The Kalacakra Tantra on the Sadhana and Mandala

The Kalacakra Tantra on the Sadhana and Mandala: A Review Article

The Kalacakra Tantra: The Chapter onSadhana, Together with the Vimalaprabha Commentary, Translated from Sanskrit, Tibetan, and Mongolian, Introduced and Annotated by Vesna A. Wallace. (Treasury of the Buddhist Sciences Series, Tengyur Translation Initiative). New York, The American Institute of Buddhist Studies, Columbia University Center for Buddhist Studies, and Tibet House US, 2010.

The Kalacakra Tantra was the last Buddhist tantra to appear in India, before the disappearance of Buddhism there, roughly a thousand years ago. This is the third book on Kalacakra by Vesna Wallace. We must be very grateful to her for another helpful contribution to our knowledge of this complex system. Her first one, The Inner Ka ̄lacakratantra: A Buddhist Tantric View of the Individual (New York, 2001), provides an overview of the whole system, drawing on all five chapters of the Kalacakra Tantra. Her next one, The Kālacakratantra: The Chapter on the Individual together with the Vimalaprabha ̄ (New York, 2004), presents a translation of the second chapter of the Kalacakra Tantra along with the indispensable Vimalaprabha ̄ commentary thereon. The Kalacakra Tantra is written entirely in the sragdhara metre, in which the length of every syllable is regulated. When a complex system is presented in a complex metre, we have a text that is hard to understand in the extreme. It would be almost incomprehensible without the full and detailed Vimalaprabha ̄ commentary.

Dr Wallace’s third book on Ka ̄lacakra, the book under review, presents a translation of the fourth chapter of the Ka ̄lacakra Tantra along with the Vimalaprabha ̄ commentary thereon. This chapter and her previous translation of Chapter Two are the only chapters of these texts so far published in English translation. This fourth chapter is on the Ka ̄lacakra meditation practice, or sa ̄dhana. Here, the elaborate Ka ̄lacakra man.d.ala with all its 722 deities (according to the count current among Gelugpas) is described in full detail. This is to be visualised in meditation. So this chapter, whether directly or indirectly, forms the basis of all the Ka ̄lacakra practice texts, or sa ̄dhanas, ever written. The Ka ̄lacakra Tantra is the core text of a system of much importance and influence. The present Dalai Lama has given the public Ka ̄lacakra Initiation more than thirty times, throughout the world. So the Ka ̄lacakra Tantra, and this chapter in particular, is likely to have an unusually large readership. For this reason, a careful review is warranted.

The task of reviewers of translations of Sanskrit texts for academic journals is much like that of proofreaders. They must carefully compare the translation with the original Sanskrit text. They are called upon to pass over in silence the thousands upon thousands of well-translated words and phrases and sentences, and take note only of those that might be improved or refined. It is in this way that our knowledge advances. Merely praising a book does not

JRAS, Series 3, 22, 2 (2012), pp. 439–463 ⃝C The Royal Asiatic Society 2012 doi:10.1017/S1356186312000223

440 David Reigle

advance scholarship. The resulting review, therefore, will necessarily be very one-sided. This is a very difficult text, and there are many problems to be solved in translating it. No one can solve all of them the first time through. Once the hard groundbreaking work of translation has been done, a reviewer can then come in and with comparative ease offer suggestions for improvement. That is what I have done. It is worth trying to understand this text as accurately as we can. My suggestions follow.

p. 11, title: “The Great Exposition on the Location, Protection, and Disclosure of Sins”. stha ̄na-rak.sa ̄-pa ̄pa-de ́sana ̄di-mahodde ́sah.. The word “etc.” (a ̄di) is omitted after “Sins”. (This also occurs on pp. ix, 4, and 25.) In the title, “The Great Exposition on the Location (i.e., the places for practice), Protection, Disclosure of Sins, etc.”, the “etc.” brings in such additional topics as the purification of the mouth (vaktra- ́suddhi), rejoicing in the merits of others (pun.ya ̄numodana ̄), and the definition or characteristics of emptiness ( ́su ̄nyata ̄-lak.san.a).

I will be using hyphens in my citations of Sanskrit in order to show exactly how I understand the word breaks. These are not in the printed texts, and the printed texts are customarily cited without them in order to show exactly what is there. But in the printed texts themselves it has become customary for editors to introduce spaces showing word breaks, where the devana ̄gar ̄ı script allows. These are not in the manuscripts, where everything is written together. So the use of hyphens in Romanised text only takes farther the usually helpful process of showing word divisions that the editors of printed texts have already established.

p. 12, line 18: “Here, in the splendid garden of the southern Malaya [mountains], in the town of Kala ̄pa”, iha ́sr ̄ımati kala ̄pa-gra ̄ma-dak.sin.a-malayodya ̄ne. The word “south” (dak.sin.a) refers to the Malaya garden or park that is located to the south of the town of Kala ̄pa, rather than to the southern Malaya mountains. See Vimalaprabha ̄.t ̄ıka ̄, vol. 1, 1986, p. 26, line 20: kala ̄pa-gra ̄ma-dak.sin.ena malayodyanam, “To the south of the town of Kala ̄pa is Malaya park”.

p. 13, verse 1: “and four [higher initiations]—the vase and secret initiations, and the wisdom and gnosis initiations—” kumbhaguhya ̄bhi.sekah. prajn ̃a ̄jn ̃a ̄na ̄bhi.seko . . . caturthah.. The word “fourth” (caturtha) may be used as the actual name of the fourth initiation, also called the “word” initiation. This should be: “the vase and secret initiations, the wisdom and gnosis initiation, and the fourth [initiation]”.

p. 14, lines 18–20: “The day is the sun, uterine blood, and lotus; . . . the night is the moon, semen, and vajra”. dinam. su ̄ryo rajo vajram. bha ̄va-bhedair ni ́sa ̄ ́sa ́s ̄ı | ́sukram. padmam. . . . The words “lotus” and “vajra” should be reversed: “The day is the sun, uterine blood, and vajra; . . . the night is the moon, semen, and lotus”.

p. 16, line 2 (and following): “astrological houses”, lagna. The lagna is the astrological rising sign or ascendant. It is the sign of the zodiac that appears to be rising on the eastern horizon at any given moment of the day or night. So there are twelve of these in twenty-four hours. These are not the same as the astrological houses. For the astrological houses in Indian astrology, the term bha ̄va is used.

p. 18, footnote 57: “The Tibetan translation reads the word ‘adept’ (sgrup pa po) in the genitive instead of in the instrumental”. After briefly noting that the typo “sgrup” should be corrected to “sgrub” in the Tibetan word for “adept”, Sanskrit sa ̄dhaka, and that a sa ̄dhaka is likely to be only a “practitioner” of the sa ̄dhana and not yet an adept, we get to the main

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 441

point. It would seem that the genitive ending here, “yi” instead of the instrumental ending, “yis” is merely a typographical error found only in the blockprint of Bu-ston’s edition of the Ka ̄lacakra Tantra (The Collected Works of Bu-ston, edited by Lokesh Chandra, Part 1 (ka), 1965, folio side 138 as numbered in this reproduction, line 7).

Now that the collated Kangyur published in China has become available, we can easily check eight editions at once: the sDe-dge, gYung-lo, Li-thang, Pe-cin (or Peking), sNar- thang, Co-ne, Khu-re (or Urga), and Zhol (or Lhasa) editions. The collated Kangyur has the instrumental “yis” here, with no variants reported (vol. 77, p. 190, line 14). Similarly, the collated Tengyur published in China allows us to quickly check the Ka ̄lacakra Tantra as it is found repeated in the Tengyur in the sDe-dge and Co-ne editions. The collated Tengyur, too, has the instrumental “yis” here, with no variants reported (vol. 6, p. 133, line 16). The Jonang edition of the Ka ̄lacakra Tantra annotated by Phyogs-las rnam-rgyal, which has recently become available in a nicely typeset edition in the Jonang Publication Series, also has the instrumental “yis” here (vol. 17, p. 110, line 9). So the genitive “yi” found in the blockprint of Bu-ston’s edition appears to be only a typographical error.

To clinch the case, we are now able to check a reproduction of a manuscript in cursive (dbu med) script of Bu-ston’s edition, published in the 7-volume Paltseg Kalachakra Commentary Series (part of the 17-volume set, Phyag bris gces btus, Beijing, 2007). There we find the instrumental “yis” (vol. 2, folio side 138 as numbered in this reproduction, line 5). Thus, neither the Tibetan translation (if we may speak of “the” Tibetan translation) nor Bu-ston’s annotated edition of this translation has the discrepancy noted; it is a mere copyist’s error in the blockprint of his edition.

Regarding the blockprint of Bu-ston’s edition, which is the Tibetan translation used by Dr Wallace, a difficulty with the references arises. The references given in the footnotes include the folio numbers of the blockprint, which is a very helpful feature. However, these are not the folio numbers of the reproduction that is listed in the bibliography, which was edited by Lokesh Chandra and published in the S ́ata-pit.aka Series, New Delhi, 1965. This puzzled me for a while, as I tried to check the references. The only other reproduction I knew of was one that the Dalai Lama’s Office had done, perhaps in the late 1980s. I had obtained a copy from them in January, 1991. It consists of much of the five Ka ̄lacakra volumes, rearranged. It was done in loose-leaf format, and without publication data, but to each folio side was added a number. Upon checking, I see that this is the reproduction used by Dr Wallace.

To match the folio numbers of the more widely available 1965 reproduction in the S ́ata- pit.aka Series, 474 must be added to the numbers given by Dr Wallace in the footnotes. This is because in the 1965 reproduction, the sa ̄dhana chapter starts on the folio side numbered 475, while in the Dalai Lama’s Office reproduction, the sa ̄dhana chapter starts on the folio side numbered 1. So for folio side 20 as given in the footnotes, one must go to folio side 494 in the 1965 reproduction. The same thing is true of Dr Wallace’s earlier translation of the second or adhya ̄tma chapter. But that chapter starts on the folio side numbered 305 in the Dalai Lama’s Office reproduction, and thus in Dr Wallace’s footnotes, but starts on the folio side numbered 1 in the 1965 reproduction. It must also be noted that both of these Vimalaprabha ̄ chapters are found in volume 2 of the 1965 reproduction, while the bibliography lists only volume 1, S ́ata-pit.aka Series vol. 41.

442 David Reigle

p. 19, footnote 61: In the mantra, one “hra ̄h.” is missing. There should be four of these. This mantra is found in the printed Sanskrit edition, vol. 2, p. 35, line 24.

p. 19, line 5: “pericarp” karn.ika ̄. The translation of karn.ika ̄ as pericarp has been in use since Horace Hayman Wilson’s pioneering Sanskrit-English Dictionary in the early 1800s, and has been adopted in the subsequent standard Sanskrit-English dictionaries of Monier Monier-Williams and Vaman Shivaram Apte. The texts make it clear that the karn. ika ̄ is the central portion of a lotus flower, not including the petals (e.g., Vimalaprabha ̄, 3.45, p. 51, line 15: kamala-tri-bha ̄ga-karn.ika ̄ya ̄m a.s.ta-dala ̄ni varjayitva ̄). But if you look up “pericarp” in English language reference books available today, this is not what you find. This always caused me to wonder, and I kept looking for a more accurate English term for it. Wilson in his 1840 translation of the Vi.sn.u Pura ̄n.a, besides pericarp (2.2.37), had also used “seed-cup” for it (2.2.9). This described it accurately enough, but did not come into use as a translation term. Finally, a few years ago I contacted botanist Steven Miller of the University of Wyoming, who kindly gave me a full explanation.

In brief, the central portion of most flowers consists of their female parts, and the most accurate collective or general term for these is the gynecium (or gynoecium). In the case of the lotus and a small number of other flowers, a botanically correct term for the central portion is the receptacle. Since gynecium is as obscure to most readers as pericarp, I have chosen to adopt “central receptacle” for karn.ika ̄, adding the adjective “central” for clarity.

p. 20, line 2: sfp.s ́sxka. The footnotes 65 and 66 here explain that “The letter f here marks a modification of visarga, called ‘upadhma ̄n ̄ıya’ (‘on breathing’), which is pronounced before the letters pa and pha”; and that “The letter x marks here a modification of visarga, called ‘jihva ̄mu ̄l ̄ıya’ (‘formed at the root of the tongue’) and pronounced before ka and kha”. I see no need to coin new transliterations for these forms of the visarga, f and x, when the standard transliteration for the visarga, h., works perfectly fine. Since in the Sanskrit text these forms of the visarga are always indicated by giving them with a following p or k, there can be no possible confusion as to what is meant. That is, the upadhma ̄n ̄ıya is always listed as h.p, and the jihva ̄mu ̄l ̄ıya is always listed as h.k. So there is no need to transliterate these as fp and xk. This string of letters can simply be transliterated as: sh.p.s ́sh.ka.

p. 27, verse 8, line 4: “and the pavilion”, pan ̃jaram. va ̄. First we note that “and” should be “or” (va ̄). This seemingly trivial thing becomes significant here in instructions for meditation, where one may visualise either this or the ku ̄.ta ̄ga ̄ra, the “multi-storied palace”. The basic and standard meaning of pan ̃jara found throughout the Sanskrit writings is a “cage” and it will retain something of this idea even when used as an architectural term. This is lost in its Tibetan translation, gur, meaning “tent”. In descriptions of the man.d.ala visualisation found in Buddhist tantric texts it is sometimes compounded with bandhana, literally a “binding” and commonly a “prison” (e.g., Hevajra Tantra, 1.3.3: pan ̃jara-bandhana, Tibetan, gur bcing ba; also in the Sa ̄dhanama ̄la ̄). So we get the idea that it keeps inside what is inside, and by extension, that it keeps outside what is outside. Indeed, as shown by the title of the text, S ́ani-vajra-pan ̃jara-kavaca (from the Brahma ̄n.d.a Pura ̄n.a), it may be seen as a kind of “armour” (kavaca), or protection. The Buddhist tantric Abhisamayaman ̃jar ̄ı specifically speaks of the rak.sa ̄pan ̃jara, the “pan ̃jara of protection” (Sarnath, 1993, p. 8, lines 13, 15).

I have never found a clear definition of pan ̃jara in relation to man.d.ala architecture, but it is always described as being on top of the walls. The various sa ̄dhanas found in the Sa ̄dhanama ̄la ̄

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 443

give the same sequence for visualising a man.d.ala. The vajra-ground (vajra-bhu ̄mi) is below; then come the vajra-walls (vajra-pra ̄ka ̄ra), and above these is the vajra-pan ̃jara. So it seems to be a kind of roof, possibly a dome. David Snellgrove, perhaps influenced by its Tibetan translation (gur, “tent”), translated it as “canopy” in his pioneering 1959 translation of the Hevajra Tantra. But a few of the sa ̄dhanas in the Sa ̄dhanama ̄la ̄ give a group of six items in the sequence of visu- alising the man.d.ala, adding three more to the three already stated (Benoytosh Bhattacharyya (ed.), pp. 424, 487, 491). After the pan ̃jara comes the vita ̄na, “canopy”, followed by the ́sara-ja ̄la and then the jva ̄la ̄nala ̄rka. So a canopy of some sort is above or beyond the pan ̃jara.

Prasanna Kumar Acharya’s extensive Dictionary of Hindu Architecture (London, 1934) does not help us here, under pan ̃jara. But under ́sikhara, he gives just what I would imagine the pan ̃jara to be: “a spherical roof rising like an inverted cup (Latin cupa) over a circular, square or multangular [sic] building”. The pan ̃jara must be a roof of some kind, and I take it as a domed roof having the connotation of a protective sphere.

p. 30, line 16: “‘O king’ is an invocation”. ra ̄jann iti sam. bodhanam. I would translate this as: “‘O king’ is a vocative”.

p. 32: lines 2–4: “The five pure colors correspond to the five collections, beginning with ethical discipline and the like”. ́s ̄ıla ̄di-pan ̃cabhih. skandhaih. pan ̃ca-varn.am. vi ́sodhitam. I take the past passive participle vi ́sodhitam, “purified”, as the verb-form that goes with the instrumentals in all of the verses that are quoted here, rather than as an adjective modifying pan ̃ca-varn. am. , the “five colours”. So I understand this as: “The five colours are purified by the five collections, . . .” That is, the five colours are the purification of the five collections. This, of course, is the language used here and throughout this text to say that an element or deity of the man. d. ala symbolises an element of the outer or inner world. In this case, it is saying that the five colours of the man. d. ala symbolise the five collections. The construal of vi ́sodhitam as a verb-form rather than an adjective here is confirmed by the Tibetan translation: tshul khrims la sogs phung po lngas | kha dog lnga ni rnam sbyangs pa (Bu-ston edition, folio side 490, line 4, and Peking and sNar-thang editions; or rnam par dag in the sDe-dge and Co-ne editions; collated Tengyur, vol. 6, p. 762, line 2). There is no word for “correspond to” anywhere in these verses, but was added by the translator as being implied.

p. 32, line 5: “fences” pra ̄ka ̄ra. While pra ̄ka ̄ra can mean “fence” I think it must be understood in its more usual meaning of “wall”. In relation to the Ka ̄lacakra man.d.ala, there is a threefold wall and a fivefold wall. When the visualisation of the man. d. ala is first described, in the previous chapter 3, verse 23 (Vimalaprabha ̄.t ̄ıka ̄, vol. 2, p. 23, lines 13–14), we read: tatora ̄.s.tra-rak.sa ̄rtham. ra ̄.s.tra-s ̄ıma ̄ya ̄m. pan ̃ca-pra ̄ka ̄ram. bha ̄vayet.“Then,forthesakeofprotecting the realm, one should visualise a fivefold wall at the boundary of the realm”. If the Great Wall of China was a fence, it could hardly have protected the realm from the Mongol horsemen.

That “wall” rather than “fence” is intended here is confirmed by the use of the word bhitti for this in verse 20, and by the use of the compound, pra ̄ka ̄rabhitti. The word bhitti means “wall” (not “fence”), and the compound pra ̄ka ̄rabhitti is used in descriptions of Indian temples to specify that a surrounding wall or enclosure wall is meant, as opposed to just a wall, such as of a building or a partition wall. This compound is used in the present chapter, in the Vimalaprabha ̄ on verses 20 and 33, where it is declined in the locative case: pra ̄ka ̄ra-bhittau (Sanskrit edition, p. 163, line 27, p. 164, line 3, and p. 169, line 11). It was

444 David Reigle

translated into Tibetan as a .sa.s.ti or genitive tatpuru.sa compound: ra ba’i rtsig pa la. We can hardly take this as “on the wall of the fence”, so Dr Wallace translates this as “on/at the wall of the enclosure” (pp. 48, 58).

We see “fence” used frequently in translations made from the Tibetan. It would seem that the Sanskrit pra ̄ka ̄ra and its Tibetan translation ra ba overlap at opposite ends of their respective meaning spectrums. Thus, while ra ba can mean “wall”, its more usual meaning seems to be “fence”. In footnote 25 here, I do not understand the sentence, “The Derge edition reads, ‘rab gsum,’ instead of ‘rab bkral gsum’”. All editions read ra ba gsum (collated Tengyur, vol. 6, p. 762, line 3; Bu-ston, folio side 490, line 4).

p. 32, lines 5–8: “The three fences in the man.d.alas of the mind, speech, and body correspond to the three vehicles, to the five spiritual faculties of faith ( ́sraddhendriya) and the like, and to the five powers (bala), faith and so on”. tri-pra ̄ka ̄ra ̄s tri-ya ̄nai ́s ca pan ̃ca- ́sraddhendriya ̄dibhih. | ́sraddha ̄dibhir balaih. pan ̃ca citta-va ̄k-ka ̄ya-man.d.ale. We know that the mind man.d.ala is surrounded by a threefold wall, and the speech and body man.d.alas are each surrounded by a fivefold wall. The word pan ̃ca, “five”, in the second line refers to these latter two. It is a nominative, and cannot modify the instrumental balaih., “powers”. This sentence is saying: “The three walls [are purified] by the three vehicles, and the [two sets of] five [walls] [are purified] by the five spiritual faculties, faith and so on, and by the powers, faith and so on, [respectively,] in the man.d.alas of mind, speech, and body”.

p. 32, line 9: “pavilions”, vedika ̄. On p. 27, pan ̃jara was translated as “pavilion”, while on p. 12, man.d.apa was translated as “pavilion”. But vedika ̄, pan ̃jara, and man.d.apa are not synonyms. These three were at these places in the Ka ̄lacakra Tantra and Vimalaprabha ̄ translated into Tibetan respectively as kha khyer (Peking, sNar-thang eds.) or stegs bu (sDe-dge, Co-ne eds.), gur, and khang bzangs (Peking, sNar-thang eds.) or sgo khyams (sDe-dge, Co-ne eds.). Of these three, the man.d.apa is commonly a kind of open pavilion having columns or pillars, and I did not comment on it at its occurrence on p. 12. There, however (Sanskrit (ed.), p. 149, line 21), its placement at the limit or boundary (avasa ̄ne) of the eastern doorway of the Ka ̄lacakra man.d.ala house shows that it is an attached entrance hall rather than a separate pavilion “near” the eastern gate (as avasa ̄ne was there translated). This is typical of many Indian temples. I have already commented on the pan ̃jara.

Although the vedika ̄ is reported to have once been a hall or pavilion in which the Vedas were read (P. K. Acharya, A Dictionary of Hindu Architecture, pp. 564, 567), it is not a pavilion in Buddhist texts. There it is generally described as a railing, such as going around a stu ̄pa (Acharya, pp. 567–568; A. K. Coomaraswamy, “Indian Architectural Terms”, Journal of the American Oriental Society, XLVIII (1928), p. 273; F. Edgerton, Buddhist Hybrid Sanskrit Dictionary). But in Buddhist tantric texts, the vedika ̄ is not a railing, either. There it is essentially a platform, a narrow platform. This brings us closer to the central meaning of vedika ̄ as an altar. In the Ka ̄lacakra man.d.ala the vedika ̄ is like a walkway or sidewalk that goes along the bottom of a wall, and that happens to provide a place for the deities to sit or stand. I do not know of an English word that means this.

The term vedika ̄ has sometimes been translated as “plinth” since a projecting foundation at the base of a wall is a meaning of plinth. However, I do not see the vedika ̄ as a plinth, since the foundation of a wall is not its function, and other words are used for a plinth in Indian architecture (Acharya, op. cit.: janman, upa ̄na, pa ̄duka; Coomaraswamy, op. cit.: adhi.s.tha ̄na,

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 445

a ̄lambana). Jeffrey Hopkins has used “apron” for it in the 1985 book, The Ka ̄lachakra Tantra: Rite of Initiation, p. 78. This would be in the sense of “a small area adjacent to another larger area or structure” (Concise Oxford Dictionary, tenth edition).

So how do we know that the vedika ̄ is a narrow platform rather than a pavilion? We know this because it is described in the Vimalaprabha ̄ as being half the width of the door (2.156, vol. 1, p. 253, line 25), or twice the width of the walls (3.39, vol. 2, p. 47, line 6). The “measure of the door” (dva ̄ra-ma ̄na) is a basic unit of measurement in the construction of the Ka ̄lacakra man.d.ala.

p. 32, line 11: “jewelled strips of fabric”, ratna-pa.t.tika ̄. The ratna-pa.t.tika ̄ in man.d.ala architecture is not a strip of fabric. In the Ka ̄lacakra man.d.ala the ratna-pa.t.tika ̄ (Tibetan, rin chen snam bu) is, as stated by Edward Henning, a “jewelled frieze running around the length of the wall and overhanging it” (“Man.d.ala literalism”, www.kalacakra.org). This is not to be confused with the devata ̄-pa.t.tika ̄. For this, see below.

p. 32, footnote 32: “The Sanskrit Ka manuscript and Dwivedi’s edition read, ‘krava ́s ̄ır.saka’ instead of ‘krama ́s ̄ır.saka’”. Actually, Dwivedi’s edition reads “kava ́s ̄ır.saka” (p. 157, line 4).

p. 35, lines 2 and 16: “sounds”, svara ̄h.. These thirty-two and sixteen “sounds” are “vowels” as svara must be translated here, and as it is translated shortly hereafter on p. 36, line 17, and p. 37, line 2.

p. 35, line 4: “on the moon disc, or on the discs of the moon, sun, and Ra ̄hu”, candra- su ̄rya-ra ̄hu-man.d.alopari candra-man.d.ale. There is no word for “or” in the Sanskrit or Tibetan here. I would not mention such a small thing, except that a controversy arose over whether Ka ̄lacakra stands on three or four seats here. See: Ornament of Stainless Light: An Exposition of the Ka ̄lacakra Tantra, by Khedrup Norsang Gyatso, translated by Gavin Kilty (Boston, 2004), pp. 326–331, where this line is quoted on p. 329.

p. 35, lines 5–10: “The thirty-two [signs] are a pentad of the first digit of the moon . . . a pentad of the second digit of the moon . . . a pentad of the third digit of the moon”. prathama- kala ̄-pan ̃cakam . . . dvit ̄ıya-kala ̄-pan ̃cakam . . . t.rt ̄ıya-kala ̄-pan ̃cakam. I understand these thirty- two [vowels, rather than signs] as pentads of the first five digits of the moon, the second five digits of the moon, and the third five digits of the moon.

̄
p. 35, line 13: The .l and .r should be long .l and .r ̄.

p. 35, line 15: ah. ”; and note 53: “The Tibetan reads, ‘a ̄h. ’”. As we saw with footnote 57 on p. 18, only the reading found in the blockprint of Bu-ston’s edition is being reported for the Tibetan. The collated Tengyur reports “ah.” here for all four editions: sDe-dge, Pe-cin (Peking), sNar-thang, and Co-ne (vol. 6, p. 763, line 12). Similarly, for the Vimalaprabha ̄ as it is found repeated in three editions of the Kangyur, the collated Kangyur reports “ah. ” here for all three: sDe-dge, Li-thang, and Khu-re (or Urga) (collated Kangyur, vol. 99, p. 476, line 19). Likewise, the printed Jonang edition of the Vimalaprabha ̄ has “ah.” here (vol. 20, p. 17, line 17). The Seventh Dalai Lama’s full-length Ka ̄lacakra sa ̄dhana, found in his Collected Works, volume VIII, as reprinted in Gangtok, 1976, also has “ah. ” here (folio side 329, line 3). The blockprint of Bu-ston’s edition is alone in having “a ̄h.” here (Part 2 (kha), folio side 492, line 1). That it is only a typographical error in the blockprint is proved by the reproduced manuscript of Bu-ston’s edition, where we find “ah.” (vol. 3, folio side 243, line 2).

p. 35, lines 16–17: “These thirty-two signs of a great man are within a degree of the latitude of the moon”, eta ̄ni dva ̄trim. ́san-maha ̄puru.sa-lak.san. a ̄ni candra ̄m. ́se; and from footnote

446 David Reigle

54: “The Tibetan reads, ‘an ̇ga’ (yan lag) instead of ‘am. ́sa’”. That is, “limb” or “body” instead of “part” or “degree”. Here in the printed Sanskrit edition is a footnote giving the variant reading “candra ̄n ̇ge” as occurring in “ga. ca. bho.” The abbreviations “ga.” and “ca.” refer to the two old palm-leaf manuscripts used in preparing this edition, and “bho.” refers to the Tibetan translation in the sDe-dge edition. When we see the combination of these three oldest sources agreeing in a footnote giving a variant reading, it is almost invariably the correct reading. The other readings found in the later paper manuscripts are almost always incorrect, however reasonable they may have appeared to the copyists and editors at the time. A new edition is underway by the editors at the Central Institute of Higher Tibetan Studies, Sarnath, and these later readings will no doubt be corrected.

So I believe that “an ̇ga” is the correct reading here instead of “am. ́sa”. The phrase “candra ̄n ̇gam” occurs in verse 2 of this chapter, where it refers to “the single Ka ̄lacakra”, as translated by Dr Wallace on p. 14, line 6. In the Vimalaprabha ̄ commentary thereon, this phrase is glossed as “eka ̄n ̇gam”, showing that candra, “moon”, is one of the characteristic word-numbers used in Ka ̄lacakra, standing for eka, “one”. This is explained a little farther on by the translator in a footnote (81) to verse 12, on p. 39. She has translated “candra ̄n ̇gam” in the commentary on verse 2 as “a single body” (p. 16, line 1). I think that this is the correct translation of the correct reading, “candra ̄n ̇ge”, here as well: “These thirty-two signs of a great man are on a single body”; i.e., the body of Ka ̄lacakra. Probably a double meaning is intended here, because the corresponding thirty-two vowels sit on the candra ̄n ̇ga, literally the “body of the moon”.

p. 35, line 25 (and footnote 58): .s.sa”. As stated in the footnote, Dwivedi’s edition omits “.s.sa”. However, it has been restored in not quite the right place in the translation. The order of this class or varga should be: ssa, h. ph. pa, .s.sa, ́s ́sa, h. kh. ka.

p. 37, footnote 67: “The Tibetan translation misses the following: . . . ” It is only the blockprint of Bu-ston’s edition of the Tibetan translation that misses this line. This line is found in the Tengyur editions of the Tibetan translation. It is also found in the reproduced manuscript of Bu-ston’s edition, vol. 3, folio side 244, lines 6–7.

p. 37, line 13: Between “four faces” and “twelve eyes” the words “three necks” (tri-gr ̄ıvam) are missing.

p. 37, verse 11: “whose legs are in the a ̄l ̄ıd.ha [pose] and whose very playful feet are on the hearts of Rudra and Anan ̇ga”, rudra ̄nan ̇ga-dvayor h.rt-sulalita-caran.a ̄l ̄ıd.ha-pa ̄dam. I do not think that feet or legs are spoken of twice here, with the synonymous words caran.a and pa ̄da. The word pa ̄da is here a metrically required substitute for pada, the normal word for “pose” or “posture” in these texts. Like for many such words used in Ka ̄lacakra, this meaning is not found in our dictionaries. As the translator pointed out in footnote 70 on this page, “The Tibetan translation reads, ‘whose very playful feet are in the a ̄l ̄ıd.ha pose’”. The Tibetan word for “pose” here is stabs, coming immediately after g.yas brkyang, or a ̄l ̄ıd.ha, where it clearly translates pa ̄da. We see this same metrical lengthening in this phrase elsewhere in the verses of the Ka ̄lacakra Tantra (3.35b, 3.88b, 4.66c, 4.135b). But in the prose commentary, unless quoting the verse, it is always pada; and even twice in the verses, where the placement allowed, it is pada (4.68b, 4.109b). For pada, and for pa ̄da as metrically lengthened, we find the same Tibetan translation, stabs, in this context of describing poses or postures.

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 447

The oldest extant source describing these postures is Bharata’s Na ̄.tya- ́sa ̄stra, which is referred to by name in verse 69 of this chapter of the Ka ̄lacakra Tantra. In the Na ̄.tya- ́sa ̄stra they are called stha ̄na rather than pada (chapter 11, verses 50 ff.). The term stha ̄na was retained for them in the later Sam. g ̄ıta-ratna ̄kara as well (chapter 7, verses 1017 ff.). But like Buddhist tantric texts, some other Hindu texts use pada for them (e.g., Vi.sn.udharmottara Pura ̄n.a, Citrasu ̄tra, 3.39). The Na ̄.tya- ́sa ̄stra describes six of these: vai.sn.ava, samapa ̄da, vai ́sa ̄kha, man.d.ala, a ̄l ̄ıd.ha, and pratya ̄l ̄ıd.ha. Five of these are used in Ka ̄lacakra (4.66–68, 109), omitting vai.sn. ava, while a different sixth one is added, called lalita (4.69).

The term sulalita found here in verse 11, however, is likely to be largely a metrical filler in the difficult middle segment of the verse line, where six short syllables in a row are needed. In describing feet in the a ̄l ̄ıd.ha posture, as it does here, I would take sulalita in its meaning of “very graceful” rather than “very playful”. Besides the fact that Ka ̄lacakra is here trampling on the hearts of Rudra and Anan ̇ga, which does not seem very playful to me, the rasa or sentiments that are associated with this posture are the v ̄ıra or heroic and the raudra or fierce (Na ̄.tya- ́sa ̄stra, 11.68). The a ̄l ̄ıd.ha posture is to be used to portray aggressiveness, shooting arrows at enemies, etc. (11.69).

p. 40, footnote 90, on the word “hatchet”, saying: “Sanskrit texts read here ‘pa ́su’, literally meaning ‘cattle’, ‘an animal’. This use of the word pa ́su for a hatchet indicates that this type of weapon was used mostly for butchering animals”. Actually, they read par ́su (or the variant spelling para ́su), meaning an “axe” or “hatchet”.

p. 40, verse 14: kha.tva ̄n ̇ga with smiling faces”, kha.tva ̄n ̇ga-vikasita-mukham. The kha.tva ̄n ̇ga is an implement that is supposed to have originated as the leg (an ̇ga) of a bed (kha.tva ̄), on which three severed heads are usually mounted. It would seem incongruous for these heads to be smiling. The word vikasita normally means “opened”, like when a flower blooms. It is used in this text in the phrase vikasita-vadana (verses 10, 176, and 193 of the present chapter, and verse 117 of the second chapter), which has duly been translated as “opened mouth”. Here, however, it would seem similarly incongruous for these heads to have opened mouths. For a question like this we would turn to the annotations of Bu-ston.

The translator gives Bu-ston’s annotation to this phrase from the verse as it is repeated in the Vimalaprabha ̄, in footnote 92 on p. 41, as: “Bu ston [22]: ‘Three smiling vajra faces’”. The Tibetan text of the Vimalaprabha ̄ here is “kha.tva ̄m. ga rnam par rgyas pa’i kha”, on which Bu-ston’s note is: “ni rtse mo’i rdo rje kha gyes pa”. Something seems amiss. We do not see “three” here, nor do we see “smiling”. Bu-ston appears to be saying that the vajra at the top has an opened mouth. We know that kha.tva ̄n ̇gas usually have either a vajra or a trident at the top, above the three heads. So Bu-ston’s note would assume that this kha.tva ̄n ̇ga has a vajra at the top, and be saying that the prongs of this vajra are opened or separated at the end. This is what a wrathful vajra has as opposed to a peaceful vajra, whose prongs join at the end. I was quite unsure of this interpretation, because I have not yet come across a description of a wrathful vajra in Sanskrit texts to see if vikasita is used to describe it, and I have no familiarity with how the Tibetan gloss, gyes pa, is used in native Tibetan. This same gloss for rnam par rgyas pa (vikasita) is also given in the annotated Jonang edition by Phyogs-las rnam-rgyal (vol. 20, p. 21, line 12). So I asked Tibetan translator Gavin Kilty about the meaning of this.

Gavin Kilty replied that gyes pa does mean separated here, and is used, for example, to describe how the channels separate out again and again to make 72,000. So, he explained,

448 David Reigle

this gloss is saying that the root meaning of rgyas pa, “expand”, here has the sense of “separate out”, gyes pa. Therefore, it does refer to the prongs of the vajra separating out at the top. He added that Dr Wallace was probably thinking of dgyes pa, “delighted, pleased” for her “smiling”.

p. 41, line 16: “origination” prave ́sa. To this, footnote 94 is added: “Bu ston [22] ‘origination’”. The word prave ́sa was translated into Tibetan here as rab tu zhugs, which Bu-ston glosses by adding: pa ste rdzogs par bskyed. This gloss was translated into English as “origination”. The glosses of Bu-ston are extremely helpful, if not indispensable. The same thing is true of Sa ̄yan.a’s commentaries on the Vedas. But the gloss is not the text. When Horace Hayman Wilson produced the first English translation of the R. gveda, he necessarily drew heavily upon Sa ̄yan.a’s commentary. He was later criticised for giving in it translations of Sa ̄yan.a’s glosses of Vedic words rather than translations of the Vedic words themselves. I do not think that prave ́sa “entry”, can per se be translated as “origination”, but only glossed as referring to that.

p. 44, verse 18: “ambrosia” am.rta. In the Vimalaprabha ̄ commentary here, am.rta is translated as both “nectar” and “ambrosia” (p. 45, lines 1–2). As we know, in Greek and Roman mythology ambrosia is the food of the gods, while nectar is the drink of the gods. So the question is whether am.rta is a solid or a liquid. In one meaning of am.rta, there are two solids and three liquids among the five am.rtas spoken of in Ka ̄lachakra Tantra 2.125: feces, urine, semen, blood, and human flesh. Pills can be made with these (4.169). The five am.rtas are explained differently in two commentaries on the Hevajra Tantra (1.2.20), as curds, milk, ghee, cow urine, and cow dung (Ka ̄n. ha’s Yogaratnama ̄la ̄, Snellgrove (ed.), p. 111, line 19), or as milk, curds, ghee, honey, and sugar (Vajragarbha’s Hevajra-pin.d.artha-.t ̄ıka ̄, or S. a.tsa ̄hasrika ̄-hevajra-.t ̄ıka ̄, Malati J. Shendge (ed.), 2004, p. 54, line 9). Perhaps this led David Snellgrove to translate am.rta as “ambrosia” in his 1959 translation of this text. But there and here, these substances are largely symbolic (see Vimalaprabha ̄ on 4.113, and on 5.127, vol. 3, pp. 69, 72). The more central meaning in Ka ̄lacakra hardly differs from the standard meaning in Hindu mythology, where am.rta is the nectar of immortality, a liquid. In the latter two of the four parts of the Ka ̄lacakra sa ̄dhana, am.rta is a liquid that flows (sravate) in the form of drops (e.g., Vimalaprabha ̄ on 4.110, quoting the mu ̄la-tantra, p. 205, line 9: sravate bindu-ru ̄pen.a am.rtam). Therefore, I think “nectar” is more appropriate than “ambrosia” as the translation of am.rta.

p. 44, verse 18: “In P ̄ıta ̄’s right hands there are, in sequence, a conch, a flute, a damaru [sic], and a jewel”. The order of the last two should be reversed: “a jewel and a d.amarusa-man.i-d.amarukah.. This is confirmed in the commentary. On a typographical note, d.amaru is given throughout the book without the diacritic, damaru.

p. 45, verse 19: “The black and white [deities] on the moon and the red and yellow [deities] on the sun are present in the intermediate directions”. k.r.sn.a ̄ ́svetendu-mu ̄rdhni tv atha vidi ́si gate rakta-p ̄ıte ‘rka-mu ̄rdhni. The meaning of Sanskrit sentences such as this depends on where one places the implied “is/are”. To get the required meaning, explained in the Vimalaprabha ̄ commentary on this and the following verse, we must place these as follows: “The black and white [deities] are on moon [discs], but also [the goddesses] present in the intermediate directions; the red and yellow [deities] are on sun [discs].” Besides the exception made for the goddesses in the intermediate directions, who are to be placed on moon discs,

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 449

an exception is also made for the gods in the cardinal directions. These are to be placed on sun discs, says the Vimalaprabha ̄ on verses 19 and 20.

p. 46, footnote 110: “Bu ston [26]: ‘The palace of gnosis’”. In this review, I am comparing the Sanskrit text, and only occasionally looking at Bu-ston’s Tibetan annotations. I happened to check this one (26 should be 25, corresponding to folio side 499 in the 1965 reproduction). It reads: shes pas gzhal yas khang yang. This refers to a palace, but does not mention gnosis.

p. 47, line 10: “Above is Us.n. is.a [sic], who is dark in color . . . ”; and p. 48, line 23: “who is dark like Aks.obhya” ́sya ̄ma. While ́sya ̄ma does often mean “dark” my impression is that it is normally used in Ka ̄lacakra in its meaning of “green”. The Tibetan translation here, ljang khu, “green”, supports this. On another typographical note, it so happens that the diacritic is missing on the “i” in Us.n. ̄ıs.a here. But missing diacritics on the countless Sanskrit words in this book are remarkably rare.

p. 48, bottom paragraph: “Here, the black and white, or the eastern and northern, male and female deities who stand above in the east and north, must be placed on the disc of the moon”, u ̄rdhvastha ̄ ́s candra . . . ; and p. 49, top: “Likewise, the red and yellow, or the southern and western, male and female deities who stand below, are on the discs of the sun”, adhasta ̄c ca. This leaves out a lot of deities. For the proper meaning, we need a ca, “and”, in the Sanskrit text after u ̄rdhvastha ̄ ́s, thus saying: “and those who stand above”. We have a ca after adhasta ̄c, and this must be included in the translation: “and those who stand below”. That is, those deities in the east, in the north, and above the man.d.ala should be placed on moon discs, while those deities in the south, in the west, and below the man. d. ala should be placed on sun discs.

In 2010, an excellent old palm-leaf manuscript of the Vimalaprabha ̄ was reproduced by Lokesh Chandra in the book, Sanskrit Manuscripts from Tibet (S ́ata-pitaka Series, vol. 629, New Delhi). This is as good as the old palm-leaf manuscript, preserved at the Asiatic Society Calcutta, which was also once used in Tibet, and is designated as ca in the printed Sanskrit edition. Although I have long had a microfilm of the Calcutta manuscript, I will refer to this book since it is available for anyone to check. In using this book, one must note that the folio sides without numbers are placed out of order. Five folio sides are reproduced on each page. So I will cite this book by page number and first through fifth folio side on that page.

Sure enough, on p. 79, second folio side, line 3, we find our needed ca: u ̄rdhvastha ̄ ́s ca candra . . . So the Sanskrit edition must be corrected, p. 164, line 8, and the English translation emended accordingly. While doing so, there is an extra “east” and “north” that should be deleted, p. 48, bottom paragraph. Also, footnote 122 on p. 48, saying that the Tibetan translation reads “below” instead of “above” should be cancelled. I only checked this after finding the Sanskrit ca, but the Tibetan translation in Bu-ston’s edition reads: ‘dir shar dang byang dang steng na gnas pa. I understand this as: “Here, located in the east, the north, and above, …”

p. 49, verse 21: “[A male deity], who has an emblem in the palm of his first right hand, has a mudra ̄ without a lotus”. yac cihnam. yasya savye prathama-kara-tale sa ̄sya mudra ̄bja-h ̄ına ̄. Besides the fact that all the deities hold an emblem (cihna) or implement (a ̄yudha) in their first right hand, I do not think the word mudra ̄ is here referring to a female partner. So there would also be no need to assume only a male deity as the subject. I think that mudra ̄, “seal” or “stamp”, is here being used to mean the primary cihna, “emblem” or “sign” of a deity. When

450 David Reigle

the deities are generated in the sa ̄dhana, from a seed-syllable and then an emblem, we find that the emblem they are generated from is usually the implement that is held in their first right hand. So I understand this as: “The emblem that is in the palm of the first right hand of a particular [deity] is the seal (mudra ̄) of that [deity], except for a lotus”. The lotus is excluded.

p. 52, verse 25: “With the exception of eight goddesses, Dhu ̄ma ̄ and the others, sometimes in the lotus of the Lord of Jinas there is a splendid wheel in the centre, which consists of twenty-five [deities], O king”. a.s.tau dhu ̄ma ̄di-dev ̄ır jinapati-kamale varjayitva ̄ kada ̄cit | ́sr ̄ı-cakram. garbha-madhye bhavati narapate pan ̃cavim. ́sa ̄tmakam. ca. The ́sr ̄ı-cakra, “splendid wheel” or “glorious circle”, is glossed here in the Vimalaprabha ̄ as the citta-man. d. ala or “mind man. d. ala”, the centremost of the three individual man. d. alas that together comprise the Ka ̄lacakra man. d. ala. It is always there. I understand this sentence as: “And when leaving aside the eight goddesses, Dhu ̄ma ̄ and the others on the lotus of the Lord of Jinas, the glorious circle inside the inner chamber comes to consist of twenty-five [deities], O king”.

The Vimalaprabha ̄ commentary adds, as translated on p. 52, bottom lines: “Thereafter, although the Divine Lord has joined in, it becomes like in the glorious [Guhya]sama ̄ja”. upavi.s.to ‘pi tada ̄ bhagava ̄n bhavati ́sr ̄ı-sama ̄javat. I think we must take the api here in its meaning of “also” rather than “although”. I understand this as: “Then, the Bhagava ̄n also being settled in [or included], it becomes like in the glorious [Guhya]sama ̄ja.” This apparently has reference to what must have then been a standard Guhyasama ̄ja man.d.ala consisting of twenty-five deities, although the thirty-two deity Guhyasama ̄ja man.d.ala became standard later in Tibet.

We may deduce from the annotations of Phyogs-las rnam-rgyal (vol. 20, p. 29) that the Ka ̄lacakra mind man. d. ala is here being said to consist of twenty-five lotuses. On each of these is a pair of deities, male and female. We know that there are eight pairs of tatha ̄gatas, twelve pairs of bodhisattvas, and four pairs of krodha-ra ̄jas, or wrathful protectors. Then, making twenty-five, is the central eight-petalled lotus on which stand Bhagava ̄n Ka ̄lacakra and his consort, surrounded by the eight goddesses, Dhu ̄ma ̄ and the others. These are each on a petal of the central lotus, and thus are left out of the count.

Regarding this, the Vimalaprabha ̄ commentary explains, as translated on p. 53, top two lines: “There is no mistake here because of its being without a lineage”. atra do.so na ̄sti niranvayatva ̄t. It is not that there is no “mistake” here, but rather that there is no “fault” (do.sa) here in teaching this. The reason given for this brings in an important term in Ka ̄lacakra, niranvaya, a term that was subject to various interpretations. This is however too large a topic to introduce here; suffice to say that Bu-ston’s edition and the Peking and sNar-thang editions of the Tengyur translate niranvaya as rigs med pa, “without lineage” here, or following Ronald M. Davidson’s Man ̃ju ́sr ̄ı-na ̄ma-sam. g ̄ıti translation, “without causal connection”, while Phyogs-las rnam-rgyal’s Jonang edition and the sDe-dge and Co-ne editions of the Tengyur here translate niranvaya as ris med pa, “without partiality” or “without following one line”.

The verse concludes, as translated on p. 52: “For the sake of initiation, the man.d.ala lacks the outer circle due to the power of the families”. seka ̄rtham. man.d.alam. vai bhavati kula-va ́sa ̄d ba ̄hya-cakra-prah ̄ın. am. It may be clearer to directly account for the bhavati by translating bhavati prah ̄ın. am as something like “becomes devoid of” rather than just “lacks”. Only the deities of the mind man.d.ala are needed for granting initiation, not the deities of the speech and body man. d. alas. It is not that the latter are not there, only that they are not needed. So I understand

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 451

this sentence, still awkward in expression because of the original being metrical, as: “For the sake of initiation, the man.d.ala becomes devoid of the outer [two] circles [i.e., man.d.alas] due to the power of the families”.

p. 53, lines 21–22: “in accordance with the natures of the body, speech, and mind” citta-va ̄k-ka ̄ya-svabha ̄vena. As we often see in the prose Vimalaprabha ̄, the word order given is intentional. Here it is intended to correspond to that of the three circles in the first part of the sentence. So the translation should follow the literal order: “in accordance with the natures of the mind, speech, and body”.

p. 54, verse 26: “On the eight petals of the lotuses outside [the mind man.d.ala] there are yogin ̄ıs, Carcika ̄ and the others, . . . and are accompanied by their eight respective goddesses of the eight directions”. ba ̄hye ca ̄.s.ta ̄.s.takena ̄.s.tasu kamala-dale.sv a.s.ta-dig-devat ̄ıbhir | yoginya ́s carcika ̄dya ̄h.. The first clause must be moved to the end of the sentence in order to get the required meaning. It is not Carcika ̄ and the other seven who are on the eight petals of the lotuses. Rather, they are each at the centre of a lotus and are each surrounded by a group of eight goddesses on the eight petals of each of the eight lotuses. So this sentence should be rearranged and modified to read: “Outside [the mind man.d.ala] there are yogin ̄ıs, Carcika ̄ and the others, . . . and are accompanied by the eight groups of eight goddesses in the eight directions on the eight petals of the lotuses”.

As always, the translation of the Vimalaprabha ̄ commentary will have to be adapted accordingly. Additionally, line 20 on this page includes an emendation by the translator: “Carcika ̄ . . . is on the eastern petal”. As stated in footnote 146 thereon, “Dwivedi’s edition and the Tibetan translation read, ‘lotus’ instead of ‘petal’”. In fact, “lotus” is quoted directly from the verse itself, and is correct. So “lotus” must be restored in place of “petal” following the Sanskrit text and Tibetan translation.

p. 56, lines 4–5: “Kauma ̄r ̄ı has a jewel and a goad” kauma ̄rya ̄ ratnam. pa ̄ ́sah. . This is a mere slip, and should be: “Kauma ̄r ̄ı has a jewel and a noose”.

p. 56, fourth paragraph: “In the petals of the lotuses of Carcika ̄ and the others – where the goddesses of the petals of Carcika ̄ and the others are to be known as turning toward the right, or toward the east and so forth – Bh ̄ıma ̄ is on the first petal, . . . ” carcika ̄di-kamala- dale.supu ̄rva ̄di-dak.sin.a ̄vartenacarcika ̄d ̄ına ̄m. patra-devyoveditavya ̄h.,tatraprathama-patrebh ̄ıma ̄.The phrase, “or toward the east and so forth” is not glossing “turning toward the right” but is giving necessary additional information. As such, it will need to be translated differently. I take pu ̄rva ̄di here in its basic meaning as “beginning in the east” rather than its usual paraphrase as “the east and so forth”. Also, in agreement with the translator’s footnote 151 on this page, we must correct yatra “where”, in the Sanskrit edition (p. 168, line 2) to patra, “petal”. But the “where” is still in her translation, making a subordinate clause for what should be the main verb of the sentence, veditavya “are to be known”. So I understand this sentence as: “On the petals of the lotuses of Carcika ̄ and the others, the goddesses of the petals of Carcika ̄ and the others are to be known, turning toward the right beginning in the east. Of these, on the first petal is Bh ̄ıma ̄”. Then follow seven more names, going clockwise around the lotus in sequential order.

p. 57, line 4: Suparamavijaya ̄ is just Vijaya ̄ here in the Vimalaprabha ̄ commentary. The suparama is probably just an addition in the Ka ̄lacakra Tantra verse to fit the metre, as it is in the middle segment of the line that requires a string of short syllables.

452 David Reigle

p. 58, line 8: Suparamaturita ̄ is just Turita ̄ here in the Vimalaprabha ̄ commentary.
p. 58, line 11: S ́a ́sadharadhavana ̄ should be S ́a ́sadharavadana ̄.
p. 58, bottom two lines: “in the right and other sections of the eastern gate”, pu ̄rva-

dva ̄rasya savya-bha ̄ga ̄dau. The text is not speaking of sections of a gate as the locations for the twelve lotuses of the body man.d.ala, but rather the sections of the man.d.ala going all the way around. Like before, as noted regarding a similar phrase on p. 56, we must here take a ̄di as “beginning with” rather than “and other”. I understand savya-bha ̄ga ̄dau as “in the sections beginning on the right (of the eastern gate)” rather than “in the right and other sections (of the eastern gate)”. Again, this has much relevance when enumerating deities in sequential order (yatha ̄-sam. khyam) going around the man. d. ala, as follows here in the text.

p. 60, verse 36 (and Vimalaprabha ̄, p. 62, line 2): “The ten lunar days of caitra . . . ” vasu-kara-tithayah.. When word numbers are used together, like here, they are to be read backwards, not added together. Thus, vasu, the (eight) Vasus, or eight, and kara, “hand”, or two, are to be read as twenty-eight, not as ten. So “the twenty-eight lunar days of caitra” are spoken of here. Then in this verse, “the two pu ̄rn.a ̄s” are added to the twenty-eight lunar days, making the thirty days of the month. Although “Pu ̄rn.a ̄ is a name of the fifth, tenth, and fifteenth lunar days”, as stated in footnote 158, it is here the latter; i.e., the new moon and full moon days.

p. 61, footnote 162: “the day of the new moon . . . It is the twenty-fifth day of the dark half of every lunar month”. Correct “twenty-fifth” to “fifteenth”.

p. 62, lines 5–6: “Their names . . . end with vajras”, vajra ̄ntam. na ̄ma. This should say, “Their names . . . end with vajra ̄”; that is, their names are n ̃a-vajra ̄, n ̃i-vajra ̄, n ̃.r-vajra ̄, etc.

p. 62, lines 10–11: “They are the secondary female deities because they move to the locations of the others”. a ̄sa ̄m. para-stha ̄na-gamana ̄d anuna ̄yika ̄tvam. The deities in the man. d. ala are stationary. The Sanskrit word here is not a verb, but rather is a noun, gamana, “the moving” or “the going”. More literally, the second clause says “because of their moving or going to the location or place of another” something they have already done. This is saying that, in the case of Ma ̄r ̄ıc ̄ı for example, although she is yellow and belongs to the Vairocana family in the west, she has taken her place in the east as the consort of black N ̄ıladan.d.a, of the Amoghasiddhi family.

p. 62, lines 12–13: “Therefore, the families of the east and other directions move toward the location of Vajra ́sr.n ̇khala ̄ and the others”. ata a ̄sa ̄m. pu ̄rva ̄di-kulam. vajra ́s.rn ̇khala ̄d ̄ına ̄m. gamanam abhimukha-stha ̄ne. The families do not move. This obscure sentence pertains to what implements or weapons are held in the hands of the secondary female deities. Literally, it says: “Therefore, the family of these [goddesses] beginning in the east is the going of Vajra ́sr.n ̇khala ̄ and the others to a facing place”. That is, although Vajra ́sr.n ̇khala ̄ is located in the west, she holds the implements characteristic of her own family, that of Amoghasiddhi in the east. Specifically, she holds the same ones held by the corresponding male deity in the east, N ̄ıladan.d.a. Similarly, although Bhr.kut. ̄ı is located in the north, she holds the same implements held by the corresponding male deity of her own jewel family in the south, T. akkira ̄ja. Thus it goes, beginning in the east and proceeding clockwise, with the other goddesses as well.

p. 63, lines 15 and 31: “Karkot.a”. This name is “Karkot.aka” in both the Ka ̄lacakra Tantra verse and the Vimalaprabha ̄ commentary.

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 453

p. 63, lines 17–19: “Atin ̄ıla ̄ has a skull and a bell. Raudra ̄ks. ̄ı has a na ̄ga’s noose and a kha.tva ̄n ̇ga. It is likewise in the case of N ̄ıladan. d. a, T. akkira ̄ja, Maha ̄bala, and Acala, . . . ” It would be helpful to add a note here explaining that N ̄ıladan.d.a holds the same implements in the same hands as does Vajra ́sr.n ̇khala ̄, T. akkira ̄ja holds the same ones as Bhr.kut. ̄ı, Maha ̄bala holds the same ones as Ma ̄r ̄ıc ̄ı, and Acala holds the same ones as Cunda ̄.

p. 64, line 1, etc.: “Va ̄suki and S ́an ̇khapa ̄la are in the southern fire man.d.alas”, etc., dak.sin. e vahni-man. d. ale, etc. The na ̄gas Va ̄suki, etc., are here described as sitting in the various directions on lotus seats on pairs of individual man.d.alas of the elements. The translation would be clearer if given more literally as “in the south, on fire man.d.alas”, etc.

p. 64, line 25: “forty-five million bhu ̄tas”, sa ̄rdha-tri-ko.ti-bhu ̄ta-. This should be “thirty- five million bhu ̄tas”.

pp. 64–65, verse 40: “A red preta, [and seven more creatures] are the seats of Ca ̄mun. d. a ̄ and the others, respectively, in the cardinal and intermediate directions of the lotus” rakta-pretam. . . . ca ̄mun. d. a ̄deh. kramen. a prabhavati kamala ̄ny a ̄sanam. dig-vidik.su. As noted regarding verse 26 on p. 54, the eight main goddesses of the speech man. d. ala are each on their own lotus; they are not on petals of a central lotus. This latter idea seems to have influenced the translation here. What is being said may be seen from the Vimalaprabha ̄’s statement introducing this verse: ida ̄n ̄ım. ca ̄mun.d.a ̄d ̄ına ̄m. kamala ̄sana ̄ny ucyante, “Now, the lotus seats of Ca ̄mun.d.a ̄ and the others are stated” (rather than “the seats of Ca ̄mun. d. a ̄ and the others in the lotus”). These are the eight individual lotus seats (kamala-a ̄sana, Tibetan, padma’i gdan) of Ca ̄mun.d.a ̄ (another name of Carcika ̄) and the other seven main goddesses of the speech man.d.ala, namely, the eight creatures listed here, a red preta, etc.

This fact makes it necessary to modify the translation of the sentence in the Vimalaprabha ̄ commentary given in lines 8–9 of p. 65 as: “The goddesses, Ca ̄mun.d.a ̄ and the others, are on the eight petals” a.s.ta-dale.su ca ̄mun.d.a ̄di-devyah.. As the Tibetan translation shows, ‘dab ma brgyad la tsa mun. d. i la sogs pa’i lha mo rnams so (Bu-ston (ed.), folio side 513, line 5), this should be understood as: “The goddesses of Ca ̄mun.d.a ̄ and the others are on the eight petals”. The eight goddesses of Ca ̄mun.d.a ̄ that are on the eight petals of her lotus have been named in verse 29 of this chapter of the Ka ̄lacakra Tantra, and earlier in the Vimalaprabha ̄ on verse 63 of chapter 3: Bh ̄ıma ̄, Ugra ̄, etc. Likewise, the eight goddesses on the eight petals of each of the lotuses of the other seven main goddesses of the speech man. d. ala have been named here in verses 30–33, and earlier in the Vimalaprabha ̄ on 3.63–64.

Similarly, the translation of another sentence here on p. 65 should be modified, the sentence in lines 12–13: “These are, in sequence, the seats within the cardinal and intermediate directions of the lotus” iti kramen.a ̄sanam. kamalasya dig-vidik.su. The phrase “of the lotus” should go with “the seats”. Thus: “These are, in sequence, the lotus seats in the cardinal and intermediate directions”.

Then, the phrase in line 15 bringing in the deities of the body man.d.ala, “a red preta is a seat in the lotus of Nairr.tya” would be clearer as, “a red preta is the lotus seat of Nairr.tya” (nair.rtya-kamala ̄sanam. rakta-pretam). We cannot here go into the question of whether such a deity is located directly on a lotus, which is mounted on one of these creatures, or whether such a deity is mounted directly on one of these creatures, which stands on a lotus. Suffice to say that the Gelugpa tradition accepts the former, and depicts them this way in their sand man.d.alas and paintings, while the Jonangpa writer Phyogs-las rnam-gyal

454 David Reigle

accepts the latter, and the revised Jonang translations of the Ka ̄lacakra texts reflect this understanding.

p. 65, footnote 178: “Daitya is another name for Vis.n.u”. Actually, Daitya is another name for Nairr.tya, also called Ra ̄ks.asa. Vis.n.u is a different deity in Ka ̄lacakra.

p. 65, verse 41: “a heron”, krun ̃ca; with footnote 180 saying that the Tibetan and Mongolian translations read, “a crane”. The existing Sanskrit-English dictionaries are notoriously unreliable for specialised words such as this. Monier-Williams gives “a kind of snipe, curlew”. V. S. Apte repeats Monier-Williams’ curlew, and adds to it the much larger heron (“a curlew, heron”). The krun ̃ca or kraun ̃ca, made famous by the story from the beginning of Va ̄lm ̄ıki’s Ra ̄ma ̄yan.a, has at long last been accurately identified thanks to the work of K. N. Dave and Julia Leslie. Julia Leslie showed that the one described in the Ra ̄ma ̄yan. a story is the Indian Sarus Crane, in her article, “A Bird Bereaved: The Identity and Significance of Va ̄lm ̄ıki’s Kraun ̃ca” (Journal of Indian Philosophy, XXVI (1998), pp. 455–487). She also drew on K. N. Dave’s Birds in Sanskrit Literature (Delhi, 1985), which shows that while it can denote other large water-birds such as flamingoes, storks, and herons, “in later literature, kraun ̃ca tends to denote specifically the Common Crane” (Leslie, p. 458). It would seem that the Tibetan translation, and the Mongolian translation made from the Tibetan translation, got it right.

p. 67, verse 43, first sentence: “The eight [goddesses] in the interior, who are in the sky and at the base of the verandah, are to be placed beneath the portals” garbhe ‘.s.tau vedika ̄ya ̄m. gagana-tala-gate toran. a ̄dho niyojyo. This line of the verse is referring to the twelve goddesses of worship or offering goddesses (pu ̄ja ̄-dev ̄ı) spoken of in the previous verse. These are located in the garbha, the “inner chamber” which in relation to the Ka ̄lacakra man.d.ala is a specific term for the mind man.d.ala, rather than a general term for the “interior” (see Vimalaprabha ̄ on verse 36 of chapter 3). Eight of these are “on the vedika ̄” (locative, vedika ̄ya ̄m), a narrow platform running along the bottom of the walls. This term was discussed above, at its occurrence on p. 32, line 9. There in the book it was translated as “pavilion” as it is again in the commentary just preceding the present verse (p. 66, bottom line), and in the second line of the present verse. But here in the first line of this verse, and in the commentary on the second line of this verse, it is translated as “verandah”. Two more of these goddesses are located “in the sky” (gagana), that is, above the man.d.ala, and two more are located “at the base” or “bottom” (tala), that is, below the man.d.ala (not the base “of the verandah”). However, the goddesses above and below the man.d.ala cannot be shown above and below in a two-dimensional representation such as a particle man.d.ala, commonly a sand man.d.ala. Therefore these four are to be represented there as “beneath” (adhah.) the “portals” (toran.a), or “arches” as toran. a was translated on pp. 29–31. So I understand this line as: “In the inner chamber [the mind man.d.ala], eight [goddesses] are on the vedika ̄, and those who are located in the sky and at the bottom [of the man. d. ala] should be placed beneath the toran. as”.

The Vimalaprabha ̄ commentary hereon explains this using contrast, although this is not reflected in the translation on p. 67, lines 16–19: “Certain goddesses who are in the sky and at the base within the sand man. d. ala should be displayed beneath the eastern and western portals. During meditation, the guardians of the directions and the others are in the previously mentioned locations” rajo-man.d.ale gagana-tala-gata ̄ devyo ya ̄h. ka ̄ ́scit ta ̄h. pu ̄rva ̄para-toran.a ̄dho dar ́san ̄ıya ̄h. | bha ̄vana ̄ya ̄m. punar dikpa ̄la ̄dayo yathokta-stha ̄na eva. I understand this as: “In a

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 455

particle man. d. ala, whichever goddesses are located in the sky and at the bottom [of the man. d. ala] should be displayed beneath the eastern and western toran.as. But (punar) in meditation, the guardians of the directions and the others are [to be visualised] just in their places as stated”. The contrast between where these deities should be placed in a particle or sand man. d. ala and where they should be placed in a visualised man.d.ala makes clear what is meant in the first line of the verse. The reference to the guardians of the directions additionally provides a clear example. The guardians of the “above” direction, Us.n. ̄ıs.a and consort Atin ̄ıla ̄, are visualised in the sa ̄dhana above the mind man.d.ala, while in a sand man.d.ala they are represented by an additional lotus at the eastern door of the mind man.d.ala.

A correction is required in the next sentence of the Vimalaprabha ̄ commentary as translated in lines 19–20: “The four, Samantabhadra and the others, are on the right of the gates”. This must say, “on the left of the gates”, dva ̄rasya ̄vasavye. The reason for this statement here in the Vimalaprabha ̄ is that the blue and green colours of the four bodhisattvas, Samantabhadra, Vajrapa ̄n.i, S ́abdavajra ̄, and Dharmadha ̄tuvajra ̄, might place them below and above the man.d.ala. But instead, they are to be placed to the left of the four doors of the mind man.d.ala, respectively. It may not be superfluous to note that in the Ka ̄lacakra man.d.ala left and right are always from the standpoint of the central deity, whose four faces face the four doors: sarvatra va ̄me bhagavata ́s catur-mukha-bhedatah. (Vimalaprabha ̄ on 3.59, p. 61, lines 9–10).

p. 67, verse 43, second sentence: Dha ̄ran. ̄ıs are on the porch”, dha ̄rin.yah. pa.t.tika ̄ya ̄m. This line of the verse is not talking about the dha ̄ran. ̄ıs, if by this is meant mantric formulas as we must assume in the absence of a glossary. The word here is dha ̄rin. ̄ı, Tibetan gzungs ma rather than just gzungs, which latter is the normal translation of dha ̄ran. ̄ı. We do not find dha ̄rin. ̄ı in our dictionaries in the sense used in Ka ̄lacakra, so we must find its meaning in the Vimalaprabha ̄ commentary. There on verse 62 of chapter 3 we read (Sanskrit (ed.), p. 63, line 16): garbha-vedika ̄ya ̄m aneka ̄h. pu ̄ja ̄-devatyo dha ̄rin. yah. samasta ̄ lekhya ̄h. , “On the vedika ̄ of the inner chamber [the mind man.d.ala], the many offering goddesses, the dha ̄rin. ̄ıs, are all to be drawn”. So dha ̄rin. ̄ı is an offering goddess, pu ̄ja ̄-devat ̄ı. This seems to be a more general term than pu ̄ja ̄-dev ̄ı, which is used specifically for the twelve offering goddesses of the mind man.d.ala. We may assume that the dha ̄rin. ̄ıs, meaning “bearing” or “holding”, get their name from bearing or holding offerings.

These goddesses are on the pa.t.tika ̄, which was translated here as “porch”. This is another word that is not found in our dictionaries in the sense used in Ka ̄lacakra. It is glossed as vedika ̄ here in the Vimalaprabha ̄: pa.t.tika ̄ya ̄m. vedika ̄ya ̄m iti. The word vedika ̄ is yet another that is not found in our dictionaries in the sense used in Ka ̄lacakra. But as discussed above, the vedika ̄, and therefore also the pa.t.tika ̄, is a narrow platform that runs along the bottom of a wall. This sense of pa.t.tika ̄ apparently derives from its meaning of a “strip”. As derived from the idea of a “strip” or “band” we also have in the Ka ̄lacakra man.d.ala the decorative ratna-pa.t.tika ̄, or “jewelled frieze”, at the top of a wall, discussed above at its occurrence on p. 32, line 11. This is to be distinguished from the kind of pa.t.tika ̄ spoken of here, called the devata ̄-pa.t.tika ̄ in the Vimalaprabha ̄ introducing verse 3.46, on which deities stand or sit. The Vimalaprabha ̄ on verse 3.43 specifically speaks of the dha ̄rin. ̄ı-pa.t.tika ̄, as we have here, saying: vedika ̄ ́sveta ̄ sa ̄ ca dha ̄rin. ̄ı-pa.t.tika ̄ | rakta ̄ tad-upari ratna-pa.t.tika ̄, “The vedika ̄ is white, and that is the dha ̄rin. ̄ı-pa.t.tika ̄. Above that is the red ratna-pa.t.tika ̄.” So the dha ̄rin. ̄ı-pa.t.tika ̄ is the pa.t.tika ̄ specifically for the offering goddess deities, and this is the vedika ̄. We may deduce from other

456 David Reigle

references that the vedika ̄ goes along the outside of the walls (3.39–41), while the pa.t.tika ̄ for the main deities goes along the inside of the walls (later note: Edward Henning informs me that the pa.t.tika ̄ is separated from the walls by a small space).

The second part of the second sentence of this verse was translated as: “prat ̄ıccha ̄s, who are associated with the family of snakes (phan.i), are in the pavilion”. I have already mentioned that “pavilion” here translates vedika ̄, which will need to be modified. So will the phrase, “who are associated with the family of snakes”, phan.i-kula-sahita ̄h.. This is saying “along with”, rather than “who are associated with”. The family of snakes is the group of ten na ̄gas who are on the vedika ̄ of the body man.d.ala along with the prat ̄ıccha ̄s.

In the third sentence of this verse, to be consistent with the other names here, the name “Paus.t.ika ̄” should be “Paus.t.ike[ccha ̄]”.

p. 68, verse 44: A few words in this verse that are relevant to the meaning of the names of the goddesses have been omitted in the translation. One of these, ka ̄ye, “on the body”, has been accounted for in footnote 189, giving the meanings of the names: “Desire for Scratching the Body”. This is scratching in the sense of scratching an itch. Another, payasi, “in water”, would go with “Desire for Swimming”. A third, ́sayane, “on a bed”, becomes quite relevant in reference to “Desire for Lying”. This is not desire for telling lies, but rather is desire for lying down on a bed.

p. 68, line 15: “Likewise Va ̄dyeccha ̄” etc., “are on the eastern and other verandahs”, pu ̄rva ̄di-vedika ̄ya ̄m. Like similar phrases with a ̄di discussed above, this should be “are on the vedika ̄ beginning in the east”.

p. 68, line 17: Between “Vais.n.av ̄ı” and “Pla ̄vaneccha ̄” a line is missing: an ̇ge maleccha ̄ va ̄ra ̄h ̄ı-janya ̄ | n.rtyeccha ̄ kauma ̄r ̄ı-janya ̄ | a ̄saneccha ̄ raudr ̄ı-janya ̄. “Maleccha ̄, on (or in regard to) the body, is born from Va ̄ra ̄h ̄ı. Nr.tyeccha ̄ is born from Kauma ̄r ̄ı. A ̄saneccha ̄ is born from Raudr ̄ı”. The descriptive word, an ̇ge, “on (or in regard to) the body”, is not part of the name, Maleccha ̄ (desire for dirt, or impurities, or impure bodily secretions). So this name need not be written “[An ̇ga]maleccha ̄” as it is in verse 44 here. Similarly, like these descriptive words in the verse that are declined (ka ̄ye, an ̇ge, payasi, ́sayane), so the descriptive vadana-gata, “found in the mouth”, that is given undeclined in a compound, is not actually part of the name, Kaphotsarjaneccha ̄. It is found in the middle segment of the verse line, where a string of short syllables is required. So again, this name need not be written “Vadanagatakaphotsarjane[ccha ̄]” as it is in verse 44 here.

p. 68, verse 45: “The activities of the eight, Ca ̄mun.d.a ̄ and the others, are iccha ̄s of those born from krodhas on the earth. They are Sam. ta ̄pe[ccha ̄]”, etc. ca ̄mun. d. a ̄dy-a.s.ta-k.rtya ̄ny api ca bhuvi-tale krodhaja ̄na ̄m tatheccha ̄, santa ̄pe, etc. The first part of this line is referring to the last seven iccha ̄s or desire goddesses listed in the previous verse and the eighth given in the Vimalaprabha ̄ commentary thereon. These must be carried down to form the first part of this sentence, which then goes on to bring in the next group of iccha ̄s, or personified desires. So this line says: “[The eight last-named personified desires] are the activities of the eight, Ca ̄mun.d.a ̄ and the others; and so also on the surface of the earth the iccha ̄ [goddesses] of the wrath-born [guardian deities] are Sam. ta ̄pe[ccha ̄]”, etc. The Vimalaprabha ̄ on the previous verse lists which of these eight last-named iccha ̄ goddesses is born from which of the eight main goddesses of the speech man.d.ala, Ca ̄mun.d.a ̄ and the others. On this verse, it lists which of ten of the iccha ̄ goddesses named here is born from which of the ten wrathful

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 457

guardian female consort deities, here called krodhaja, the “wrath-born” (cp. Vimalaprabha ̄, 3.151). It then goes on to list two more iccha ̄ goddesses, who are born from two of the pracan.d.a ̄ goddesses. These are continued in the following verse 46. That is why in verse 46 Ucchis.t.abhakte[ccha ̄] is called the fifth, even though coming in a long list of iccha ̄ goddesses, and being the third one listed in this verse. She is the fifth of those born from the pracan.d.a ̄ goddesses.

p. 70, lines 12–13: “Similarly, in the external pavilion within the external man.d.ala, whatever other activity of sentient beings there is . . . ” evam. ba ̄hya-man. d. ale ba ̄hya-pa.tya ̄m aparam api yat kin ̃cit sattva-k.rtyam. The phrase, aparam api, “also the other” goes with what precedes it rather than what comes after it. So it does not refer to “other” in “whatever other activity”, but rather goes with the preceding ba ̄hya-pa.tya ̄m. The reading ba ̄hya-vedya ̄m, found in the verse that is being glossed here (see footnote 194 on p. 69 of the translation), does not change the meaning, since the outer pa.t.tika ̄ is the vedika ̄. As discussed above, this is the narrow platform going along the bottom of the walls on the outside of the walls. In the mind man.d.ala, the pu ̄ja ̄-dev ̄ıs or offering goddesses stand on it; in the speech man.d.ala, the iccha ̄s or desire goddesses stand on it; and in the body man.d.ala, the prat ̄ıccha ̄s or counter-desire goddesses stand on it. These latter two groups of goddesses have just been described. So this sentence is saying: “In this way, in the outer man.d.ala, on the outer pa.t.tika ̄ [where the prat ̄ıccha ̄s stand] and also the other [the outer pa.t.tika ̄ of the speech man.d.ala where the iccha ̄s stand], whatever activity of sentient beings there is . . . ”

It is worth noting that the indeclinable evam, taken by me as “in this way”, was given in the translation as “similarly”, one of its stock meanings that is not applicable here. What is being said here is not similar to what was said above. The same thing occurs in the verse here (46), where the indeclinable tatas was translated in a stock meaning, “afterward” rather than its applicable meaning, “therefore”. This is frequent with indeclinables throughout the translation.

p. 70, verse 47: “in the environment, in the body, elsewhere, and in expansion and contraction”, ba ̄hye dehe pare ca spharan.a-nidhanate. Here in the first seven-syllable metrical unit we have one of the most characteristic phrases and ideas of the whole Ka ̄lacakra system: in the outer (ba ̄hye), or in the environment, in the body (dehe), or in the inner, and in the other (pare), here translated as “elsewhere”. The “other” refers to the Ka ̄lacakra man.d.ala, as reiterated in the Vimalaprabha ̄ commentary here, and as more famously stated in a verse apparently from the lost mu ̄la Ka ̄lacakra Tantra quoted in the Vimalaprabha ̄ on 3.55 (p. 57, lines 18–19):yatha ̄ba ̄hyetatha ̄deheyatha ̄dehetatha ̄pare|trividham. man.d.alam. jn ̃a ̄tva ̄a ̄ca ̄ryoman.d.alam. likhet, “As in the outer, so in the body; as in the body, so in the other. Having understood the threefold man.d.ala, let the teacher draw the man.d.ala”.

p. 72, footnote 200: “Dwivedi’s edition reads ‘ali kali’ instead of ‘a ̄li ka ̄li’”. The contrast is presumably to Biswanath Banerjee’s Critical Edition of S ́r ̄ı Ka ̄lacakratantra-ra ̄ja (Calcutta, 1985). It is always necessary to consult this edition, because the verses of the tantra as given in Dwivedi’s edition of the Vimalaprabha ̄ were mostly based on a single paper manuscript. Upon checking this, we see that it does read “a ̄li-ka ̄li ”. But this is a silent emendation on the part of Banerjee. He has no note here giving variant readings. Yet one of the manuscripts he used, preserved in the Cambridge University Library, forms the basis of the 1966 edition by Raghu Vira and Lokesh Chandra, and this edition has “api kali” here. The point is that the

458 David Reigle

syllables are short, not long. That the original reading is “ali-kali” is confirmed in the old palm-leaf manuscript from Narthang reproduced by Lokesh Chandra in Sanskrit Manuscripts from Tibet (S ́ata-pit.aka Series, vol. 81. New Delhi, 1971), folio side 139, line 6. This, of course, is because of the metre, which requires short syllables here in the middle segment of the pa ̄da or metrical foot. The meaning remains unchanged: these are the vowels and consonants of the Sanskrit alphabet.

The same thing is seen in the very first verse of the Ka ̄lacakra Tantra, in the middle segment of pa ̄da d, kaliyugasamaye, where it appears to say, “at the time of kali-yuga”, the dark age. But the Vimalaprabha ̄ commentary makes it clear that this is to be understood not as kali- yuga-samaye, but rather as kali-yug-a-samaye, and is talking about the consonants (kali) and the vowels. See Vimalaprabha ̄.t ̄ıka ̄, vol. 1, p. 46. John Newman translated this phrase as, “the A collection that possesses the KA line” (The Outer Wheel of Time, PhD thesis, 1987, p. 319), and well explained this in a later footnote (p. 381).

p. 73, line 13: “the threefold speech has the [ka-.sa] contraction (pratya ̄ha ̄ra)”, va ̄g api trividha ̄ pratya ̄ha ̄ren. a. Why the “[ka-.sa]”, we may wonder. A little farther in the commentary on this verse, p. 74, line 14, we read, “Here, a collection of all the consonants is the syllable k.sa”. atra sarva-vyan ̃jana-samu ̄hah. k.sa-ka ̄rah.. To this, footnote 212 is added: “According to Bu-ston [46], the syllable k.sa is a contraction of all the consonants, which begin with ka and end with .sa”. This is like the well-known pratya ̄ha ̄ra or abbreviation used in Pa ̄n. ini’s system of grammar to represent all the consonants, hal. But for the threefold speech here, we would expect not just the ka-.sa pratya ̄ha ̄ra representing all the consonants. The vowels would also have to be included. We would expect something like the pratya ̄ha ̄ra used in Pa ̄n.ini’s system to represent the whole alphabet, al.

In Pa ̄n.ini’s system, pratya ̄ha ̄ras are made from the fourteen S ́iva-su ̄tras, which are usually given at the beginning of his A.s.ta ̄dhya ̄y ̄ı. Although pratya ̄ha ̄ras from this system are quoted in the Vimalaprabha ̄ (e.g., on 1.5, vol. 1, p. 56, line 3), for pratya ̄ha ̄ras in Ka ̄lacakra, a rule and nine su ̄tras different from the S ́iva-su ̄tras are given in the commentary on verse 8 of chapter 1 (p. 60, lines 27–31). Bu-ston adds a note here identifying the source of this rule for pratya ̄ha ̄ras as the ‘jam dpal gyi bya ̄ ka ra n.a, or Man ̃ju ́sr ̄ı-vya ̄karan.a (Bu-ston, part 1, folio side 441, line 2). Its text, the Man ̃ju ́sr ̄ı- ́sabda-lak.san.a, does not seem to have survived in the original Sanskrit, but is preserved in Tibetan translation in the Tengyur (see: Pieter C. Verhagen, A History of Sanskrit Grammatical Literature in Tibet, vol. 1, Leiden, 1994, pp. 126–129, 199–200, 300–304).

p. 73, lines 18–19: “Thus, the syllable om. is the pran.ava. The heart is called a ‘lotus’” evam. om. karah. pran. avah. | h.rdayam ucyate kamalam iti. Here is a good example of where the punctuation in the printed Sanskrit edition should be corrected on the basis of the Tibetan translation: de ltar om. yig pra n. a ba ni snying por gsungs te padma’o. This shows that there should be no dan. d. a after pran. avah. . In India, one would not normally need to be told that the syllable om. is the pran. ava. This is saying: “Thus, the syllable om. , the pran. ava, is called the heart; i.e., the lotus”. The point is that the om. , consisting of the letters a, u, and ma, is being compared in this verse with a lotus, consisting of the bulbous root and stalk, the petals and filaments, and the central receptacle with its moon and sun seats.

p. 74, lines 10–12: “Thereafter, one should immediately observe their bodies, which have become the nature of the circle of the man.d.alatatas te.sa ̄m. svaka ̄ya ̄n man.d.ala-cakra-svabha ̄v ̄ı- bhu ̄ta ̄n jha.titi pa ́syet. Here in the Vimalaprabha ̄ the meditator has just been instructed to draw all sentient beings into the man.d.ala, initiate them, and transform them into man.d.ala deities.

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 459

The words used for this are dhya ̄ya ̄t, the standard word for meditation in general, so “one should meditate on”, here applied to a visualisation, and bha ̄vayet, the favored word in tantric writings for the visualisation type of meditation, so “one should meditate on” in the sense of “one should visualise”. Now comes the sentence using the word pa ́syet, the standard word for seeing, so “one should see”; but where the clear implication is “one should picture”, or “one should visualise”, not merely see or observe. So I would construe this sentence as: “Thereafter, one should visualise that their own bodies have instantaneously become of the nature of the circle of the man.d.ala”.

p. 74, lines 15–16: “According to this suggestive rule, he who has the first name has consonants and should do all this” tena jn ̃a ̄pakena yasya yat prathamam. na ̄ma tasya vyan ̃janam. tena tat sarvam. kartavyam. We must try to make sense of this. There are always questions of what a pronoun refers to. Here we have several pronouns. The so-called yat-tat correlative of a relative pronoun and a demonstrative pronoun, widely used in Sanskrit, is not used in English. Here it is doubled, making the Sanskrit phrase even more awkward for us in English. Very literally, it says, “According to this suggestive rule, what first name of who/what, its consonant, with that all this/those should be made”. The idea seems to be that all subsequent names should be made with the consonant of the first name. This is what we see in the following paragraph that gives thirty-two seed-syllables of various deities corresponding to the thirty-two marks of a great person. All of these seed-syllables begin with the conjunct consonant “k.s”.

That the demonstrative pronoun tasya should here not be correlated with the relative pronoun yasya, but rather should be construed separately, is supported by the Tibetan translation. The Sanskrit words prathamam. na ̄ma tasya vyan ̃janam, literally “first name, its consonant”, are translated into Tibetan in Bu-ston’s edition as ming gi gsal byed dang po, meaning the “first consonant of the name”. The genitive pronoun tasya has here been merged into the genitive noun ming gi, “of the name”. With tasya out of the way, or construed separately, we are left with the demonstrative pronouns tena tat to correlate with the relative pronouns yasya yat. These make a nice yat-tat correlative. When construed in this way, this sentence gives the idea or sense stated by me above.

p. 74, line 20: K.s.l” should be “K.sha”; line 21: “k.s.l.l” should be “k.sha ̄”. In both cases, this is the conjunct consonant k.s plus the letter h, not merely an alternate transliteration of k plus .s. The mantras as given in the printed Sanskrit edition cannot be fully relied on, because they are often based on inaccurate late paper manuscripts. The readings of the more accurate old palm-leaf manuscripts are sometimes not recorded or are recorded incorrectly in the notes that give variant readings. The old Tibetan translations preserve the seed-syllables in transliteration more accurately, clearly showing the letter h here, where it is considered to be one of the semi-vowels. The mantras in general, however, have many scribal errors in the Tibetan blockprints, and these must be allowed for. When several different editions of the Tibetan texts can be checked, as is now possible, and when these agree with the old palm-leaf Sanskrit manuscripts, the correct form of the mantras can in most cases be established with a high degree of certainty. Mantras are of fundamental importance in the Ka ̄lacakra system, and their accuracy is crucial to it.

p. 75, line 1: “Then, having generated in detail every single male deity with every single seed syllable within the private organ of the goddesses, one should emanate them”, atha vistaratah. pratyekaika-b ̄ıjena dev ̄ı-guhye pratyeka-devata ̄m. ni.spa ̄dya uts.rjet. The deities spoken of

460 David Reigle

here as being generated are not only male but also female, and they are generated in the private organ or secret place of one particular goddess only, Vi ́svama ̄ta ̄. I understand this sentence as: “Now, in detail: Having generated each deity with each individual seed syllable in the secret place of the goddess, one should emanate them”.

p. 76, lines 24–28: “Having perceived the blue-coloured syllable hu ̄m. in the fluid of the moon as shining and causing purity, he transforms the vajra from that [hu ̄m. ] and diffuses it by means of that [syllable]. [Seeing] himself generated, the Divine Lord, a yog ̄ı, has the vajra jewels and the tiara with the Lord of Jinas, and is embraced by the wisdom [being] as before”, tatra candra-drave hu ̄m. -ka ̄ram. n ̄ıla-varn. am. d.r.s.tva ̄ sphurad amala-karam. tena parin. atam. vajram. tena spha ̄ritam iti ni.spannam a ̄tma ̄nam. yog ̄ı bhagava ̄n vajra ̄lan ̇ka ̄ra-yukto jina-pati-muku.tah. prajn ̃aya ̄lin ̇gita ́s ca pu ̄rvavat. At the time the four goddesses sing their song to awaken the Bhagava ̄n Ka ̄lacakra, he is supposed to have previously melted into a drop of moon fluid or bodhicitta. He arises from this drop of moon fluid in three stages. First is the syllable hu ̄m. , which then transforms into a vajra, and this vajra in turn becomes the completed or perfected (ni.spanna) full form of Ka ̄lacakra. This is how all the deities arise, being a general rule in tantra that is applicable everywhere (Vimalaprabha ̄, 4.73, p. 187, lines 24–25): evam. b ̄ıjena cihnotpa ̄dah. , cihnena devatotpa ̄dah. sarvatra ̄vagantavyo yogineti tantra-niyamah. , “Thus, from a seed [-syllable] arises an emblem; from an emblem arises a deity. This should be understood by the yog ̄ı everywhere. It is the rule in tantra”. So the translation of the clause, “he transforms the vajra from that [hu ̄m. ] and diffuses it by means of that [syllable]”, should be modified.

Because the passive construction is used so often in Sanskrit, and by comparison is used so little in English, some translators routinely change passive constructions into active constructions. However, there are times when this does not work out satisfactorily. Sometimes the passive verb-forms do not imply an active agent, which is supplied by the translator as “he” etc. I understand this clause as being such a case. So rather than “he transforms the vajra from that [hu ̄m. ]” we would have simply, “transformed from that [hu ̄m. ] is a vajra”, tena parin. atam. vajram. . This idea could be expressed in an active construction as: “that [hu ̄m. ] transforms into a vajra”.

The next part of this clause, translated as “and diffuses it by means of that [syllable]” is also a passive phrase in Sanskrit, tena spha ̄ritam. But besides this, there are also other issues here. I understand spha ̄ritam to mean “emanated” or “manifested” in the Ka ̄lacakra writings, where we find nidhana, “destruction”, used in a contrasting pair with the cognate spharan.a, “creation” or “manifestation” or “emanation”. This is not the idea that one gets from “diffuse”, as spha ̄ritam is translated here, or “expand”, as its cognates are translated in the next couple verses. So we would have “emanate” rather than “diffuse” in meaning, and the passive “emanated” rather than the active “diffuses” in form. We must now take the pronoun tena as “from that”, like in the previous phrase, rather than as “by means of that”, and we must take it as referring to the vajra rather than to the syllable. So I understand the phrase tena spha ̄ritam as “emanated from that [vajra]”. For what is emanated from that, we must bring in the rest of the Sanskrit phrase.

The Sanskrit phrase, tena spha ̄ritam iti ni.spannam a ̄tma ̄nam, shows us that what is emanated from the vajra is a ̄tma ̄nam, “himself” and he is now ni.spannam, “generated” as “completed” or “perfected”. Bhagava ̄n Ka ̄lacakra is now generated in his full and finished form, complete with vajra ornaments (vajra ̄lan ̇ka ̄ra), a crown (muku.ta), in embrace with his wisdom consort

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 461

(prajn ̃aya ̄lin ̇gita), etc. So as expected, here in the third stage Bhagava ̄n Ka ̄lacakra is “emanated from that [vajra]”, which in turn was “transformed from that [hu ̄m. ]”. In the lines quoted above, the translation of the words ni.spannam a ̄tma ̄nam was placed at the beginning of a new sentence: “[Seeing] himself generated, . . . ”

Earlier in the preceding sentence quoted above we find the phrase, “shining and causing purity” (sphurad amala-karam), describing “the blue-coloured syllable hu ̄m. ”. In this oft- repeated phrase, amala-karam is consistently translated into Tibetan as dri ma med pa’i ‘od zer, “immaculate rays of light” and taken as the object of sphurad, “radiating” or “shining”. I see no compelling reason to override this and translate it instead as “and causing purity”, so I take this phrase as: “radiating immaculate rays of light”.

The sentences quoted above began with, “Having perceived the blue-coloured syllable hu ̄m. ”. If Bhagava ̄n Ka ̄lacakra is seeing or perceiving the syllable hu ̄m. , how is he at the same time emanating from it? Because, as discussed in my comments regarding p. 74, lines 10–12, words for “seeing” in these writings sometimes have the sense of “picturing” or “visualising”, not just seeing or perceiving. The verbal d.r.s.tva ̄ used here, “having seen” occurs twice in this long Sanskrit sentence, the first time in the first part of it that was not quoted above. In the translation, this long Sanskrit sentence is broken into smaller sentences, as required for English. In the earlier part of it, Bhagava ̄n Ka ̄lacakra “perceives the entirety of the three realms, characterised by desire, form, and formlessness, as similar to an illusion” (p. 76). This first occurrence of d.r.s.tva ̄ was translated into Tibetan as gzigs nas (Bu-ston (ed.), fol. 522, line 6), while the second occurrence of d.r.s.tva ̄ was translated as bltas te (fol. 522, line 6). It would seem that the Indian pandit and Tibetan translator team understood these to be somewhat different in import. In seeing that the three realms are similar to an illusion, Bhagava ̄n Ka ̄lacakra is perceiving this. In seeing the syllable hu ̄m. , he is picturing this.

As discussed in the above six paragraphs, I understand all these lines as: “There in the moon fluid, having seen (or pictured) a blue-coloured syllable hu ̄m. radiating immaculate rays of light, [picturing] a vajra transformed from this [hu ̄m. ], and [picturing] himself emanated from that [vajra], thus generated as perfected, the yog ̄ı, the Bhagava ̄n, has vajra-ornaments, has the Lord of Jinas on his crown, and is embraced by the wisdom [consort], as before”.

The long Sanskrit sentence just discussed, found in the commentary to verse 49, actually pre-glosses the coming verse 51. There in the translation we find some of the same issues. The first two lines of this verse are translated (p. 82) as: “Hearing that song, perceiving the entire three worlds as illusory, and expanding his shining and pure emblem, the vajr ̄ı creates [the man. d. ala]” g ̄ıtam. ́srutva ̄ sa vajr ̄ı tri-bhuvana-sakalam. tv indraja ̄lopamam. vai, d.r.s.tvotpattim. karoti sphurad amala-karam. spha ̄rayitva ̄ sva-cihnam. The word spha ̄rayitva ̄ is translated as “expanding”. As noted above with the cognate word spha ̄rita, there translated as “diffuses”, these words mean “emanate” or “manifest” in Ka ̄lacakra. Keeping in mind the three stages mentioned above, it is not that he has expanded his emblem (cihna), the vajra, but rather has emanated it. As also noted above, the phrase sphurad amala-karam, there translated as “shining and causing purity” and here translated as “shining and pure” is better translated as “radiating immaculate rays of light”. Then, regarding the last phrase here, “the vajr ̄ı creates [the man. d. ala]”: The vajr ̄ı does not here create the man.d.ala, but rather he “arises” utpattim. karoti. It is in the last line of verse 51 that, after arising, he creates the man.d.ala. I understand these two lines as: “Hearing [their] song, and perceiving the entire three worlds as being like an illusion, he, the vajr ̄ı,

462 David Reigle

arises, having emanated his emblem radiating immaculate rays of light”. That is, from the vajra that he emanated, which radiates immaculate rays of light, he arises.

The last line of verse 51 is translated as: “Moreover, O king, the entire generation of the man.d.ala is by means of the wisdom and method [beings]” prajn ̃opa ̄yena ra ̄jan punar api sakalam. man.d.alotsarjanam. ca. Here we carry down the karoti from the phrase in the second line, utpattim. karoti, “arises”, making the corresponding verb phrase here, utsarjanam. karoti, “generates”. It is here that he generates the man.d.ala (man.d.alotsarjanam). The point of this line is not to say how he generates the man.d.ala, but to say that he generates the man.d.ala. This is missed when an implied “is” is placed before the dependent clause “by means of the wisdom and method [beings]” as the verb of the sentence. Also, I see no indication that “wisdom and method” refers to “beings” here. It is glossed here in the Vimalaprabha ̄ as prajn ̃opa ̄ya-sama ̄pattya ̄, “by the attainment of wisdom and method”, which was so translated on p. 82. Then, here at the beginning of the second part of the sa ̄dhana, the indeclinable punar is used in its meaning of “again” rather than “moreover”. The commentary leaves no doubt about this, saying pu ̄rvavat, “like before”, so this last line is saying: “And again, O king, he generates the entire man.d.ala, by wisdom and method”. It is only after first arising, by way of the three stages mentioned above, that the vajr ̄ı can generate the man. d. ala, and he now does so.

The sequence spoken of in this verse is made clear in the song of the four goddesses, immediately preceding it, quoted from the lost mu ̄la Ka ̄lacakra Tantra in the Vimalaprabha ̄ commentary on verse 50. Each of the four individually requests him to “arise” from dissolution or “get up” from this kind of sleep, using the imperative verb, uttha. Each of them also requests him to “desire me” (ma ̄m), using the imperative verb, ka ̄ma. Then together, they request him to “emanate” or “manifest” the triple man.d.ala, using the imperative verb, spha ̄rayasva. Again, this cognate word is translated as “expand”. However, it is not that he is being exhorted here to “expand the man.d.alas of the body, speech, and mind” (p. 81), but rather that he is being exhorted to once again “emanate the man.d.alas of body, speech, and mind”.

We cannot leave the song of the four goddesses without noticing another translation issue. As recognised in the annotations by Bu-ston, the verses of this song make reference to the kinds of “result” or “fruit” (phala); and indeed, Ka ̄lacakra Tantra verse 50 speaks of “sama-sukha-phala-de”. The Maha ̄vyutpatti (2271–2277) lists five of these: ni.syanda-phala, adhipati-phala, puru.saka ̄ra-phala, vipa ̄ka-phala, and visam. yoga-phala, and these are found in the Abhidharmako ́sa (2.56 ff., 4.87 ff.), the Abhidharmasamuccaya (part 2, chap. 4, near beginning); the Maha ̄ya ̄nasu ̄tra ̄lam. ka ̄ra-vya ̄khya ̄ (17.31), the Bodhisattvabhu ̄mi (Wogihara (ed.), p. 102, line 16 ff.; Dutt (ed.), p. 72, line 12 ff.), etc.; so the Indian Buddhist audience of the Ka ̄lacakra Tantra would have been familiar with them. Here in the Vimalaprabha ̄, like in the Hevajra Tantra (2.4.56–58), only four are given, adopting three of these five, ni.syanda, vipa ̄ka, puru.sa(ka ̄ra), and using a different name for a fourth, vaimalya, rather than visam. yoga. So the translations of these terms in the song of the four goddesses must be altered.

In the first verse of the song of the goddesses, ni.syande yogina ̄m. sthita ̄ does not mean “present in the yog ̄ı’s emission” (p. 81), but rather refers to the ni.syanda-phala, the “natural outcome result”. In the third verse, puru.se yogina ̄m. sthita ̄, does not mean “present in the spirit of yog ̄ıs”, but rather refers to the puru.saka ̄ra-phala, the “man-made result”. Although

The Ka ̄lacakra Tantra on the Sa ̄dhana and Man.d.ala 463

Sa ̄m. khya terms and ideas are used in the Ka ̄lacakra writings, puru.sa does not here mean “spirit” as it does in Sa ̄m. khya. The translations of the second and fourth verses, speaking of vipa ̄ke and vaimalye, also referring to “results” (phala), must likewise be altered.

p. 76, lines 28–31: “Furthermore, because of the newborn child’s cry at spiritual awakening, one should emanate a man.d.ala that is of the nature of the body, speech, and mind and that consists of wisdom and method”. punah. prajn ̃opa ̄ya ̄tmakena citta-ka ̄ya-va ̄g- dharmena man. d. alotsarjanam. kurya ̄j ja ̄tasya ba ̄lakasya prabodha ̄krandana ̄d iti. This sentence comes immediately after the long sentence discussed at length above. Like in the translation of the similar last line of verse 51, also discussed above, the import and intent of the Sanskrit is lost in the translation of this sentence. The meaning “furthermore” for the indeclinable punar, like the meaning “moreover” given for it in verse 51, is not applicable here. The indefinite article “a” before man.d.ala should be made the definite article “the”. The sa ̄dhana is being described, and we are at the beginning of the second part of it. Here Bhagava ̄n Ka ̄lacakra has been reawakened by the song of the four goddesses just prior to once again emanating the threefold man.d.ala, like he did in the first part of the sa ̄dhana. In the previous sentence he has arisen from a syllable hu ̄m. , and then a vajra, and is now in full form. Now, says the text, he should again (punar) generate or emanate the man.d.ala (man.d.alotsarjanam. kurya ̄t).

This brings us to the end of the translation of the second section of the fourth chapter of the Ka ̄lacakra Tantra and Vimalaprabha ̄, and this review has gotten too long. It is time to conclude. We have seen that if one does not take full account of the description of the man. d. ala given in the third chapter, significant errors of interpretation occur in the translation of the description of the sa ̄dhana given here in the fourth chapter. Similarly, we have seen that if one does not take full cognizance of what is happening in the sa ̄dhana here in this chapter, significant errors of interpretation again occur in the translation. There is much involved in translating even a single Sanskrit sentence from this complex system, and this gives us a perspective on translating a text that has thousands of difficult sentences. If one wants to complete the translation, it is not always practical to deal with all the problems of translation that each of these poses. This is where reviews come in and can make their contribution to the understanding of these texts.

When the Buddhist texts were translated from Sanskrit into Tibetan around a thousand years ago, this was usually accomplished by a translation team consisting of an Indian pandit who had been taught the text and a native Tibetan translator. We do not have this luxury today. We must do the best we can using the avenues now available to us, and try to support each other in this difficult work in whatever ways possible. I am personally very grateful to Dr Wallace for her tremendous labour in making this translation, thereby opening up access to the core texts of the important Ka ̄lacakra system.

David Reigle

International Kalachakra Network