consists of mea."'l.s and insie:ht. "Arising" from the tantra means the samputa-equipoise; "arising" is the process of arising. Such a state consists of the intrinsic
nature (and universal nature) made up of the stationary

and the movin,g. What is characterized by this is the "char- acte:ristic."
Or the word "samputa" as the nidana of all tantras means Vajrasattva. By the word "secret" is meant con- struction of mantras according to the precepts and the character of initiation and deities of the mandala ret- inue.

This is a tantra consisting of means and insight.

Listen to what I declarel

Having imagined voidness first, one washes away the impurity of embodied beings.
One imagines "sphere of form" amon,!! the voids and does the same with "sound."
One imagines "sphere of odor" among the voids and

does the same with "tae:te."

One imagines "sphere of tangibles" among the voids and does the same with "thought."
Vajragarbha questioned:

If one imagines "true form of the eye" amone: the voids, how can there be perception in it?
If one imae:ines "true form of the ear" among- the voids, how can there be perception in it?
If one imagines "true form of the nose" among the voids, how can there be perception in it?
If one imagines "true form of the tongue" amonp.: the voids, how can there be perception in it?
If one imagines "true form of the body" amontr the voids, how can there be perception in it?
If (one imagines) "mind as external" among the voids, how can there be perception in it?

In such a case, there is neither form nor seer, neither sound nor hearer,
neither odor nor one who smells, neither taste nor


neither tan ibles nor one who touches, neither thought nor thinker.
'l'he Bhagava t answered:

Listenl Since the path to reality avoids duality and is nondual,
is imperceptible and non-rational, it is not compre- hended by logicians.
'l'he secret of all Buddhas is like the sameness of the


r vakas do not discern it since they are hindered by the darkness of ignorance.
Those manifested as Pratyekabuddhas (do not discern)

the knowledge which leads to the end of habit- energy,
Indeed, none of them know the secret which is within the range of Buddhas.
There is said to be the character of a ford for

sentient beings turning in the middle (of sa sara), who have minds gone to the ford out of fear of the
ocean of sam.s ra •

One should observe the destiny of knowables, looking at the knowable with the knowledge
of minds which have the perception and knowled e of the act of characterizing, the thin s character­ ized, and the characteristics.
Minds which observe the destinies are able to @:O

wherever they wish.

They o to destinies of former lives; the yogin (goes) to many thousands.
When the characteristic -- that is, the char cteriRtic

mentioned above -- is contemplated in the void, the wise one casts the seed into the void fields
of body.

That which has issued from the beautiful head of the channel is said to be free from dust.
The bodhicitta is brightly shinin like pure crystal. That entity of five knowledges is as small as a mus­
tard seed.

The deity located in the middle of that has a form both manifest and unmanifest.
Half the size (of a letter), very subtle, it has the

form of a drop and is made of mind.

Possessing light rays of great splendor, it always dwells in the middle of the heart.
It iR at the limit of the twelve (sense bases) and

at the limit of the nine (orifices), leaving out the soles of the feet and the head.

The sin le thread having issued, (that issue) is placed in the middle of the nave.
Amounting to only five parts, it expands into the shape of a Serpent Chief.
A small portion of that having been emitted, it is

drawn out of the middle of the vajra.

The seed is nut into the middle of the yoni. When the dharmadhatu melts,
it transits sequentially in all nine orifices.

Accordingly, (the seed syllable of) fire (is applied to) the (orifice of) Brahma •.
As for the twofold seed of earth, precisely that is applied to the eyes.
As for the seed of the voids of wind, (it is applied)

mentally to the nostrils and ears.

As for the good seed of ambrosial water, the lord (applies it to) the mighty tongue organ.
The seed of delusion (is applied to) the neck while that of (another) defilement is likewise (applied to) the two arms.
(The seed syllable of) vibration (is put into) places

in the heart while that of animals (is placed) in two roots of the navel.
One should hold on to the eight members of seeds with the rite of eight members.

The Pervader pervades the whole world, the station­ ary first and then the moving.
Brahma, along with other Suras and the Asuras, is the substratum of this.
The abode in which the master roams is called "bhae:a."

As long as he is stationed in the body, he has action and no action.
He performs all acts, either good or bad.

The yoga of sameness is said to be the contemplation of worship.
When the action body is broken onen, he becomes divine

like that one (worshipped).

At that time, he who pervades the stationary and the moving is his own akti.
One should know his color to be like the sameness of the sky.
The Hero stands in nirvan.a, unsullied and immaculate •
Mother, Sister, Daughter, and Female Kin,

Brahman Maid, K atriya Maid, Vaisya Maid, and §udra Maid, Dyer, Dancer, Outcaste, and Ca galini --
These one should worship by the ritual of means and insight with devotion to their reality.
They are to be served with zeal so that separation does not occur.

If the secret is not kept, one reaps suffering from creatures of the earth, (animals such as) snakes and (humans such as) thieves.
The mudras are said to be five-fold when classified by the division of Families.
The Brahman Maid is born into a Twice-born family;

and she belongs to the Tathagata (Family).

The K atriya Maid is of royal lineage, born into the Peacock and Moon families;
she is said to belong to the ambrosial Vajra (F'amily).
The Vai ya Maid is a cowherder; and she belongs to the Karma Family.
The Stldra Maid is a servant; and she belongs to the

Great Vairocana (Family).

The Dancer belon s to the Padma Family, and the Dyer to the Karma Family.
The Outcaste belongs to the Vajra Family; Ca 9 lini

belongs to the Ratna (Family).

The five mudras are determined in this way. Concisely, these are called the Families of the Ta­

Having gone to Thusness, the Glorious One has come exactly thus.
By way of this insightful principle, he is called


The Families are said to be five-fold; there is also a hundred-fold (division) and an infinity of Families.
Finally, with the distinctions of body, speech, and mind, one arrives at a three-fold division.
Since the Families are the intrinsic nature of the

five elements and five skandhas,

they are the Families of (the Steps) of holdin the vajra, the jewel, the lotus, and the gem.
There is neither contemplator nor object of contempla- tion; there is neither mantra nor divinity.
The unelaborated intrinsic nature abides in mantra and


Along with Ak obhya, Vairocana, Amogha, Ratna, Arolika, and S ttvika,
he is called Erahma, Vi u, Siva, Sarva, Vibuddha, and Tattva.
He is called Brahma, free from obscuration and awakened;

he is called Vi pu because he pervades •
He is called Siva because he is always auspicious;

Sarva because he dwells in all selves.

He is called Tattva on account of sublime bliss; Vibuddha due to his comprehension of pleasure.
He is called "divine" because he is "born in the body." He who "possesses the bhaga" is called "Bhagavat."

People speak of six kinds of bhagas, all the merits of lordliness, etc.
Or, he who is called "Bhagavat" is the "one who destroys the defilements."
Insight is called Mother because she engenders all

living beings.

Insight is called Sister because she shows potentiality.

Insight is called Dyer because she colors sentient beings.
Insight is called Daughter because she draws the milk of ood qualities.
Insight is called Dancer because she moves •L;o and fro

with great compassion.

Because Bhagavati is untouchable, she is called Outcaste.

What is called engaging in muttering is enunciating the series of vowels and consonants.
What is called a mandala is a footprint because a mandala makes an impression.
A rnudra is the opening of the hand and also the pres­ sing with a finger.
Whatever is meditated upon is thought; thus, what is thought is meditated upon.
That bliss which is receive0 from the Father is the

bliss which is enjoyed by one's self.

With whatever bliss one dies here, that bliss is what is called meditation.

Here ends he first chapter on the reality of con­ templating the subject to be expressed and its expression, the production of the bodhicittc, etc.

(Chapter 2)

(The Bhagavat announceds)

I shall now proclaim the thirty-seven Dharmas Accessory to Enlightenment.
Because he dwells observing bodies outer body, both inner and outer body
inner body, he is rightly
insightful. He possesses mindfulness toward the world

which is to be trained but which is hesitant due to ignor- ance,
Because he dwells observing feelings inner feel- ings, outer feelings, both inner and outer feelings -- he is rightly insightful, He possesses mindfulness toward the world which is to be trained but which is hesitant
due to ignorance,

Because he dwells observing dharmas -- inner dha . outer dharmas, both inner and outer dharmas -- he is ri ht­ ly insightful. He possesses mindfulness toward the world which is to be trained but which is hesitant due to ig­ norance,
Because he dwells observing thoughts -- inner thoughts, outer thoughts, both inner and outer thoughts -- he is rightly insightful. He possesses mindfulness toward the

world which is to be trained but which is hesitant due to ignorance.
These are the four Stations of Mindfulness.

In order to stop the production of sinful and de­ meritorious dharmas not yet produced, he enerates longing, he endeavors, he exerts striving. He takes hold of his mind and exerts it fully.
In order to eliminate sinful and demeritorious dhar­

alr.eady produced, he generates longing, he endeavors, he exerts strivin . He takes hold of his mind and exerts it fully.
In order to produce meritorious dharmas not yet pro­

duced, he generates longing, he endeavors, he exerts striv­ ing. He takes hold of his mind and exerts it fully.
In order to preserve meritorious dharmas thus gener­

ated, in order to fulfill them by way of the immeasurable­ means contemplation, and in order to enhance them further by way of the sublime contemplation, he generates longing, he endeavors, he exerts striving. He takes hold of his mind and exerts it fully.
These are the four Right Elimination-exertions.

He contemplates the basis of magical power which is accompanied by the elimination-exertion motivation of the longing sami!idhi. He thinks, "Since I have become certain due to discrimination, certain due to aversion, certain due to cessation, and have become ripe for escape, let not my

longing become too faint or too grasping.

He contemplates the basis of magical power which

is accompanied by the elimination-exertion motivation of the striving sarnadhi. He thinks, "Since I have become certain due to analysis, certain due to aversion, certain due to cessation, and have become ripe for escape, let not my striving become too faint or too grasping.
He contemplates the basis of magical power which is accompanied by the elimination-exertion motivation of the examination samadhi. He thinks, "Since I have become cer­ tain by analysis, certain by aversion, certain by cessa­ tion, and have become ripe for escape, let not my examina­ tion become too faint or too grasping.
He contemplates the basis of magical power which is

accompanied by the elimination-exertion motivation of the thought sarnadhi. He thinks, •"Since I have become certain through analysis, certain through aversion, certain through cessation, and have become ripe for escape, let not my thought become too faint or too grasping,
These are the four Bases of Magical Power, together with coursing in desire.
Having faith in the mundane right view, he trusts in the ripening of his own karma in each case. He thinks, "Whatever action I perform, I shall experience the matura­ tion of that action, either virtuous or sinful." He will not instigate a sinful action even for the sake of his own

life. This is called the organ of belief.

Whatever dharmas he believes with the ?rgan of belief, those dharmas he acquires with the organ of striving, This is called the organ of striving.
Whatever dharmas he acquires with the organ of striv- ing, those dharmas he preserves with the organ of mindful- ness, This is called the organ of mindfulness,
Whatever dharmas he preserves with the organ of mind-

fulness, those dharmas he makes the object of a single area of thou ht with the organ of sam dhi. This is called the organ of sam dhi.
Whatever dharmas he makes the object of a single area of thought with the organ of sam dhi, those dharmas he penetrates with the organ of insighta and that engenders discrimination among these and those dharmas. This is called the organ of insight.
These are the five Organs (which are also) the five


(The latter are) as followsa power of belief, power of striving, power of mindfulness, power of sam dhi, and power of insight.
These are the five Powers.

What then are the seven Limbs of Enlightenment? They are as followsa the mindfulness limb of enlightenment, the higher analysis of dharmas limb of enlightenment, the striv­ ing limb of enlightenment, the joy limb of enlightenment,

the cathartic limb of enlightenment, the samadhi limb of enlightenment, the equanimity limb of enlightenment. He should contemplate the seven Limbs of Enlightenment being certain due to discrimination, certain due to aversion, certain due to cessation, having become ripe for escape, and possessing the higher analysis of dharmas.
This is the eightfold Noble Path.

What is right view in it? It is that which is supra­ mundane which does not arise with the view of self-- of soul, being, person, feeder, man, actor, or knower, or of something eternal or annihilated, or of existence or non­ existence. It (does arise) with the view of the prophecy
of Family. That view is called right view as long as it arises with the view that there is neither samsara nor
nirvan.a, (neither sams ra nor parinirv
With whatever discursive thoughts lust, hatred, de- lusion, and other defilements arise, those discursive thoughts are not right discursive thoughts. With whatever discursive thoughts morality, samadhi, insight, the know­ ledge and vision of liberation, and the skandhas arise, those discursive thoughts are right discursive thoughts. This is called right discursive thought.
Whatever speech does not enflame or defile, or ridicule oneself or others, that is accompanied by that speech which is joined to equipoise. With whatever speech one understands the rightly noble, that is called right speech.

Whatever is a black action with a black maturation, that karma (the yogin) does not instigate, Whatever is a white action with a white maturation, that karma he does not instigate. Whatever action has a black-white matura­ tion, that karma he does not instigate. Whatever action he dwells upon for the purpose of exhausting white and
black maturation, that karma he does instigate. True action

is the support of right bodily action, This is called right bodily action.
Whenever one has the qualities of a disciplined man like the Nobles, (his) livelihood is authorized by the Nobles -- having conquered faintheartedness, lacking hypoc­ risy, chattering, and enjoyment of stories, being self­ regulated by the norm, joined to the Lord of the other
world, blameless with regard to one's own gain and one's own satisfaction. This is called right livelihood.
Whatever effort is deviant and not authorized by the Nobles, (the yogin) does not wish to follow; (such efforts) are characterized by lust, hatred, delusion, and other de­ filements, Whatever effort focusses upon the path leading to nirv a, accompanied by understanding of the truth of
the right Noble Path, that effort he rightly follows, This

is called right effort.

Wherever there is a mindfulness which is steady and not shaking, accompanied by a body which is straight and not crooked, displaying (mastery) with regard to the faults

of sa sara, it leads along the path to nirv a. This is joined to the Noble Path. This is called right mindfulness.
In whatever samadhi he dwells for the sake of the liberation of all sentient beings, he is rightly stationed; he is progressing toward nirv a. This is called right samadhi.
Having explained these, (I now explain) the Joys as abodes.
The home of the Bodhisattvas is the same for the true form of the organs.
The home of all Buddhas is the same for the skandhas in a distinguished way.
The best lovers of the Buddhahood of Buddhas and Bodhi-

sattvas are accompanied by the thirty-two bodhi­ cittas.
Whatever becomes fulfilled in this life among existing

organs and skandhas, that is the infJ..nite Buddha- hood.
Buddhahood is in one's own body and is not located


Because they are hindered by ignorance, people say that Buddhahood differs from the body.
The Great Knowledge is located in one's own body; it is free from all discursive thought.
It pervades all given things; located in the body, it

is not body born.

Vajragarbha askeda

What are the channels in the body? The Bhagavat declareda
According to the variety of four cakras, there are one hundred and twenty.
According to the true form of bodhicitta, thirty-two

are said to be most important. They are as follows:
abhedya, suk marup . divya, varna, vamani, kurmaja, bhavak!, sek . do a, vistha, mata, arvari, It , do m , lalana, rasan , avadhuti,
nravan . hr.s.t.a, varn.a, samanya, hetudayika.
viyoga, premani, siddha, p vakt, suman ,

t varta, kamini, geh , candika, and maradarika. (Vajragarbha askeda)
0 Bhagavat, of what sort are these channels which develop in the three realms of existence?
(The Bhagavat declareda)

They avoid every apprehender and everything appre­ hended.
Here ends the second chapter entitled "The Understand­ ing of Bodhicitta" which explains fully the fo'...!:r. Stations
of Mindfulness, the four Right Elimination-exertions, the four Bases of Magical Power, the five Organs, the five Pow­ ers, the seven Limbs of Enlightenment, and the eightfold Noble Path.

(Chapter 3)

Then, all the Tathaeatas, having paid homa e and having bowed, petitioned the Bhagavat thuss
0 Bhagavat, please explain the heart, the supreme knowledge of the secret.
Then, the Bhagavat, reco izing the petition of all the Tathagatas, became equipoised in the sam dhi called "The (Diamond of) Knowledge Light which is the Heart of All Tantras" and declared this to be the secret of all tantrasa
When one enjoys the supreme secret, he is always abiding in the universal self.
The being consisting of all Buddhas is Vajrasattva who has supreme bliss.
This is the Bhagavat's yoga, supremely fixed, su­ premely eternal,
arising as Manmatha, having the intrinsic nature of being difficult to master.
The Buddhas, Vajradhara and the rest are said to do what is needful
by way of the yoga of diverse acts for those persons desiring diverse rites.

That one who is all Buddhas becomes all the states of existence of the stationary and the moving.
He who is joined to all Buddhas has the vow of the net of dakinis;
and by this yoga of illusion, he is supreme pervading all things.
Because he is trained by all Buddhas, he is a siddha, supreme oal of all sentient beings.
By transformations of his own form, he accomplishes himself as all illusory women.
This mudra of dive se illusion is called in the ver- nacular "dakin'i."
This means "going in the expanse of the sky;" the realm (characterized by such language) is discursively thought.
Because it is the accomplishment of movin throughout
the sky, it is popularly known as d.akini •
The vi vamudra on all sides is called the viSvasam.vara
on all sines.

Vajra stands for Vajradhara, Padma stands for Padma- dhara,
and M i stands for Ma idhara; the Families belong to these.
Then (the Bhagavat) became equipoised in the samadhi called "Station which is Free from Dust and Overpowering All Tathagatas" and declared this to be bodhicittas

It is neither void nor not void; its middle cannot be perceptually reached,
United with the perfection of insi ht, the means is among those who have compassion.
Then, the means with oodly compassion is clarified by

the perfection of insi ht.

In dharmas lackin discursive thought there is neither contemplation nor the object of contemplation.
With (continual) conviction dus to nondiscursive thought, he should think discursively every dis­ cursive thought.
Among dharmas of nondiscursive thought, he reflects

upon the aim of sentient beings,

The dharma consisting of Tathagatas has neither a dhar­

mg possessor nor an underlying nature.

Then this speech of dharmas arises with a likeness to an echo,
Then, with expansions of virtue by way of contempla­ tion arising from the Mahayana, all the Tathagatas of the three times extolled him with this king of praise:
Homage to the Lord of yoga who liberates sentient beingsl Homage to him who causes all selves to be onel Homage to him who dispels the delusion of the ocean of samsaral Homage to him who reveals the one Knowledge of all realityl

Then, having completed this worship, bowin , they spoke thuss
0 Bhagavat, please explain the heart, the body uni­ fying all dharmas.
The Bhagavat declareda

Whatever the organ and whatever i.ts path, it is the intrinsic nature of that.
By the yo a after equipoise, he is continually equi­ poised.
For that reason, the perception which is the self of all dwells in the skandha.
Some ordinary beings deluded in mind are incapable of understanding this.
Mind and what arises from the mind is the foi•m of the mind; perception has the true form of the knowable.
The efficient cause for the results is the contempla­ tion of a (certain) dharma among dharmas.
Therefore, the many dharmas are like a flowing river or an ocean.
Liberation is singular, and the many do not reach it perceptually.
That which can be gained from the mouth of the

is located within.

H re ends the third chapter giving a brief account of reality.

(Chapter 4)

tthe Bhagavat announced:)

Now I will proclaim a single exegesis for all tantras. Due to his similarity to all sentient bein s, (the
Lord) had the far extended form of the three (Bodies).
He is located among these: Brahm (and ethel Suras) along with (the Asuras).
Precisely that is the perfection of insight having a form of conventional imagery.
Precisely that which transcends objects of sense is located in the heart of all sentient beings.
Therefore, whatever thin s are extensive, precisely

that is Buddhahood concisely.

That Buddhahood which was attained in an incalculable number of tens of millions of aeons,
you will attain in this birth along with sublime bliss.

Either as Vajradhara or as a Cakravartin, (you will attain) either the eight great siddhis or anything else you desire.
Delusion, lust, hatred, pride and jealousy are the five defilements.

By those very members with which sentient beings are bound, (sentient beings) conquer.
By these (defilements), sentient beings are bound to births, circling in samsara by way of the six destinies.
They are confused by the defilements and commit much sin.
For the destruction and burning of these, however,

there is the principle manifested by the Buddha who has seen the suffering and lack of a among
those who have fallen into the ocean of sam.sara •
By virtue of the manifestation of means and insight, the defilements become a matter of belief.
Indeed, this is the result characterized as shining immaculate in the three realms.
Whatever creatures are installed with whatever princi-

ple, they have the form of that.

By reason of (the principle) which burns the fuel, one becomes a divine danc r who delights the Jinas.
That which is characterized as pure consists of the (five) skandhas, the sense bases, and the realms.
Just as an enemy being capable destroys someone, so

that one being capable destroys the enemy. Delusion (is destroyed) by the purity of delusion;
hatred (is destroyed) by the purity of hatred.

Lust (is destroyed) by the purity of lust;

great pride (is destroyed) by the purity of pride. Jealousy is destroyed by the purity of jealousy;
the purity of each wields a thunderbolt.

Given the pure form of these, the five defilements are equal.
These are the five Families, the five knowledges, and

the five Buddhas.

From these are born the sentient bein s of the three realms, (including) Vajragarbha and the Wrathful ones.
Precisely this is an inner variety which one gains from

the mouth of the guru.

Whatever santient beings lack gurus, they are the ones who do not acquire mantras and mudras.
Whoever shines brightly and (praises) the Buddha here

in Jambudvipa becomes the (n-ev-a.m") of pleasure in the middle of the pure triangle with the form of "e."

When there is pleasure in the triangular mandala, it is called vajrarali.
It is called the bhaga of the Lady," and also the source of dharmas."
The lotus that occurs in the middle of that has a peri- carp and eight petals.
Therein the vowels and consonants commingle arranged

in eight sections.

One performs the set of ritual acts when the (deities) are the embodied with the form of mantras.
Precisely these fifty letters belong to the Vedas.

While the tantras and mantras have their external forms as treatises,
their true forms are vajragarbhar and there is nothing whatsoever which is different from these.
Those fifty letters are the classesa g, ka, gg, a,
ta, • • s"' a; they are manifested as the same.

They occur on the lotus in the vajrarali, on the petals known as in the eight (cardinal and intermediate) directions.
At the filament in the middle of this, is known to be

the Supreme Female Lord.

Her supreme letter is surrounded by eight sections.

The letter "g" is the best of all letters; chief of its class, it is of great purpose.
All mantras of the embodied ones rightly arise precisely from there.
These are the members (called siddhis)a "sword," "eye ointment," "foot ointment," "little ball," "under-
world," and "yaks• a;"
followed by (the actions) "wandering throughout the three worlds" and "mercurial destiny."
All who have great power, having the desire, (can accom­ plish these fifty) in their own homes.

These classes themselves issue forth from the best vowel occurring in the middle of the eight classes,
Whatever speech of persons is of a larger sort, it is scarcely understood,
It is precisely for that reason that every such sort

arises with the form of mantras.

For all the embodied ones, the mystic sound is called "mantra."
The dharm rali develops from the great place of various


Whatever mantra is not certain, it is established ac­ cording to the magical power of siddhi.
Verily, the supreme Lord of the classes with his mantra does not arise by intrinsic nature.
He is called supreme who has the characteristics "arising from saf!lpu a."
The syllable "e" is h.T.own as earth, karmamudra, and


Possessing great compassion and great means, she ranges everywhere in her true form,
She is located in the nirmana-cakra, at the navel, in a lotus of sixty-four {petals).
The syllable "va " is known as water, dharma-mudra, and


She has the characteristic nature of love and fervent aspiration and is the female deity raised in the Vajra Family.
She is located in the dharma-cakra, at the heart, in a good lotus of eight petals.
The syllable is called fire, maha-mudra, and

Pat} ara.

By the union of delight and power, she is the female deity raised in the Padma Family.
She is ocated in the sam.bhoga-cakra, at the throat,
in a lotus of sixteen petals.

The syllable "E" has the nature of wind and is the destroyer of all defilements.
It is the mahasam ya-mudra and the female deity raised in the Karma Family.
By the union of equanimity and knowledge, she is Tara

who carries one across the cyclical flow.

She is located in the cakra at the head called maha­ sukha, in a lotus of thirty-two petals.
The syllable "_g" is called insight, and the syllable

"V8.J!111 possesses the means.

Adorned with the syllable certainly beautiful.

"-v-am.11 , the syllable "e is
It is equipoised below and above by the (joint) intrinsic

nature of means and insight.

The term "eva " signifies that (the Lord) has pro­ claimed the ultimate (meaning) as the collection of tantras,
(or that he will proclaim it).

As for the phrase "maya srutam," great craving is always located there. The syllable "6ru" is revealed to be the "leaking," while the syllable "tam" is said to be the Lord of Great Bliss. The phrase "maya f3'rutam" sig­ nifies hearing with ear cognition but without comprehen- sion.
(Vajra arbha asked:)

0 Bhagavat, what is at fault in the lack of compre­ hension?
The Bhagavat explained:

There is no difference between the compiler of the scriptures and the explainer of them.
Moreover, when the candidate is transfixed, that very comprehension is both the compiler and the ex­ plainer.
I am the explainer; I am the Dharma.

I am the hearer with the congregation as myself. (Vajragarbha asked&)
How can this be realized? (The Bhagavat replied:)
Whe ever the Lord of Great Bliss dances, he dances

according to the phrase "the taste of the one and the many."

Whoever says "maya•••ekasmin" is called a Son of the Fam­ ily by the Bhagavat at "one" time, in a moment. "Heard" means being transported; and it has that (aspiration) thought which goes with the saying, "He has obtained the samadhi of the inconceivable (liberation)."
" mave" means a measure of time; there are said to

be three kinds of time:

good time, bad time, and inconceivable time.

When (wind) comes into the nostril like the flow of milk, that is called good time.
When it goes out with the nature of fire, that is called bad time for one of these two (winds).
When time is one without companions, the inconceivabl

kind occurs.

Neither cravin nor aversion, or something in between is perceptually reached.
There, craving has the character of capability; aversion

is called cessation.

When a middle distinct from these two does not come to mind, then these three are not reached perceptually.
The mixture of craving and aversion is limpid.

Likewise, out of craving and aversion comes the one instant of single taste which is the single taste of all entities.
(The time of contemplating is said to be "one.")

"Bhagavat" is explained as follows: He possesses lordliness, a beautiful form, fame, glory, knowledge, and exertion toward the goal. These are called the "bhagas" of six kinds. Since he possesses them, he is called
"Bhagavat." Or, he is called "Bhagavat" because he is the one who has "broken" those discordant dharmas.
With regard to the "body, speech, and mind of all Tathagatas," he dwells in what can be called their heart, Vajra, Lady, or bhaga.
(Vajragarbha askeda)

0 Bhagavat, what is meant by the phrase "was dwelling?" The Bhagavat declared•
It means, "He has seized the minds of the candidates

with the diverse means of the single taste of all Tathagatas."

When the Bhagavat was dwelling in the source of dharmas which has the intrinsic nature of vajra and is the heart
of the body, speech, and mind of all the Tathagatas, at that time (it is said) "by me it was heard."
(What was heard) is as follows: Insight destroys

the (primary) defileme ts and the secondary defilements, For that reason, insight is called . Every Tathagata dwells in that bhaga accompanied by his Lady.
(Vajragarbha exclaimed&)

Oh, the Incomparable Knowledge which is the Omniscience of all the Budd•hasl

(The Bhagavat declareda)

Whatever is the blissful goal of a Tathagata, it is accomplished in the company of his consort.
The Sons of thE! Family whose words are "eva maya srutam" stand permanently at the head of the Dharma. They are the pure ones generating the siddhis of right
Buddhahood. What has not been explained by Vajrap ti, is

now explained (to you) by me. By means of these words, sentient beings have accomplished the characteristic (of going to a destiny) very far beyond sa sara. Realizing
that suddenly and directly, with a relaxed mind, he reaches Buddhahood or the state of Vajrasattva in this life. Who- ever has not reached an inconceivable base, such persons
are Sugatas and Buddhas,

Whoever is accompanied by a service of speaking rightly, while (thinking of) the act of characterizing and the thing to be characterized, falls into Avici hell. Hav­ ing a defiled mind, he loses the truth.
The glorious sadhaka (who goes beyond out of) fear and dread of a defiled existence, contemplates a pure mind, Just as by way of the union of means and insight a charac­ teristic consisting of pure introspection is well reached,
(so) he who has been transformed by insight and means shines in the three worlds like the sky.

His character is scarcely capable of being char- acterized, He is like a wish-gem, repudiating the causes and conditions for going everywhere, yet accomplishing the goal within the worlds, This is the best of all siddhis.
Here ends the first part of the King of Investiga­
tions, the Glorious Sam• put•odbhava concerning the secret of the nidana of all tantras.


Works Cited

Bhattacaryya, Benoytosh, ed. Guhyasamaja Tantra. Baro­ da: Oriental Institute, 1931.

• Journal of the Bihar Orissa Research Society,

1 4: 343, cited by Cintaharan Chakravarti, Tantras:

Studies on their Religion and Literature, p. 21.
Calcuttaz Punthi Pustak, 1963.

Chakravarti, Cintaharan. Tantras: Studies on Their Re­ ligion and Literature:-Calcuttas Punthi Pustak, 19 3•

Conze, Fdward, trans. The Perfection of Wisdom in Fi.qht Thousand Lines. Bolinas: Four Seasons, 1973.

Das, Sarat Chandra. A Tibetan-Fnglish Dictionary. Delhiz Motilal Banarsidass; reprinted,, 1970.

Ed.qerton, Franklin. Buddhist Hvbrid Sanskrit Grammar and Dictionary. 2 vols. Delhi: Motilal Banarsi ass, 1970; reprinted., 1953.

Elder, George . "Problems of Lane:uae:e in Buddhist Tantras." History of Relie:ions 15.3 (F bruary, 1976): 231-50.

Eliade, Mircea. Patterns in Comparative Relie:ion. Cleve­ land: World Publishing Company, 1963.

• Yoe:a: Immortalitv and Freedom, Princeton:

P rinceton University, 1969; first published, Paris,


George, Christopher, ed. and trans. The CanQamaharo§ana Tantra, Chanters I-VIII. New Havens American Ori­ ental Society, 1974.

Hoffmann, Helmut. The Relie:ions of Tibet. New York: Macmillan Company, 1961.

Hopkins, Jeffrey, trans. "Tson,sr-ka-pa's Introduction to Tibetan Tantra." Preface by His Holiness Tenzin Gyatso, the Fourteenth Dalai Lama. Forthcomin,2'. (Xerographed,)

Lessin.e:, Ferdinand D. and Wayman, Alex, trans. Mkhas grub r,je's "Fundamentals of the Buddhist Tantras." The Ha.e:ue: Mouton, 1968.

Matsunami, Seiren, comp. A Catalo.e:ue of the Sanskrit Manuscripts in the •rokyo University Librar;v. Tokyos Suzuki Research Foundation, n.d.

Monier-Williams, Monier. A Sanskrit-Fne:lish Dictionary.
Oxford: Clarendon Press, 1899.

Snell rove, D. L., ed. and trans. The Heva.ira Tantra.
2 vols. Londona Oxford University, 1959.

Stony Brook, New York. Institute for Advanced Studies of World Religions. Sanskrit Mss. No. MBB-1971-17.
Sa putodbhavasarvatantranidanarnahakalpara.ia. (Microfiche.)

Institute for Advanced Studies of World Reli­ ions. Tibetan Buddhist Canon, Japanese Photographic dition of the Peking Fdition. Vol. 55s 251-JOO. Suravajra, Rin chen phren ba.

Tibetan Buddhist Canon, Japanese Photo.e:raphic Fdition of the Pekin.e: Fdition. Yail da.e: par sbyor ba. z s bva ba'i rgyud chen po. (Xerograph; source, A. Wayman.)

Tokyo. Tokyo University Library. Sanskrit Mss. Nos.
427-28. Sam u odbhavasarvatantranidanamah kal a­ rEi,ja. (Microfilm.

Tsuda, Shinichi, ed. and trans.
Selected Chapters. Tokyo•

The Samvarodaya Tantras Hokuseido Press, 1974.

Ui, Hakuji et al., ed. A Complete Catalogue of the Ti­ betan Buddhist Canons. Sendai, Japan, 19)4.


lv'ayrnan, Alex. The Buddhist Tantras: Tibetan Esotericisrn. New York:
Light on Indo­
Samuel Weiser, 1973.

=-• Review of The Heva.ira Tantra, ed. and trans. by

D. L. SnellP.rove, in Journal of the American Orien­
tal Society 80.2 (1960): 159-62.

• Yo.e:a of the Guhyasama.iatantra. Delhis Motilal Banarsidass, 1977.


Selected Works Also Consulted

Avalon, Arthur. Tantra of the Great Liberation. New York: Dover Publications, 1913; reprinted., 1972.

Basham, A. L. The Wonder that Was India. New York: Grove Press, Inc., 1954.

Beyer, Stephen. The Cult of Tara: Magic and Ritual in Tibet. Berkeley: University of California Press, 1973.

Bharati, Agehananda. The Tantric Tradition. New York: Doubleday and Company, Inc., 1965.

Bhattacaryya, Benoytosh. The Indian Buddhist Iconogra­ phy. Calcutta: K. L. Mukhopadhyay, 1958.

Blofeld, John. 'rhe Tantric IV ysticism of •ribet. New York:
F. P. Dutton and Company, Inc., 1970.

onze, Edward. Buddhist Thought in India. Ann Arbor: University of Michigan Press, 1962.

Danielou, Alain.
New York:

YoRa: The Method of Re-Integration. University Books, 1949.


Dasgupta, Shashi Bushan. dhism. Berkeleya

An Introduction to Tantric Bud­ Shambhala, 1958.

• Obscure Religious Cults. Calcutta: K. L. hlukhopadhyay, 1946.

Dayal, Har. The Bodhisattva Doctrine in Buddhist Sanskrit Literature. Delhi: h1otilal Banarsidass, 1932.

Fvans-Nentz, w. Y., ed. The Tibetan Book of the Great Liberation. Commentary by c. G. Jung. London: Oxford University Press, 1954.

• Tibetan Yo a and Secret Doctrines. London: Oxford University Press, 1958.

Govinda, Lama Anagarika. Foundations of T•ibetan :vlvsticism.
New York: Samut Weiser, 1960.

Jaschke, H. A. A Tibetan-Fnglish Dictionary. London: Routled e and Kegan Paul, Ltd., 1881; reprinted,, 1968.

• Tibetan Grammar, New York: Frederick Ungar

P-ublishing Co,; renrint ed., 1954.

Jung, c. G. "Psycholo ical Commentary on Kumlalini Yoga." Snrin , 1975: 1-)4; 1976: 1-Jl.

Kazi, Lama Dawasamdup, comp. An English-Tibetan Diction­ arv. Calcutta: Baptist Mission Press, 1919.


La Vallee Poussin, Louis de, trans. L'Abhidharmakosa de Vasubandhu. Paris: Paul Geuthner, 19)1.

Macdonnell, Arthur A. A Sanskrit Grammar. London: Ox­ ford University Press, 1927.

Monier-Williams, Monier. English-Sanskrit Dictionary.
New Delhi: Munshiram Manoharlal; reprinted., 1976.

Nyanaponika Thera. The Heart of Buddhist I iedi tation.
New York: Samuel Weiser, 1962.

Robinson, Richard. The Buddhist Religion. Belmont, Cal.: Dickenson Publishing Company, Inc., 1970.

Sakaki, R., comp. Mah vyutpatti and Index, 2 parts.

Suzuki, Daisetz Teitaro, trans. The Lankavatara Sutra.
Londons Routledge and Kegan Paul, Ltd., 1932.

'l'ucci, Giuseppe. The Theory and Practice of the Mandala.
New Yorks Samuel Weiser, 1961,

• Tibetan Painted Scrolls. 2 vols. Rome: La

L i.breria delle State, 1949.

Waddell, L, Austine. Tibetan Buddhism, New Yorks Dover Publications, 1895; reprinted,, 1972.

Wayman, Alex. "Buddhism," Hifltoria Religionum. Leidenr F, J, Brill, 1971, pp. 372-464.

Wayman, Alex an1 Wayman, Hideko, trans. The Lion's Roar
of Queen Srim la, New Yorkr Columbia University Press,

Whitney, William Dwie:ht. The Roots, Verb-Forms and Pri­ mary Derivatives of the Sanskrit Language. Leipzig: Breitkopf and Hartel, 1885.

• Sanskrit Grammar. C Lmbridge, Mass.: Harvard

U-niversity Press, 1889.