1

Pelliot 349: a Dunhuang Tibetan text on rDo rje Phur pa

Robert Mayer1
[1] Introduction
In 1978, the late R.A.Stein published a paper under the title A propos des documents
anciens relatifs au phur-bu (kla).2 In this paper, Stein made a wide-ranging analysis of
ancient texts concerning the kla, including two Dunhuang texts (Pelliot 44 and Pelliot 349),
as well as various materials from the later canonical collections of the Kanjur, Tenjur and the
rNying mai rgyud bum (henceforth NGB). Much of what Stein reported in that article
remains valid at the time of writing over two decades later; but inevitably, with the passage of
time, some of his findings have also become more debatable. What remains valid is his
important identification of a strong connection between the kla materials and the
Guhyasamja tradition; and also the importance of the occurrence of the shared identity of
the forms Vajrakla and Amtakualin. I believe Stein's article was the first to remark that
Vajrakla and Amtakualin often and on significant occasions share a merged identity:
while Bischoff's 1956 study of the Mahbala-nma-stra had showed that Vajrakla and
Amtakualin occur as different deities in the same maalas, Bischoff had not yet
encountered any evidence of the merged identity these forms often share, a feature that took
particular significance in the important Guhyasamja literature and its many derivatives.3
What have become more questionable however are some of Stein's minutiae in historical and

An earlier version of this paper was delivered on December 12, 2000 at the University of Hamburg's Institut
fr Kultur und Geschichte Indiens und Tibets, at the kind invitation of Professor David P. Jackson. This text was
one of those studied by my advanced reading group during my time as Visiting Professor at the Humboldt
University of Berlin, 1999-2001. Two students made contributions to this paper: Gudrun Melzer most
significantly by preparing the presentation of items 5.1 and 5.2 in the Appendix, by locating texts within the
Sdhanaml, and by discovering the parallel text in the Piiktasdhanopyikvttiratnval (see under
Appendix 5.2). Melzer also prepared the initial transcription of the Tibetan text, a difficult task that she worked
on with Kerstin Grothmann. My thanks to these two students, whose outstanding keenness also made short work
of the arduous tasks of locating P349 within the largely uncharted microfilm, and printing it out in a readable
form. My many thanks also to Dr. Ralf Kramer of the Bodleian Library, Oxford, who went to a great deal of
trouble to procure for me a clearer view of the text than we had been able to achieve in Berlin. Our thanks also to
Professor Matthew Kapstein for his help on an item of vocabulary. At a later stage, Professor Cristina ScherrerSchaub made a number of extremely useful points on the presentation and the contents of this paper, for which
we are very much indebted. Above all, many thanks to Dr Cathy Cantwell for making her studies in the Sa skya
pa and rNying ma pa Vajraklaya literature available to me, and for many other extremely useful and learned
suggestions made in the course of reading this paper through in its final stages of preparation.
2
See Proceedings of the Csoma de Krs memorial Symposium, ed. L. Ligeti, Budapest, pages 427-444.
3
Vajrakla in this context is generally depicted as a deified stake, rather than a heruka who wields a stake as his
main implement (more on this distinction below). The occurrences of Amtakualin/Vajrmta merged with
Vajrakla in Guhyasamja Chapter 13 (verses 74-75) and also (if less specifically) in Chapter 14 (verses 59ff)
seem to have proven very influential, and are certainly the distant basis of the verses and mantra in the
concluding lines of the text Pelliot 349 analysed here; although Pelliot 349's concluding lines follow the specific
arrangement as found in Ngrjuna's commentarial Pacakrama rather than the root text of the Guhyasamja
itself. Not surprisingly, many important Guhyasamja commentarial texts such as Candrakrti's
Pradpoddyotana and Ngrjuna's Pacakrama reaffirm the identity of the deity described in Guhyasamja
Chapters 13 and 14 as Amtakualin appearing as Vajrakla, as do subcommentarial texts like Munirbhadra's
Pacakramaippai. Boord 2002:26 ff gives very useful citations from the Guhyasamja literature of
Amtakualin/Vajrmta identified as Vajrakla. Perhaps based on such materials, Boord's earlier work (Boord
1993:6 and the whole of Ch. 2) argued at length that Amtakualin or Vajrmta is in fact the main precursor of
Vajrakla. However I find this a considerable oversimplification. For a different and more broadly cultural and
Indological (rather than narrowly Buddhological) view of the precursors of Vajrakla, see Mayer 1991.

2
textual data.4
Of the two Dunhuang kla texts Stein addressed in his paper, the longer and more
significant one had already been studied in some detail by Bischoff and Hartmann (1971),
who produced both a transcription of the Tibetan text, and a full translation. Stein made
considerable use of their work, while adding more of his own insights. The text in question is
the famous P44, a text that describes Padmasambhava bringing the Vajraklaya tradition to
Tibet, and which includes the famous and important Yang le shod narrative that re-occurs in
much later literature, in which Padmasambhava meditates at Pharping in Nepal and subdues
demons there by means of the Vajraklaya tradition brought to him from Nland. Bischoff
and Hartmanns study was of course a pioneering work, and it might well prove fruitful to
retranslate their renderings when time and oppurtunity permits (note that the first chapter has
already been retranslated recently by Matthew Kapstein in his book published in 2000, The
Tibetan Assimilation of Buddhism).
The other Dunhuang text Stein addressed is the shorter Pelliot 349, which has not to
our knowledge so far been transcribed and translated. Although apparently the first person to
look at this text, Stein himself made no effort at all to present a transcription of the Tibetan,
nor did he present any translation. Rather, he limited himself to a very brief summary
4

For examples of what has become questionable, on page 428 Stein stated categorically that the Guhyasamja
was not translated into Tibetan before about 1000 C.E., yet we now know diferently - for a clear resum, see
Toru Tomabechi's paper Selected Tantra Fragments from Tabo Monastery. See also Kenneth Eastman's 1980
more detailed study and stemma codicum of the Guhyasamja made from all the extant Tibetan versions
available to him at the time, to which Tomabechi refers. We can add to Eastman's and Tomabechi's findings that
in fact several witnesses of this famous scripture as found in the NGB (Rig 'dzin vol. Tsa; sDe dge vol. Na;
gTing skyes vol. Tsa), give very specific colophonic information that the main Guhyasamja mlatantra (i.e.
chapters 1-17 without the Uttaratantra or 18th chapter) was first translated by Vimalamitra and sKa ba dpal
rtsegs (paita bi m la dang lo tsa ba ska wa dpal rtsegs kyi bsgyur pa'o//), i.e. around 200 years earlier than
the date Stein gives, and moreover that in these particular editions, the Uttaratantra (i.e. Chapter 18) was
translated later by Buddhaguhya and a certain Brog mi dpal ye shes (rgya gar gyi mkhan po sangs rgyas gsang
ba dang/ /bod kyi lotstsha ba 'brog mi dpal yeshes bsgyur ba'o//). The Rig 'dzin and sDe dge edition colophons
also suggest that the famous Rin chen bzang po translation was a reworking of the earlier translation (slad kyi
mkhan po tsrya shraddha ka ra war ma dang / zhu chen gyi lo tstsha ba dge slong rin chen bzang pos bsgyur
te gtan la phab pa'o//), although it is not absolutely clear if this refers to the whole text or only to the
Uttaratantra. One should note however that not all NGB editions have the so-called NGB version - some merely
reproduce the Kanjur's Rin chen bzang po version. Of course NGB colophons are not always reliable as
historical sources, but additional information comes from the Blue Annals (p.204-5), which also mentions that
there existed translations of the Guhyasamja made earlier than Rin chen bzang po, although here the earlier
translations are attributed to the comparatively late figure of Smti, who along with his near contemporary Rin
chen bzang po traditionally marks the watershed between Old and New translation periods. In addition, of
course, we are now aware of the important testimony of the Dunhuang text Tib 438 and 431, which together
comprise a complete 17-chapter Guhyasamja mlatantra in Tibetan with many marginal notes - which
Kenneth Eastman believes to date from between 800 and 900, and which he has attempted to demonstrate as the
sole source for all subsequent Guhyasamja editions in Tibetan, all of which he says merely revise this
Dunhuang original. This would include Rin chen bzang po's version, and those that followed Rin chen bzang
po's, such as the editions by 'Gos lo ts ba Khug pa lhas btsas (Tomabechi dates him as 11 century) and by Chag
lo ts ba Chos rje dpal (Tomabechi dates him as 1197-1264) and even by Tsong kha pa, who is said to have
studied many Guhyasamja manuscripts comparatively. How exactly the Dunhuang text compares with the
NGB version found in some NGB editions and attributed to Vimalamitra and sKa ba dpal rtsegs remains to be
seen - Eastman did not present a full critical or comparative edition. My thanks to Dr Adelheid Hermann-Pfandt,
who very kindly made Eastmans work available to me.
Another possible error of Stein's: on p.437-8 he explains how in following a textual clue from Ratna gling pas
rNying ma chos byung, he sought key verses in the two K la ya bcu gnyis texts found in volume Ha (29) of the
NGB (presumably, referring to the Rig 'dzin and gTing skyes editions then available to him), but could not find
the verses in either of those texts. He seemingly remained unaware that the main K la ya bcu gnyis is not in the
Rig 'dzin or gTing skyes volumes Ha at all - where only two minor texts of that name occur - but in volume Dza
(19) of the Rig 'dzin, gTing skyes and mTshams brag editions alike.

3
comprising only 9 lines in his own words of the basic gist as he saw it of the text (not all of
which we can now fully agree with). He also made a few important observations such as
pointing out its citation of materials found in Guhyasamja commentarial materials attributed
to Ngrjuna (Stein cites the Tibetan translation of the Piiktasdhana from the Peking
Tenjur vol. 61, no. 2661, p. 269; and Louis de la Valle Poussins Sanskrit edition of the
Pacakrama, found in his tudes et Texts Tantriques, Gand & Louvain, 1896. Pp. 1-2.).
These materials include the important mantra found in the root Guhyasamja's Ch.14 v58,
with which Pelliot 349 comes to its culmination.
Our effort here is therefore in the first instance to transcribe and translate the text of P
349 in full. This has by no means been easy, and we can see why Stein did not attempt either
task. Firstly, regarding the transcription, the text is damaged in parts, and it has proven
impossible to reconstruct the missing portions with any certitude. By and large, we have
resisted the temptation to attempt reconstruction except where the reconstruction is quite
uncontroversial. Secondly, the remaining parts of the text are here and there illegible in the
monochrome microfilm available to us, and might remain so even if good quality colour
images eventually become available from the Bibliothque Nationale in Paris (or even if we
get an opportunity to view the original). Hence we can sometimes only guess at the shape of
the Tibetan letters underneath the vagueness of our available images. Illegible parts of the text
are clearly marked in our transcription.
Our inability to read parts of the text of course in several places contributes to
uncertainty in our translation of connected readable portions. Moreover it is also likely that
the text itself contains some errors, for example, giving in line 8 the seed syllable (bja) Nya
when A was much more likely intended (see comments on line 6 below). Like IOL Tib J 754
(81-82) (Mayer & Cantwell 1994), this text with its untidy layout and the poor quality of its
writing, very much gives the appearance of a note or aide-mmoire for personal use, rather
than a carefully and neatly written scripture for communal use. But even if the whole text had
been readable and error-free, we could not have given entirely confident translations of all of
it. Line 21, for example, is largely readable, but remains a little ambiguous. Wherever our
translation is uncertain, we mark it clearly.
Unfortunately perhaps for students of the rNying ma canonical tradition, it is not only
ancient Dunhuang Tantric materials that are difficult to understand: considerable portions of
existing canonical materials from the NGB continue to defy the best efforts of even the most
learned rNying ma pa lamas, who, with the best will in the world, simply can not understand
them. Of course this is frequently due to transmissional errors, many of which one might hope
will eventually be removed through careful textual criticism. Take for example the famous
and widely cited rDo rje phur bu chos thams cad mya ngan las 'das pa'i rgyud chen po, in
which, if one examines all surviving witnesses, one finds that the level of transmissionally
generated confusion can be quite severe.5 But over and above transmissional obscurities, the
materials are also inherently difficult. This is particularly true of some of the more arcane
Mahyoga materials where the root texts are extremely terse condensed references to
extremely complex and highly technical ritual categories, and which imply access to a parallel
oral or commentarial tradition for filling in the details. But as is so often the case with Tantric
5

This occurs in 26 chapters at Rig dzin Vol. Sa 113v-155v, at gTing skyes Vol. Sa 141r-192r (Kaneko 336),
and at Nubri Vol. Sha 44r-96v. However, it also occurs in 28 chapters at sDe dge Vol. Zha 46r-82r and at
mTshams brag Vol. Chi 229.5-340-3 (Taipei 5102 Vol 61 pp. 164-180). The difference is caused by a number of
folio misplacements, which subsequently became incorporated into some of the transmissions with a number of
quite confusing results. Nevertheless this tantra is among the most frequently cited within commentarial
literature. Cathy Cantwell and I are currently engaged in editing this text at the Oriental Institute, University of
Oxford.

4
texts for which a specific commentarial tradition is no longer available, the remaining root
text becomes de facto partially lost when the oral or commentarial materials are lost. This is
apparently the case for sections of the Vajraklaya materials in the NGB: the famous Phur pa
bcu gnyis, for example, has no commentarial works written specifically for itself, and lamas
trying to read it have to rely on general Vajraklaya commentaries, of which there are of
course many. But the problem with relying on the general commentaries is that when the Phur
pa bcu gnyis presents unusual materials or unique rites, especially if in abbreviated and terse
outline only, the general commentaries can provide only the vaguest of clues as to the exact
meaning. As a result, I found that not even the most learned Phur pa mkhan pos of our time
could understand substantial passages of the Phur pa bcu gnyis. And so on.6 It should
therefore not surprise us to find similar difficulties in interpreting Dunhuang kla materials,
and to some extent that is the case with Pelliot 349. Nevertheless I hope much of our
translation is valid, especially where the text is not destroyed or illegible.
One of Steins great contributions in his brief study of P349 was to point out the close
relation between some Guhyasamja and Vajrakla materials. Indeed, although Stein did not
point this out, Chapter 14 of the Guhysamja root tantra even has important text uttered by
"the Blessed One, the Great Vajrakla" (bhagavn mahvajrakla, bcom ldan 'das rdo rje
phur bu chen po, Ch.14 v70-72), and this chapter in particular contains a great deal of kla
ritual. This relationship has since also been commented on briefly in Mayer 1991 and at
greater length in Boord 1993 and 2002. The full relationship between these two Mahyoga
traditions is an extremely complex issue that we do not wish to go into in much detail in the
present paper, since it requires at the very least a full length monograph treatment and
moreover the textual transmission of the Guhyasamja tradition in Tibet gives signs of being
quite complicated. Nevertheless we add as an appendix Gudrun Melzers demonstration of
some important textual parallels to passages of Pelliot 349 that occur in the Pacakrama
commentarial texts of the Guhyasamja tradition attributed to Ngrjuna, which Stein clearly
remarked but did not actually present to his public. Thus Melzer presents the Sanskrit of de
Valle Poussins Piikramasdhana edition of 1896, with additional reference to Mimakis
facsimile edition of 1994; and from the Tibetan, she adds the sDe dge and Peking Tenjur
versions of the same. In fact versions of the verses in question also occur elsewhere, in texts
Stein did not remark. Melzer found them also in the Piiktasdhanopyikvttiratnval
attributed to Ratnkaranti, although here in a discontinuous form with word by word
commentary interspersed (note the term sdhanopyik in the title, which we will comment
on below; Peking Tenjur 2690, folios 297b-298b). Other parallels - some more exact, some
less exact, and some with particularly interesting continuities to P349 - also occur in more
recent Sa skya and rNying ma pa Phur pa texts of various genres: we cite an example below of
a citation from the ritual text the Sa skya Phur Chen, and Boord 1993:107 gives a translation
(but does not give the Tibetan) of a parallel passage from a Phur pa historical text by the 18th
century Byang gter author 'Phrin las bdud 'joms. I should add, I have also found further
related passages in several NGB Mahyoga scriptural texts, for example, in the gZi ldan 'bar
ba mtshams kyi rgyud, usually classified within the Tantra sde bco brgyad section of
Mahyoga and dedicated to the Ten Wrathful Deities (daakrodha, khro bo bcu).7 In some of
6

Even for those few NGB texts where commentaries do exist, the commentaries can often be later than or
slightly at variance to the actual NGB texts they comment upon - sometimes taking a particular view, sometimes
following a slightly different text, and so to varying degrees departing from or failing to illuminate the intentions
of the authors or redactors of the actual NGB texts as transmitted. Commentaries certainly can not solve all our
problems, but they often help.
7
This text occurs as the 6th text in the Rig 'dzin NGB in vol. Dza (folio 234 -253), in the gTing skyes NGB also
in vol. Dza (pages 516-561), in the mTshams brag NGB in volume Zha (pages 533.3-580.4), and in sDe dge vol
Pha, folios 36-52.

5
these NGB scriptures, we appear to find extremely interesting evidence of Pacakrama verses
entering NGB canonical scriptures (compare Appendices 5.2 and 5.3 below); but of course a
lot more work will have to be done before we can say with any certainty quite what such
definitely existent but extremely complicated textual relationships amount to. Elsewhere in
the NGB, in the Phur pa phrin las skor section of Mahyoga, there appear to be remixes of
the Pacakrama-derived kla verses that more closely follow some of the words found here in
P349 (e.g. in chapter 16 of the Phur pa gsang chen rdo rje 'phreng ba'i rgyud;8we include this
example in Appendix 5.4 below). These parallels with the commentarial Pacakrama verses
bear interesting comparison with the Dunhuang Phur pa text IOLTibJ754,81-82, and also
several instances in the NGB such as the Phur pa bcu gnyis Ch.11, where the parallels might
follow the Guhyasamja root tantra itself or its commentaries.9
By examining the constant textual remixes and permutations so characteristic of much
rNying ma pa Tantric literature, we hope eventually to arrive at a clearer understanding of the
cultural, religious and literary processes through which these texts were produced and
reproduced as commentary, revelation and canonical scripture.
[2] Tibetan text in transliteration:
We present the transliteration according to the conventions established in Tsuguhito
Takeuchi's Old Tibetan Manuscripts from East Turkestan in The Stein Collection of the
British Library, Tokyo 1998:
$
page initial sign (mgo yig, siddha)
I
reversed gi gu
[abc]
our conjectural restorations of letters partly illegible or lost in the original
[abc?]
uncertain readings
[...]
illegible letters, number unknown
[---]
illegible letters, number known, indicated by broken line
[3]
illegible letters, approximate numbers known, indicated by numeralwith
[a(/b)]
ambiguous readings
abc [
end of line lost through damage
abc
text deleted in the original manuscript
[1] $// // phur ba'i [--]m rgyud ni/ /cho ga 'i rgyud [ni?] ki la ya [---] gnyis [k?]yi [don dang?]
[2] tan tra sde gsum kyi mdo' btus nas/ /las rnam s[-]u [2] gi cho ga 'i rgyud ni las cher
b[t?][--]
[3] la phur ba'i rtsis mgo rnam pa bzhi bstan te/ /phur bu'i no phyi ka dang phur ba'i bsam
[rgyud?]
[4] [-]/phur ba'i yon tan dang/ phur bu 'i grub pa'i rgyu dang / rnam pa bzhi 'o/ /de la phur bu []
[5] no phyi ka ni/ lha tib ta tsag kra khro bo chen po 'i sku mdog dmar/ / spyan gsum phyag
8

Rig 'dzin Vol. Sha folios 43-60; gTing skyes Vol. Sha: 93-128 ; mTshams brag vol. Ji folios 214-258 ; Taipei
5120 vol. 61 pages 316-322; sDe dge Vol. Zha folios 145-161.
9
This is not the ocassion to digress at length on such relationships - but to give a short example, the
Guhyasamja root tantra (Ch.14 65-68) has three Vajrakla mantras that strike at the samayas of body, speech
and mind. These mantras reappear in IOLTibJ754,81-82 and in NGB texts such as the Phur pa bcu gnyis Ch.11
as mantras for bestowing consecrations of body, speech and mind to the material kla. The first of the three (that
of body) reappears throughout Vajraklaya literature of all periods in the mantras of the goddess 'Khor lo rgyas
'debs ma, Vajraklaya's main consort, and also in the mantras of Amtakualin as used in Vajraklaya texts.
Whether passages such as Phur pa bcu gnyis Ch.11 relate more closely to the root Guhyasamja or to its
commentaries remains to be examined. The latter might seem more likely, but we can not be sure as yet.

6
[6] [dru]g pa/ /zhabs gcig rdo rje bu [--][rtse?] [2]/ /[khams?] kyi [g?]nod sbyin [-]an po
'tshir [zhing/(/zhig)]
[7] [bdag?] dang gnyis su [3]r dbyings [gcig?] pa ni / no phyi ka 'o/ / bsam rgyud [n?]i [...]
[8] dbyings gcig pu las/ /lag pa g.yas kyi mthil tu/ nya las zla ba'i dkyil tu [sgyur?] /
[9] thabs kyi rang bzhin yin bas/ /khro bo chen po bcu/ g.yon kyi mthil du ma las/ /
[10] nyi ma 'i dkyil 'khor tu gyur te//shes [rab] kyi rang bzhin kyi rtags [rtags?] [yin bas?]
[11] khro bo chen mo bcur dmyigs pa la rtsogs pa ni/ / [bsam?] rgyud [phu]r bu 'i yon tan
[12] nI/ /de ltar khyab pas tshe 'di la bgegs zhI ste/ /bsod nams kyi tshogs thob/ /[2]
[13] pha rol kyi mtho ris kyi gnas su phyin pas/ /ye shes kyi tshogs thob pas/ /bsod
[14] nams dang ye shes kyi tshogs rnam pa gnyis thob pas/ /yon tan/ /phur bu 'i grub
[15] rgyu ni de ltar byas nas/ /bgegs zhi ste/ yon bdag gi bsam ba grub/ /mtho [ris]
[16] kyI gnas thob/ /tshogs chen po gnyis kyang rdzogs/ /thabs dang shes rab kyi
[17] rang bzhin kyi las ma g.yos pas/ yon tan dang grub pa'i rgyu 'o / /phur bu 'di lag pa
[18] gnyis kyi bar du dril zhing gsor ba'i tshe/ /gnan bzlog gi phur bu'i rgyu ni/ / mtshon
[19] myI la babs pa'i lcags la bgyi'/ /rgya mdud kyi steng du he ru ka dgod/ /ngos bzhir phrin
[20] las kyi khyad bar dang [sbyar(/rgyud)] te dgod/ /mgul zur brgyad la ma mo chen mo
brgyad dgod/ /rtse mo la
[21] mu ka brgyad bgod nas/ /sna la bdag rang la gdab/ [-]'i bsam rgyud ni/ rgyu phun sum
tshogs pa
[22] ste/ /lhar byin kyis brlabs nas grub pa gsol te/ /'dod pa'i khams man cad du gdab / bgegs
[23] la gdab pa'i thabs ni/ /rgyu kun lhar tshogs nas/ /lag pa'i bar du drild pas zhe sdang gi
[24] bsam ba myI skyed/ snyIng rje chen pos gzhi bzung/ /'od zer dang 'phro 'du byung bas//
gang la bya ba
[25] 'i gzugs la phog pas byang cub kyi sems sk[y?]es pas/ /zhi ba chen po 'i rang bzhin [tu?]
[26] gyur par bsams nas/ /phur bu bsgrags pa'I tshig bshad 'di brjod do/ khro bo rgyal po 'di
dag
[27] gis/ /bgegs nI bkug nas rnam par gzhig/ /blo ldan rab du sbyor ba yis/ cho ga bzhin du
[28] phur bus gdab/ /rdo rje rgyal chen bdud rtsi po/ /rdo rje phur bu nyid gnas pas/ ud dpal
sngon po
[29] 'i mdog 'dra ba/ /bgigs kyi tshogs la 'og du gzigs/ /lte ba man cad cha rnams ni rtse mo
[30] lta bur rnam par [... ] de 'i sngags rnams sbyor bas yis/ / rdo rje phur nges btab na/ bgegs
[31] [lus?] shin tu myI g.yo 'o/ /o gha gha gha ta ya gha ta ya/ sa rwa tu shta ni pha/ ki la ki
la ya
[32] sa rba ba pham pha/ / h h [ba]dzra dha rod a [-] pa ya ti / [
[3] Translation as a whole, without comments
Conventions:
[---] Gaps in the text, one dash for each missing syllable.
[...]
Gaps where enumerating the number of missing sylables is impossible.
[abc] Conjectures occasioned by illegible text or difficult meaning of which we are
reasonably confident
[abc?] Conjectures of which we are a little doubtful
(abc) Occasionally we present the relevant Tibetan word in italics within round brackets,
sometimes with a Sanskrit or English gloss, to help the reader understand our thinking.
In particular, please note that the beginning of the text is damaged and illegible, and an
accurate reconstruction does not seem possible.

7
Note also that the text variously uses the words phur bu, phur ba, and klaya to
describe the implement and deity; we reproduce these usages verbatim as they occur. This is
for a reason: the names and terms are currently used differently to the way they appear in
P349 and are also subject to contemporary debate. The term phur bu (sometimes interpreted
as equivalent to klaka) in more modern usage more generally refers to the implement, while
phur pa (sometimes interpreted as equivalent to kla) can refer to the deity or the implement.
While such restricted usages might possibly be intended in our text - with the difference that
here phur pa takes the form phur ba (see lines 1, 3 and 4) which does not nowadays occur at
all except as an error - unfortunately our text is not consistent. We have phur bu'i yon tan in
line 12, referring back to phur ba'i yon tan in line 4. As for the term klaya or vajraklaya:
this is absolutely ubiquitously used in Tibetan texts of all historical periods (including the
present) to refer to the yidam form of the deity or to its tantric texts rather than the implement,
yet it is nowadays under attack from the majority of Western scholars, who wish to impose a
term more commonly used for the implement (kla or vajrakla) in its place. I have argued
however that there is little point in hyper-Sanskritising the Tibetan usage of klaya into kla.10
At all other times, technical terms are in English, or Sanskrit where no English term is
established (eg heruka or mt).
[1]
[2,3]

[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
10

As for the Phur ba [gtam rgyud, oral tradition? or bsam rgyud, Tantra meditation
tradition?], the Tantra ritual [tradition] [- -] the meaning of the two [- - -] Klaya [...]
extracted from the stras of the three sections of Tantra: regarding the ritual traditions
of the [- - -] rites greatly [- - -] four principal headings are taught for Phur ba: Phur
bu's means for attainment (no phyi ka = sdhanopyik), Phur ba's meditational
[tantra tradition?]
[-] , Phur ba's qualities, and Phur bu [as a] basis for accomplishment, four in number.
From these,
regarding the Phur bu means for attainment: the deity Dptacakra (lha tib ta tsag kra),
the great wrathful one, has a body colour of red; is three-eyed and [six]-armed;
has a single lower limb (zhabs gcig) [of a] Vajra [-] [point?][- -]; he crushes the [- -]
[yakas] of [the world?]; and
[onself?] being [one] with the [non-dual] expanse [- - -], this is the means for
attainment. As for the meditational tantra [tradition][......]:
from out of the single [non dual] expanse, on the palm of the right hand, [visualise
arising out of] the syllable Nya, a moon disc;
being of the nature of skilful means, [upon it arise?] the Ten Great Wrathful Deities.
On the left palm, from the [syllable] Ma
arises a sun maala; [since it is?] [- - -] the sign for the nature of wisdom,
meditate and so forth as [arising upon it?] the Ten Great Female Wrathful Deities. As
for the [Tantra meditation tradition's] virtuous qualities:
by pervading [one's hands] in this way, one's obstacles in this life will be pacified;
[thereby] the accumulations of merit can be attained [--]
[so that] one passes on to an abode in the transcendent heavens [where] the
accumulation of wisdom can [also] be attained;
and thus the two accumulations of merit and wisdom can both be attained: [hence
these are its] virtuous qualities. Regarding the Phur bu as a
basis for accomplishment: having done accordingly [as above], the obstacles [will be?]

All the more so since such a learned expert in Sanskrit Tantric languages as Alexis Sanderson has also argued
that the form klaya might well have been current in Indian Tantric circles to refer to the deity, and need not be a
Tibetan distortion of kla (the implement). See Mayer 1996:165-6.

8
pacified, patron's wishes will be accomplished,
[16,17] heavenly abodes will be attained, and even the two great accumulations will be
completed. Since these rites whose [very] nature is skilful means and wisdom do not
waver, they are a basis for accomplishment and qualities.
[18] On the occasion of rolling and brandishing such a Phur bu between one's two hands,
regarding the materials of the suppressing and repelling Phur bu,
[19] make it out of iron from a weapon [that has fallen on a man?]; above its knotted cords,
establish Heruka; on the four faces,
[20] establish those endowed with the [four] particular enlightened activities; on the eight
sides of the neck, establish the eight great Mtrs;
[21] having established the eight Mukhas (mu ka brgyad) on its point, [with such a tip (sna
la), one can strike at one's own self?]. As it is said in the meditation tradition: since
this is the Perfection of Material,
[22] consecrate it as the deity and request accomplishment, and one will be able to strike at
[all] the Realm of Desire (kmadhtu) below.
[23] As for the means of striking at the obstacles: having assembled all the material as the
deity, when rolling it between the hands, do not give rise to angry thoughts,
[24] [but] with great compassion, adhere to the Base (gzhi). Giving rise to the emanation
and reabsorption of rays of light,
[25] as these strike at the form of whoever [the rite] is to be done for, generating bodhicitta,
imagine that they become transformed into the nature of the Great Peace,
[26] and utter these verses of Phur bu recitation:
[27] By this wrathful king
The obstacles are summoned and totally destroyed.
Those supremely endowed with good intellect
[28] Strike with the phur bu in accordance with the rite.
The great Vajra King, the Amta being,
Abides as the Vajra Phur bu itself,
[29] Blue in colour like an utpala,
Gazing down at the hosts of obstacles.
The part below his navel
[30] Is like a point, and utterly [.....] [brlag, destroys?]
If, endowed with his mantras,
One definitively strikes with Vajra Phur [bu],
[31] The bodies of the obstacles will become quite immobilised
o gha gha gha ta ya gha ta ya / sa rva du shta ni pha // ki la ki la
[32] ya sa rva ba pham pha// h h [ba] dzra dha rod a [-] pa ya ti [...

[4] Translation with our comments

Translation of line 1-4
[1]
As for the Phur ba [gtam rgyud, oral tradition? or bsam rgyud, Tantra meditation
tradition?], the Tantra ritual [tradition] [- -] the meaning of the two [- - -] Klaya [...]
[2,3] extracted from the stras of the three sections of Tantra: regarding the ritual traditions
of the - - - rites greatly - - - four principal headings are taught for Phur ba:Phur bu's
means for attainment (no phyi ka = sdhanopyik), Phur ba's meditational [tantra
tradition?]
[4]
[- ], Phur ba's qualities, and Phur bu [as a] basis for accomplishment, four in number.

Comments on lines 1-4

Line 4:
Even if the beginning of the text has been partially destroyed rendering it untranslatable, we
can at least with some certainty interpret the term no phyi ka, which is one of the four topics
or headings that structure the text (phur bai rtsis mgo rnam pa bzhi). This term had puzzled
Stein who remarked that he could not understand it (mot inconnu), but Matthew Kapstein
has provided us with the answer. According to Kapstein (personal communuication, February
1, 2000), it is quite common within Dunhuang ritual texts, and seems to be a Tibetan vulgar
Sanskrit based on sdhanaupayika> naupayika> no p(h)yi ka, i.e it means methods for
sdhana. Following Das, Hackin and other sources, we have found various citations of the
term that support Kapsteins interpretation: DVP 553 (no pyi ka) and 554 (no pyi k), both on
page 171 of the de Valle Poussin catalogue; Hackin pages 8 and 46 (no phyi ka); see also
Sdhanaml 209, 228, 240, 248, which mention sdhanopyik.11 See also Edgerton p.146,
who cites Sdhanaml 415.5, 449.17, 468.12, and 486.3 (all colophons); and Blue Annals p.
160, which refers to the Tenjur text 'Jig rten snang byed zla ba'i no pi ka, Tg. rGyud no. 3584.
Translation of lines 5-7
From these,
[5]
regarding the Phur bu means for attainment: the deity Dptacakra (lha tib ta tsag kra),
the great wrathful one, has a body colour of red; is three-eyed and [six]-armed;
[6]
has one foot, a Vajra [-] [point?][- -]; he crushes the [- -] [yakas] of [the world?]; and
[7]
[onself?] being [one] with the [non-dual] expanse [- - -] , this is the means for
attainment.
Comments on lines 5-7
One of the questions that arises here concerns the deity Dptacakra, whose name appears on
line 5. Here in P349 Dptacakra is clearly male: lha tib ta tsag kra khro bo chen po.
Moreover we find passages in a number of other authoritative ancient sources that remix
many of the words of P349, and which also affirm Dptacakra as a male deity. Such passages
occur in the Vajraklaya writings of the early Sa skya pa master Grags pa rgyal mtshan (11471216); in an NGB Vajraklaya scripture called the Phur pa gsang chen rdo rje 'phreng ba (for
these passages, see Appendix 5.4 below); and also in the still current major Sa skya pa
Vajraklaya ritual, the Phur chen, with its commentaries (which we will discuss shortly). Yet
in much of the very recent tradition, above all in most Western translation and study of
Vajraklaya, Dptacakra is almost invariably taken to be the Sanskrit name of the female deity
whom Tibetan sources call 'Khor lo rgyas 'debs ma.12 This 'Khor lo rgyas 'debs ma is the
central Vajraklaya deity's famous female consort of union (sbyor), a more common
11

Thanks to Gudrun Melzer for discovering these titles within the Sdhanaml.
For Dptacakra used as the Sanskrit name for 'Khor lo rgyas 'debs ma, see for example my own previous work,
Mayer 1996:174 and Mayer 1998:293; or see Boord 2002:39 and Boord 2002:316. Both Boord and myself have
here been continuing the usage already well established in the modern West. See also the numerous unpublished
works on Vajraklaya produced by various Western Dharma organisations for their practitioners, for example,
the impressively extensive and detailed works of the Vajravairocana Translation Committee based in the USA to
which half a dozen leading rNying ma pa lamas and mkhan pos contributed; or those circulated among the
Western Sa skya pa community, to which several major Sa skya pa lamas have contributed: in all of these,
Dptacakra is ubiquitously used to indicate the female consort 'Khor lo rgyas 'debs ma. However, Boord 1993
temporarily changed his usage from Dptacakra to Tptacakra, without giving any reasons for doing so; I have
never encountered the form Tptacakra other than in Boord 1993, and certainly not so far in any Tibetan sources.
More recently, Boord 2002 has (albeit again with no explanation) reverted from Tptacakra back to the more
usual Dptacakra.
12

10
counterpart to Ekaja who is his female consort of killing (sgrol) - although whether these
two are really separate consorts, or two aspects of the same consort, is very ill-defined sometimes they are described as two separate consorts, sometimes as two aspects of the same.
But it is noteworthy that while so many recent sources now give 'Khor lo rgyas 'debs ma the
Sanskrit name Dptacakra, this usage is very rare in Tibetan literature, where she is, as far as I
can see, nearly always called only by her Tibetan name. In fact, I can not recollect ever having
encountered Dptacakra - nor any other Sanskrit name - offered for 'Khor lo rgyas 'debs ma in
any traditional Tibetan sources whatsoever. I am not saying there are no such - only that I
have no recollection of encountering any. On the other hand, there are several occasions when
Ekaja is identified as 'Khor lo rgyas 'debs ma, for example, in the Sa skya literature; but here
the tendency seems to be more a conflation of the two consorts.13
Nevertheless, there is at least one good justification for the modern usage of
Dptacakra - it comes from her mantra, o dptacakra hana hana h pha. But there is also a
good argument against it: it in no way translates her Tibetan name. The Tibetan name 'Khor lo
rgyas 'debs ma means something like 'She who seals with the wheel(s)'. But, as Jam mgon
kong sprul points out (following earlier commentarial tradition), the etymology of the mantra
is as follows: dpta means blazing, cakra is a wheel, and hana hana is the exclamation strike!
strike!; so the whole mantra means "strike, strike with the blazing wheel!" He adds that it is
because of the meaning of this mantra that the yum appears holding a wheel of destruction in
her right hand.14 Thus the literal Tibetan translation of Dptacakra would be 'Khor lo 'bar ba,
not 'Khor lo rgyas 'debs ma.
We must also note at this point that the root Guhyasamja (14.65) has a mantra o
chindha chindha hana hana daha daha dptavajracakra h pha which clearly anticipates
the Tibetan tradition of this consort of Vajraklaya, since she subsequently takes the greater
part of this mantra as her own; but unfortunately, the Guhyasamja root tantra is not at all
clear about what (if any) the gender implications of the words dptavajracakra might be does this point to a male name, as our several quoted traditional sources would have it, or to a
female name, as many modern and recently translated sources would have it? Or neither? Or
both? Unfortunately, the Guhyasamja commentaries are not any more clear than the root
tantra about the gender implications of the words - from what I have seen so far, they only add
to the uncertainty.15
Compare Phur Chen 16.4 ff where the usual Sa skya form of 'Khor lo rgyas 'debs ma is elaborately visualised,
with Phur Chen 36b. - 37a where with no explanation this same visualisation is lengthily praised as Ekaja; for
a similar passage, see also Grags pa rgyal mtshan p. 184.4. A myes zhabs offers no explanation in his great
commentary (see below).
14
See his famous commentary, dPal rdo rje phur pa rtsa bai rgyud kyi dum bui sgrel pa snying po bsdud pa
dpal chen pai zhal lung zhes bya ba, p.101.
15
Unfortunately, I have not had time to check on these commentaries with adequate thouroughness (there are a
great many of them - around a dozen Tenjur volumes are dedicated to Guhyasamja commentaries!).
Chintaharan Chakravarti's edition of the Sanskrit manuscript of Candrakrti's
Guhyasamjatantrapradpodyotanak from the Rahul Collection does clearly interpret Guhyasamja 14.65 as
referring to a female (page 159, paragraph 3: omitydiko niranto mantra | chinda chinda sdhyakyam | hana
hana kyabalam | daha daha kya[m] | dptavajra ca cakra ca yasy dptavajracakretymantraam | h
phaiti codanam). Boord translates this very nicely, but accepts it without further question (Boord 2002:39).
However, J.S. Jha points out in his introduction that the Rahul Collection text appears to comment on a
Guhyasamja root text that has a number of readings not found in other Guhyasamja editions - so further
research of Sanskrit sources is probably called for. Meanwhile, the Tibetan translation of this famous
commentary by Candrakrti as witnessed in the Peking and Golden Tenjurs (Peking 2650, Vol. Sa f.155b;
Kinsha rgyud 'grel vol Sa,201) does not specify a female at all (o zhes bya ba la sogs pa ni sngags te/ ming
mtha' med ces bya'o/ /tshinda tshinda zhes bya ba ni/ bsgrub par bya ba'i lus chod cig pa'o/ /ha na ha na zhes
bya ba ni lus kyi stobs choms shig pa'o/ /da ha da ha zhes pa ni lus bsregs shig pa'o/ /dpta badzra cakra zhes
bya ba ni rdo rje dang 'khor lo 'bar ba can gang yin pa la/ 'bar ba'i rdo rje 'khor lo can zhes bod pa'o/ /h pha
ces pa ni bskul ba'o/). Another Guhyasamja commentarial text from the Peking Tenjur (Vol. Sha, 243b-244a)
13

11
But here in Pelliot 349, the name Dptacakra certainly refers not to the very well
known female consort deity of Vajraklaya, but to a wrathful male deity (khro bo chen po,
khro bo rgyal po) with a Heruka upper body and a kla-shaped lower body. In the broader
Vajraklaya literature, such an iconographical form most typically represents the 'Supreme
Son' or sras mchog form of the Vajraklaya deity, frequently associated with the
quintessentially male material kla as a ritual implement and the deity's nirmakya.16
This application of the name Dptacakra to a male deity with a phur bu shaped lower
body is certainly not unknown even in literature in regular contemporary use: the Phur chen
sdhana, the major current practice of the Sa skya Khon lugs phur pa tradition, has the
following verse (starting on folio 24r line 6):
tpta ca kra phur pa'i lha/ mthing nag gcer bu ral pa can/ sku stod khro bo chen po la/
zhal gsum phyag kyang drug pa ste/ dbu la rigs lnga'i sangs rgyas rdzogs/ lte ba man
chad sku yi cha/ utpal sngon po'i 'dab 'dra ba/ 'bar ba'i phreng ba 'khrigs pa'i 'od/
lcags kyi phur pa zur gsum pa/ drag por gyur ba'i phur pa ste/ btab na lha yang brlag
par 'gyur/ gnod byed bgegs la smos ci dgos/
'Dptacakra, Phur pa deity, / Dark blue and naked, with matted hair, [Your] upper body
is a great male wrathful one./ With three heads and six arms,/ [Your] heads are
perfected by the Buddhas of the Five families./ The part of your body which is below
the middle/ Is like the petals of a blue lotus./ With light amassing in a blazing garland/
[Around] the three-sided iron phur-pa,/ This is the phur pa [which has] become
wrathful! If it were to strike, even the gods would be destroyed, / What need is there to
speak of the harmful forces and obstacles?'
The great 17th century Sa skya savant 'Jam mgon A myes zhabs wrote the definitive
commentary on the Sa skya Phur chen, and his analysis of the words tpta ca kra phur pa'i lha
etc as cited above are most interesting and quite clear. He says that: they refer to the material
kla held in one's hands, which is visualised as the 'Supreme Son' kla; that this is Vajraklaya
(ie not his consort); and that the Tibetan meaning of his name is 'khor lo 'bar ba.17 We can see
that the words 'khor lo 'bar ba follow the literal translation of dptacakra that we find in much
Tenjur commentary on Guhyasamja Ch.14, and also in Tibetan commentarial explanations
of the meanings of the Sanskrit mantra of Vajraklaya's female consort - but which is
generally denied her as her actual name in Tibetan sources, which instead call her 'Khor lo
rgyas 'debs ma.
Note also that some of the lines here from the Phur chen are parallel to the
Guhyasamja commentarial materials presented in the appendix given below: / utpal sngon
po'i 'dab 'dra ba/ 'bar ba'i phreng ba 'khrigs pa'i 'od, and also sku stod khro bo chen po la/
which is attributed to Ngrjuna, the rguhyasamjatantrasyatantraknma or dPal gsang ba 'dus pa'i rgyud
kyi rgyud 'grel pa, also comments on Guhyasamja 14.65 in such a way as to leave gender unspecified: o ni
rnam par snang mdzad do/ /tshinda zhes pa ni chod ces ston to/ /ha na zhes pa ni bsgrub bya'i lus sod cig ces par
ston to/ /dpta badzra zhes pa ni rdo rje 'bar ba ste/ /bod pa'i tshig go /h dang pha ni khros pa la'o/. Clearly,
we will need to look further in both Sanskrit and Tibetan sources before arriving at a clear decision about
Dptacakra's gender in Guhyasamja commentarial literature.
16
One should note that exceptions do of course occur. For example, in some instances forms with heruka upper
bodies and kla lower bodies can represent dharmakya deities of the five enlightened families (but these are
nevertheless still male!). But this form most typically represents the male nirmakya 'Supreme Son' or
material kla.
17
dpta tsa kra phur pa'i lha ces sogs brjod/ de dagi don ni/ dpta tsa kra zhes pas ni lag na yod pa'i sras mchog
de nyid gsal btab pa yin la/ 'o na 'di badzra k la ya yin pa la/ dpta tsa kra ste 'khor lo 'bar ba zhes brjod pa. See
'Jam mgon A myes zhabs, ed. Sopa 1973:347.

12
zhal gsum phyag kyang drug pa ste/; also lte ba man chad sku yi cha. In addition, the above
verses are close to the verses from Grags pa rgyal mtshan and from the NGB's Phur pa gsang
chen rdo rje 'phreng ba'i rgyud that we cite in Appendix 5.4, where we again find a male
Dptacakra - although there following P349 in giving the deity's colour as red rather than blue.
In fact, as we have already pointed out above, these particular verses (or, more
commonly, various remixes of them) are quite widespread in Vajraklaya literature in general,
but interestingly they are not always taken to indicate a single male deity as they do above and
in P349 - or perhaps, even where they seem to, they are often interpreted otherwise. For
example, Martin Boord has presented a translation of the verses as found in the 18th century
Byang gter author Phrin las bdud 'jom's Byang gter phur pa'i dbang gi lo rgyus legs par bshad
pa nor bu'i do shal. However, (perhaps following an uncited oral explanation?), Boord
appears to conjecturally introduce the word and into his text, to get around what he quite
understandably (but perhaps mistakenly in this case) sees as the anomaly of the name
Dptacakra being applied to the quintessentially male kla deity. In this way Boord tries to
attribute the name Dptacakra to the female consort instead. Hence he gives us a yum-yab
interpretation: Oh Tptacakra [and] the Kla god, dark blue in colour, naked, with long
dishevelled hair...; I wonder if Boord should have more simply written: Oh Dptacakra, Kla
god, dark blue in colour, naked, with long dishevelled hair.... thus accepting the transmitted
textual evidence of Dptacakra applying to a single male deity? See Boord 1993:107.18
So the question arises: is the name Dptacakra, ubiquitous in the last 30 years or so as
referring to Vajraklayas female deity of union 'Khor lo rgyas 'debs ma, being correctly used?
I regret that adequate answers to this question can probably only be achieved by a detailed
study we can not attempt here - we would have to look through enough sources to ascertain at
which point the Sanskrit name Dptacakra became applied to 'Khor lo rgyas 'debs ma. Given
that her mantra contains the element dptacakra, and the sheer depth of contemporary
opninion that calls her Dptacakra, this identification might in fact turn out to be be quite old,
and even correct. In which case, it raises the issue of the double application of the name
Dptacakra to Vajraklaya's Supreme Son and to his consort of union alike. While some
Indological scholars might argue that such name and gender ambiguities are unremarkable
from their point of view, my impression is that they are sufficiently rare in rNying ma pa
literature - at least for deities with such prominent and clearly defined personalities as these to pose an interesting question. If this is anything more than just a modernist confusion
prompted by the Sanskritising impulse of Western scholarship, then was the ambiguity
originally 'planned', a doctrinal and ritual development that was deliberate from its outset? after all, there are few Indian tantric traditions more minutely analysed than the Guhyasamja,
and few Tibetan tantric traditions more commented upon than Vajraklaya. Or was it a
possibly anachronistic anomaly arising from the gradual emergence of Vajraklaya and his
maala out of the conceptual vagueness of the pantheonic margins - where identity and
gender is more often ill-defined - into the minutely scrutinised limelight of pantheonic
centrality - where identity and gender is usually more clearly defined? Or did it originally
arise from the confusion of a faulty scribal transmission that was later rationalised? Or from

18

Boord (1993:108, note 398) seems to say that Phrin las bdud 'joms took his text from the 17th century bKa'
brgyud pa author gTsang mkhan chen 'Jam dbyangs dpal ldan rgya mtsho's rDo rje phur pa'i chos byung, but I
am not sure if this is what he means. Nevertheless it is clear that the author (whether Phrin las bdud 'joms or
gTsang mkhan chen) associates these verses with the famous Pharping narrative, which we find in Pelliot 44 and
throughout subsequent Phur pa histories: hence the author has Padmasambhava utter a version of these verses in
the Asura cave at Pharping in order to tame the various troublesome godesses there (he lists them as Sho na,
bDag nyid chen mo, and bSe mo).

13
some other kind of interpretational confusion between mantras and names?19 Or was it a
result of alternative interpretations of Guhyasamjatantra 14.65 and its commentaries?
Not without important reservations, one can also consider an additional perspective:
the Supreme Son can be functionally very close to the consort. Both can represent
Vajraklayas practical apotropaic activities of summmoning and liberating obstacles (which
typically make use of a further minor pantheon of more marginal deities), as opposed to the
more central soteriological functions closely associated with the main deity Vajraklaya
himself and his immediate circle. Or, to use rNying ma pa terminology, the Supreme Son and
the consort alike (along with other more marginal often female Vajraklaya maala deities
such as the dog-headed goddess vna20) can be especially important in the smad las, the
lower rites of eliminating obstacles, rather than the stod las, or 'upper rites' of realising
dharmat.21 Could this functional closeness of the male nirmakya form to the more
marginal female deities of activity and his consequent co-habiting of various subsidiary
maalas with them contribute to occasional name or gender ambiguity? But a serious
problem with this analysis is that it is the consort of liberation (sgrol), Ekaja, who fits this
scenario, rather than the consort of union (sbyor), Khor lo rgyas 'debs ma.
Nevertheless, gender and name ambiguity is certainly not so rare among the more
marginal deities of the Vajraklaya maala: one can point out that the twenty attendants of
the Ten Wrathful Deities (two for each) can sometimes be seen as all female, and sometimes
as ten males and ten females; likewise vna can also sometimes (but comparatively rarely)
19

Some examples of potentially confusing passages: Ch.11 of the NGB's Phur pa bcu gnyis bestows the body
consecration on to the male material kla with the mantra: o chindha chindha hana hana dptackara h, a
form of words which on its own and without added commentary might imply to the reader that dptackara is the
name of the male material kla, as in P349. Incidentally, we know this must be an old tradition, because an
almost identical pattern occurs in the Dunhuang texts IOLTibJ 754, 81-82, and IOLTib J331.III, where the
former bestows the body consecration with om tshin dha tshin dha da ha da ha ha na ha na tib ta ca kra hum
phad; and in the latter with o ha na ha na tib ta ca kra h phat. (Note: these mantras, and the other mantras
for consecrations of speech and mind that accompany them, are all derived from the Guhyasamja Ch.14). But
Phur pa bcu gnyis Ch.11 continues by seemingly indicating that the latter part of the mantra is that of the female
consort, invoked to make the body consecration of the male material kla by joining in union with the male deity
so that they can produce bodhicitta'; hence it is not necessarily the case that Dptacakra refers to the male
material kla itself or its male deity form. The Phur pa bcu gnyis Ch. 11 gives the following explanation: Then,
interlinking together one's eight fingers,/ In between one's thumbs which are in line,/ Insert the kla, and say the
following words:/ O and chindha chindha and/ Hana hana dpta and/ Cakra h; with this superlative
utterance,/ The bodhicitta of the lord's union with his consort/ Emanates superlatively, and dissolves into the
[kla]. Sparks filling a thousand worlds shoot forth,/ And [the kla] obtains a great power and glory/ Which can
attain all goals with certainty; [Upon this], present it to rest within the centre of the secret consort. The
interlinking of hands or fingers in such a context to represent a yab-yum is also mentioned in Guhyasamja 14.66
and its commentaries, as well as in most Vajraklaya literature. The chindha chindha elements occur in much
Vajraklaya literature in mantras of Amtakualin, who is often identical to Vajrakla; while the hana hana
dpta and cakra h elements are probably here taken to be those of 'Khor lo rgyas 'debs ma, Vajraklaya's
consort of union (sbyor). The placing of the kla to rest within the centre of the secret consort probably refers to
the phur khung or kla stand, which is often identified with the consort's "sky" into which the [male] implement
can be put to rest vertically. We can conclude that this mantra in this instance is not necessarily indicative of the
male deity being called Dptacakra; rather, it might simply be the mantra of 'Khor lo rgyas 'debs ma. But such
passages (especially the mantras on their own without further commentary) can easily become a source for
confusion.
20
Her Sanskrit name is variously rendered as vnamukh, vanmukh, vna, or vana and her Tibetan name
as Sho na or Shwa na. She is the most famous of the Vajraklaya protectresses, whose place in the Vajraklaya
maala traditionally goes back to her being tamed by Padmasambhava at Pharping (cf Pelliot 44).
21
Evidence for this can again be found in the Phur pa bcu gnyis, where Ch. 9 is devoted entirely to the Supreme
Son. Here the Supreme Son is envisaged as having his home in the maala of the secret consort, 'encircled by a
blazing radiance of fire, (perhaps an allusion to Dptacakra as a female, perhaps an attribute of himself), where
he co-habits with relatively marginal and mainly female 'lower rite' deities of killing and liberating and the
largely female Vajraklaya protectors such as vna and Remati.

14
have male counterparts22, and the descriptions of the other Vajraklaya protectors can also
vary quite a lot. We can conclude - if it is not simply a confusion created by modern
scholarship! - that what might be surprising about the gender and name ambiguity of
Dptacakra is not so much the ambiguity as such, but its existence between such famous and
well-defined deities as Vajraklaya's main consort and his 'Supreme Son'.
Translation of lines 8 to 11
As for the meditational tantra [tradition][......]:
[8]
from out of the single [non dual] expanse, on the palm of the right hand, [visualise
arising out of] the syllable Nya, a moon disc;
[9]
being of the nature of skilful means, [upon it arise?] the Ten Great Wrathful Deities.
On the left palm, from the [syllable] Ma
[10] arises a sun maala; [since it is?] [- - -] the sign for the nature of wisdom,
[11] meditate and so forth as [arising upon it?] the Ten Great Female Wrathful Deities.
Comments on lines 8 to 11
Line 8:
The text here reads Nya. We know however that this is quite likely an error for A because
more or less all other sources give A in this context of visualising the sun and moon on the
hands as part of the kla-wielding ritual: to mention but a few, the old Dunhuang texts IOL
Tib J 754, 81-82; the NGB's Phur pa bcu gnyis Ch. 11;23 the main current Sa skya sdhanas
(Sa skya Phur chen f. 24r and the dPal rdo rje phur pai bsnyen sgrub gsal byed bdud rtsii
od can f. 150). The process is a complex one of consecrating the hands and the kla and
solemnly wielding them both in elaborate and graceful hand gestures or mudr. P349 gives an
extremely abbreviated reference to this famous kla rite.
Lines 9 and 10:
The Ten Wrathful Deities (khro bo bcu) and their consorts are very important in the
Vajraklaya traditions and of course occur throughout many other Vajrayna texts in addition.
In this version of this rite, the Ten Wrathful Deities and their consorts are mentioned as
arising directly from the visualised sun and moon on the palms of the hands. In the Phur pa
bcu gnyis Ch.1124 and the current Sa skya traditions (dPal rdo rje phur pai bsnyen sgrub gsal
byed bdud rtsii od can folio 150; Sa skya Phur chen f. 24r), the process is more gradual,
with the wrathful deities developing in stages out of the unions of the male and female
peaceful Buddhas, all performed with elaborate visualisation and hand gestures or mudr.
Translation of lines 11-17
As for the [Tantra meditation tradition's] virtuous qualities:
[12] by pervading [one's hands] in this way, one's obstacles in this life will be pacified;
[thereby] the accumulations of merit can be attained [--]
22

The gter ma of mChog gyur gling pa (mChog gling gter sar) have both male and female vna deities: for
example, the Zab bdun mchog zab yang dag gi shwa na chen po'i zlog pa'i phrin las bcol ba (volume 17, pp.
559-569) has the passage: yab gcig shwa na mu kha che/ mthu chen bdud rgyal rnams kyi gshed../ khyod kyi
yum gcig shwa na ma/ mkha la 'khor 'das thams cad rdzogs/ (p. 562). Thanks to Andreas Doctor for these texts.
Note however that in the Shwa na dkar nag gi rgyud of the NGB (sDe dge Zha f. 260; mTshams brag Ji p.1096;
gTing skyes Sha p.493; Nubri Sa f.65 gong; Rig 'dzin Sha f.222), which is the only Tantra specifically for vna
deities with which I am currently familiar, only female forms of vna are ever explicitly mentioned (although
it is also just conceivable that male ones might also be very vaguely implied, especially with a liberal helping of
creative exegesis; at least they are not explicitly precluded).
23
gTing skyes Vol Dza p.110; sDe dge Vol Pa f.217r
24
gTing skyes Vol. Dza pp. 110-112, sDe dge Vol. Pa f. 217r-218r

15
[13]

[so that] one passes on to an abode in the transcendent heavens [where] the
accumulation of wisdom can [also] be attained;
[14] and thus the two accumulations of merit and wisdom can both be attained: [hence
these are its] virtuous qualities. Regarding the Phur bu as a
[15] basis for accomplishment: having done accordingly [as above], the obstacles [will be?]
pacified, patron's wishes will be accomplished,
[16,17] heavenly abodes will be attained, and even the two great accumulations will be
completed. Since these rites whose [very] nature is skilful means and wisdom do not
waver, they are a basis for accomplishment and qualities.
Comments on lines 11-17
lines 11-14:
Here we find a rationale for the apotropaic aspects of the Vajraklaya rites: specifically aimed
at removing this-worldly obstacles, they only do so in order to enable spiritual practice, as the
first stage of a gradualist spiritual program. This kind of rationale is also found in
hagiographic materials about early Vajraklaya practitioners: see for example the story of
gNyags Jnakumra as contained in the bDud 'joms chos 'byung.25 The reference to the
rebirth in a pure realm is noteworthy: in most Vajraklaya literature this is a virtue enjoyed by
practioners of Vajraklaya and their 'liberated' victims alike. Another of the Dunhuang Phur
pa texts, IOLTibJ331.III, makes this connection clear in its title, Zhi ba'i mchog 'pho ba'i
'phrin las bsdus pa'o - where 'phrin las refers to the Phur pa 'liberation' ritual, and 'pho ba to
the yogic transference of consciousness to the pure realm, here glossed as zhi ba'i mchog,
supreme peace.26
Lines 15-17:
This reiterates much of the above, but taking the perspective of yogic accomplishment rather
than the virtuous qualities of the practice.
Translation of lines 18-22
[18] On the occasion of rolling and brandishing such a Phur bu between one's two hands,
regarding the materials of the suppressing and repelling Phur bu,
[19] make it out of iron from a weapon [that has fallen on a man?]; above its knotted cords,
establish Heruka; on the four faces,
[20] establish those endowed with the [four] particular enlightened activities; on the eight
sides of the neck, establish the eight great Mtrs;
[21] having established the eight Mukhas (mu ka brgyad) on its point, [with such a tip (sna
la), one can strike at one's own self?]. As it is said in the meditation tradition: since
this is the Perfection of Material,
[22] consecrate it as the deity and request accomplishment, and one will be able to strike at
[all] the Realm of Desire ('dod pa'i khams, kmadhtu) below.
Comments on lines 18-22
Lines 19-21:
Phur pa bcu gnyis Chapter 1027 describes the materials for making a kla as ideally to be taken
from weapons such as knives and arrowheads, as well as from meteors or thunderbolts, and to
have qualities of cutting, sharpness, and hardness etc. It is also possible that the Phur pa bcu
gnyis advocates the use of iron from an arrowhead that has pierced a persons heart, but the
text is corrupt at that point, and the meaning ambiguous. The text here in P349 is slightly
25

Dudjom1991:601-605.
I hope to comprehensively study this important text in the near future.
27
gTing skyes Vol. Dza page 106; sDe dge Vol. Pa folio 216r
26

16
obscure (mtshon myi la babs pai lcags), but it seems a possible conjecture that the material is
meant to be iron from a weapon that has actually struck a person.
These lines also support one of the findings of Mayer & Cantwell 1994: that the classic
Tibetan kla design as we know it today was already in place by the time the Dunhuang texts
were written. Although extremely terse in describing the kla, nevertheless here at the very
least we have the knotted cords, the four-square base, and an eight-facetted shaft, features that
make unmistakeable reference to the ypa or Indian sacrificial stake (Mayer 1991).
The establishment of deities on the different parts of the kla is ubiquitous in all kla literature,
but the details of which deity is put where seems to vary from text to text and sdhana
tradition to sdhana tradition, which is perhaps understandable in that the different
Vajraklaya maalas are populated by slightly different arrangements of deities.
Nevertheless the placement of Heruka in his palace above the knotted cords (as here also)
does seem to be a constant. The deities of the four enlightened activities will probably be
those of the standard list of peaceful, increasing, powerful and wrathful activities, or else the
well-known four goddesses with iron hook, noose, iron chain and bell who summon and
bind. Mention is made here of the eight mu ka; possibly a popular Sanskritism (mukha = face
or head), referring to the famous animal-headed goddesses as found in many Vajraklaya texts
all of whose names end in -mukha; for example, eight occur in Chapter 7 of the Phur pa bcu
gnyis, in the context of the definitive arrangement of the central Vajraklaya maala.
More problematic is the culmination of this section, sna la bdag rang la gdab//. In particular
we have problems interpreting sna: the tip of the phur bu is more often referred to as dbal
than sna, and striking oneself on the nose with the consecrated phur bu is not mentioned
elsewhere! However, Cristina Scherrer-Schaub has pointed out to me that in Old terminology,
sna is an ancient term for gtso bo, which may be understood as a metaphorical expression for
dbal, meaning pinnacle or tip. I am following her suggestion here. dBal is of course widely
used to refer to the point of a phur bu, and is especially widespread in the Bon Phur pa
tradition. There is an important part of the rNying ma and Sa skya rite where the freshly
empowered kla is solemnly touched (not struck) to the five places and three gates, i.e. (1)
crown of head, (2) forehead, (3) back of right ear, (4) back of head, (5) back of left ear, (6)
forehead again (7) throat and (8) heart. More generally, from a soteriological point of view,
the ultimate function of the phur pa is to enable one to strike at the ignorance, desire and
aggression within oneself: this is always seen as the ultimate usage of the implement, a
soteriological interpretation that goes back to the Guhyasamja.
Line 22:
The Dunhuang Phur pa text IOL Tib J 331.III explains it will subsume the Vajraklaya
teachings within Seven Perfections (phun sum tshogs pa bdun). The first two of these are
gzugs phun sum tshogs pa, the Perfection of Form, and byin rlabs phun sum tshogs pa, the
Perfection of Consecration. The first details the materials and form to be used in making the
kla, and the second the methods of consecrating it. The Perfection of Material (rgyu phun
sum tshogs pa) mentioned here in P349 does not occur in IOL Tib J 331.III, but its initial
Perfection of Form seems to cover similar ground, and as in P349 is likewise followed by
consecration of the implement as the deity, which then forms the basis for the subsequent
rituals.
Translation of lines 23-26
[23] As for the means of striking at the obstacles: having assembled all the material as the
deity, when rolling it between the hands, do not give rise to angry thoughts,
[24] [but] with great compassion, adhere to the Base (gzhi). Giving rise to the emanation
and reabsorption of rays of light,
[25] as these strike at the form of whoever [the rite] is to be done for, generating bodhicitta,

17
imagine that they become transformed into the nature of the Great Peace,
[26] and utter these verses of Phur bu recitation:
Comments on lines 23-26:
The description of the actual wrathful rite makes clear its adherence to conventional Buddhist
ethics. Even while striking at the obstacles (bgegs), P349 insists the practitioner should not
give rise to angry thoughts, but should proceed with a mind of compassion that adheres to the
Base (gzhi, equivalent to Sanskrit laya, of course originally a Mahyna term but also much
adapted in rDzogs chen texts). Although not spelt out by name, the rite of forceful liberation
or killing (sgrol ba or moka) is clearly being referred to. In Tib J 754 81-2, also a
Dunhuang text dealing with the same rite, similar Buddhist principles are likewise invoked
(Mayer & Cantwell 1994). We find similar sentiments in the opening passages of IOL TibJ
331.III. The clear evidence of the Dunhuang kla killing rites taken as a whole seems to be
that they were fully ethicised and soteriologised. As we would expect from materials so
closely linked to Guhyasamja, the early Tibetan kla tradition of moka was not a sorcery
tradition, but a Mahyna Buddhist one, albeit in the final analysis most likely a bloodless
calque on non-Buddhist Tantric blood sacrificial rites of the type still so widespread in kta
religion. The rite of moka of course continues as a central practice in contemporary rNying
ma pa ritual, especially in the Vajraklaya traditions, and it is remarkable how little the rite
described in these Dunhuang texts has changed over the last millennium, if at all. For a
detailed study of moka, see Cantwell 1997.
The mention of the projection and reabsorption of light rays in line 24 is very similar to the
instruction given at the same point in the same ritual as described in IOL Tib J 754, 81-82
(Mayer & Cantwell:59-60).
Translation of lines 27-32
[27] By this wrathful king
The obstacles are summoned and totally destroyed.
Those supremely endowed with good intellect
[28] Strike with the phur bu in accordance with the rite.
The great Vajra King, the Amta being,
Abides as the Vajra Phur bu itself,
[29] Blue in colour like an utpala,
Gazing down at the hosts of obstacles.
The part below his navel
[30] Is like a point, and utterly [.....] [brlag, destroys?]
If, endowed with his mantras,
One definitively strikes with Vajra Phur [bu],
[31] The bodies of the obstacles will become quite immobilised
o gha gha gha ta ya gha ta ya / sa rva du shta ni pha // ki la ki la
[32] ya sa rva ba pham pha// h h [ba] dzra dha rod a [-] pa ya ti [...
Comments on lines 27-32:
As R.A Stein has already discussed, these verses are found also in several places in the
Guhyasamja tradition (see the appendix below). The culminating mantra does not survive
intact in P349, but is easily recognisable as a famous mantra from verse 58 of the
Guhyasamja's Chapter 14: O gha gha ghtaya ghtaya sarvadun pha klaya klaya
sarvappn pha h h vajrakla vajradhara jpayati sarvavighnn
kyavkcittavajra klaya h pha. In fact, this mantra has had a long and varied career in
Vajrayna literature, and appears in a variety of traditions beyond the Guhyasamja, for
example in the Yogin traditions of Vajravrh. This mantra is still found in the rNying ma pa

18
Vajraklaya traditions, but interestingly the deity to which it is attached can often be a female
one. In the Phur pa bcu gnyis Ch. 13 and again in Ch. 20, this mantra is given as that of rDo
rje sder mo, or Vajra claw. This is one of the special lower rite goddesses (according to Phur
pa bcu gnyis Ch. 20) who takes their orders from the central deity and perform the actual
strikes against the obstacles, in this sense closely related by function to the male deity form
with the heruka upper body and the kla lower body who here in P349 (line 28) is called rdo
rje rgyal chen bdud rtsi po, the Great Vajra Amta King. rDo rje sder mos Sanskrit name
remains unclear;28 but rDo rje sder mo under her Tibetan name also occurs in the Dunhuang
text IOL Tib J 331.111, where she also has the same mantra as appears here (O gha gha
ghtaya etc.), and the same lower-rite function as in the Phur pa bcu gnyis Ch. 20. The
Dunhuang text IOL Tib J 754, 81-82 also has the same mantra with the same function, but
does not mention the name of any deity either male or female.

[5] Appendix: some parallel Sanskrit and Tibetan texts to P349 lines 27-32 (5.1 and 5.2
prepared by Gudrun Melzer)
Piikramasdhana (PKS) of Ngrjuna:
Facsimile Edition in Mimaki 1994: A PKS 2a4-2b3
Ms. of Valle Poussin's edition
B PKS 2a3-2b1,
[5.1] Piikramasdhana
(Valle Poussin 1896, pp. 1-2)
Anena krodharpea
kyaiva vinyakn |
klayed vidhivat sarvn
prayogea tu buddhimn ||10
vajrmtamahrja
vajrakla vbhvayet |
nlotpaladalayma
jvlmlkulaprabham || 11
nbhided adhobhga
lkra vibhvayet |
rdhva krodhktii caiva
trimukhkraabhujam || 12
adho vighnagan vkya
tan mantra samudharanii |
nikhaned vajrakla tu
vighnadeheu nicalam || 13
o gha gha ghtaya ghtaya sarvadun phaiii klaya klaya sarvappn phaiv h hv
vajrakla vajradharavi jpayati sarvavighnn kyavkcittavii klaya hviii pha
[5.2]
sgrub pa'i thabs mdor byas pa (Piiktasdhana)
sDe dge vol. Ngi, 3,l.4-4,l.2; Peking 2661.
28

The rNying ma text does not attempt a Sanskrit name, although Sa skya phur chen folio 18b line 2 calls her
Vajra Tt but has here identified her as a regular member of its sambhogakya maala rather than as a special
goddess of killing in the nirmakya maala.

19

khro bo'i gzugs can 'di yis ni //

bgegs kyi dbang po nyid bkug nas //
blo dang ldan pas sbyor ba yis //
cho ga bzhin du phur bus gdab //
rdo rje bdud rtsi rgyal po che //
utpal sngon po 'dab ma'i mdog //
'bar phreng 'khrigs pa'i 'od ldan pa //
rdo rje phur bu rnam par bsgom //
lte ba'i phyogs nas smad kyi cha //
rtse mo lta bur rnam par bsam //
stod ni khro bo'i dbyibs can te //
zhal gsum phyag drug lta bur bsgom //
de yi gsang sngags legs brjod la //
bgegs kyi tshogs la 'og gzigs pas //
rdo rje phur bu nges btab na //
bgegs kyi lus ni myi g.yo 'gyur //
o gha gha gh ta ya gh ta ya / sa rva du pha pha /
k la ya k la ya / sa rva p pa pha pha h h h /
badzra k la ya / badzra dha ro dzny pa ya ti / sarva bighnn /
k ya vk ci tta / badzraix k la ya h h h pha pha
The Piiktasdhanopyikvttiratnval or mDor bsdus pa'i sgrub thabs kyi 'grel pa
rin chen phreng ba attributed to Ratnkaranti (Peking 2690: 297b l.7. to 298b l.2) contains
a slightly different version of the verses to the above. Here, the verses are broken up with
word by word commentary interspersed. Thanks to Gudrun Melzer for discovering this
passage.
/rdo rje bdud rtsi rgyal po che/
/utpal sngon po'i 'dab ma'i mdog
/'bar phreng 'khrigs pa'i 'od ldan pa/
/rdo rje phur bus rnam par bsgom/
.........
lte ba'i phyogs nas smad kyi cha/
/rtse mo lta bur rnam bsam zhing/
/stod ni khro bo'i dbyibs can te/
/zhal gsum phyag drug lta bu bsgom/
.........
de yi gsang sngags legs brjod la/
/bltas pas 'og tu bgegs kyi tshogs/
/rdo rje'i phur bus nges btab na/
/bgegs kyi lus la mi g-yo 'gyur/
.........
o gha gha gha ta ya sarba du a h pha pha /
k la ya k la ya sarba p pa pha pha h h badzra k la ya badzra dharo adzny pa ya ti/
sarba bighn k ya b ka ci tta k la ya h h h pha pha
[5.3]
gZi ldan 'bar ba mtshams kyi rgyud

20
This text of about twenty folios has no chapter divisions or titles. The text cited below is taken
from folios 274r-v of vol. Zha of the mTshams brag edition (Vol 21 pages 551-552 in the
modern pagination). It is very close to the text from the Piiktasdhana cited above in
Appendix 5.2:
/h/ khro bo'i rgyal po 'di bdag gis/
/bgegs kun bkug nas rnam par 'jig
/blo ldan rab tu 'byor pa yis/
/cho ga bzhin du phur kun btab/
/rdo rje bdud rtsi rgyal po yi/
/rdo rje'i phur bu nyid gnas pa/
/utpal sngon po'i mdog 'dra bar/
/bar ba'i 'phreng ba 'khrig pa'i 'od/
/lte ba man chad chas rnams ni/
/phur rtse lta bur rnam par sgom/
/ro stod khro bo lta bu nyid/
/zhal gsum phyag kyang drug pa ste/
/bgegs kyi tshogs la 'og tu gzugs/
/de yi sngags ni brjod bya ste/
/rdo rje phur pa nges btab nas/
/bgegs lus bzhin du mi g-yo ba'o/
/o gha gha gha ta ya sarba dustan h pha/
/kilaya kilaya sarba ppa h pha/
/h h h badzra kilaya/
/badzra darod adnya payati/
/ka ya bag cita badzra ki la ya h pha/
[5.4]
Phur pa gsang chen rdo rje 'phreng ba'i rgyud, Chapter 16
Rig 'dzin edition of the NGB, Vol. Sha folios 43v to 60r
tib ta cakra phur pa'i lha//
dmar po gcer bu ral pa can//
kun kyang khro bo chen po la//
zhal gsum phyag ni drug pa ste//
ral gri sku la phur pa'i so//
lte ba yan chad chas rnams ni//
na za rdo rje go cha gtams//
lte ba man chad chas rnams ni//
utpal sngon po'i mdog 'dra ba//
'bar ba'i 'phreng bas 'khrig pa'i 'od//
lcags kyi phur pa zur gsum pa//
btab na lha yang rlag pa'i phyir//
gnod byed dgra bgegs smos ci dgos//
k la ya/
m ra ya pha/
Phur pa'i las byang, by Grags pa rgyal mtshan
rDo rje phur pa'i sgrub skor, Sa skya bka' 'bum, vol. 4, p 182.

21
db ta tsakra phur ba'i lha/
/dmar po gcer bu ral pa can/
/sku stod khro bo chen po la/
/zhal gsum phyag kyang drug pa ste/
/ral gri'i sgra la phur bu'i so/
lte ba man chad sku yi cha rnams ni/
/utpal sngon po'i 'dab ma 'dra/
/'bar ba'i phreng ba 'khrugs pa'i 'od/
/lcags kyi phur pa zur gsum pa/
/drag po gyur pa'i phur bu ste/
/btab na yang brlag 'gyur te/
/gnod byed bgegs la smos ci dgos/
/o badzra k la ya sarba bighn ba h pha/
Notes to appendix texts:
i A, B rdhvakrodhkti ii A samudharet iii A +pha iv A +pha v A +h
vi A vajradharo vii A kyavkcittavajra viii A +h h ix Peking omits

[6] Bibliography
Tibetan Sources:
bDud 'joms 'jigs bral ye shes rdo rje: dpal rdo rje phur bu thugs gyi sgrub pa gsang ba'i rgya
can bdud 'joms gnam lcags spu gri'i las byang khrag 'thung mngon par rol pa'i dga'
ston. From the Collected Works of H.H. bDud 'joms Rin po che, vol. 10.
Grags pa rgyal mtshan: rDo rje phur pa'i sgrub skor, in Sa skya pa'i bka' 'bum vol. 4, pp. 175199. Tokyo, 1968.
'Jam mgon a myes zhabs kun dga' bsod nams: bCom ldan 'das rdo rje gzhon nu'i gdams pa
nyams len gyi chu bo chen po sgrub pa'i thabs kyi rnam par bshad pa 'phrin las kyi
padmo rab tu rgyas pa'i nyin byed, in Ngawang Sopa, ed: The Vajrakla Rites As
Practiced by the 'Khon Lineage of Sa skya, New Delhi, 1973.
Kong sprul blo gros mtha' yas: dPal rdo rje phur pa rtsa ba'i rgyud kyi dum bu'i 'grel pa
snying po bsdus pa dpal chen dgyes pa'i zhal lung zhes bya ba n.d., n.p.
rNying mai rgyud bum, The Collected Tantras of the Ancient School of Tibetan
Buddhism. gTing skyes edition: Ed. Dil mgo mKhyen brtse. Reproduced from a
manuscript copy from gTing skyes dgon pa byang monastery in Tibet. 36 vols.
Thimbu, 1973-75.
Rig 'dzin edition: original manuscripts held by the British Library, London. An
electronic catalogue by Cathy Cantwell, Robert Mayer and Michael Fischer can be
found at http://ngb.csac.anthropology.ac.uk:8080/
mTshams brag edition: The Mtshams brag Manuscript of the rNying ma rgyud
'bum, Photo-offset litho reproduction made in Delhi under the auspices of the
National Library of Bhutan. 46 volumes. Thimpu 1982.
dPal rdo rje phur pa'i bsnyen sgrub gsal byed bdud rtsi'i 'od can. From the sGrub thabs kun
btus vol Pa, p. 140 ff.
Phur pa bcu gnyis kyi rgyud ces bya ba theg pa chen po'i mdo sDe dge rNying ma'i rgyud
'bum vol. Pa, folios 176r-251v; and gTing skyes dgon pa byang monastery ms.,
Khyentse Rinpoche reprint, vol Dza 1r-100r (pp. 1-199)

22
Sa skya phur chen: dPal rdo rje gzhon nu'i sgrub thabs bklags pas don grub. Rajpur, India.
Other Languages
Bhattacharya, Benoytosh.
1968 Sdhanaml vols. 1 and 2. GOS nos. 26 & 41. Baroda.
Bischoff, F.A. & Charles Hartmann
1971Padmasambhava's Invention of the Phur-bu: Ms. Pelliot Tibtain 44. tudes
Tibtaines Ddies la Mmoire de Marcelle Lalou. Paris, pp. 11-28
Boord, Martin J.
1993 The Cult of the Deity Vajrakla. Tring
2002 A Bolt Of Lightning From The Blue. Berlin.
Cantwell, Cathy
1997 To meditate upon consciousness as vajra: ritual killing and liberation in the
rNying ma pa Tradition, In PIATS 7, Tibetan Studies, Volume 1, ed. Krasser,
Much, Steinkellner & Tauscher, sterreichischen Akademie der Wissenschaften,
Wien pp. 107-118
Chakravarti, Chintaharan, ed.
1984 Guhyasamjatantrapradpodyotanakakoivykhy. Patna.
Dudjom Rinpoche, Jikdrel Yeshe Dorje. (trans. Dorje and Kapstein)
1999 The Nyingma School of Tibetan Buddhism. Boston.
Eastman, Kenneth W.
1980 Chibetto-go Guhyasamjatantra no tonkoo shutsudo shahon. Tokyo.
Edgerton, Franklin.
1970 Buddhist Hybrid Sanskrit Dictionary. Delhi.
Fremantle, Francesca.
1971 A Critical Study of the Guhyasamja-tantra. PhD diss. SOAS, London.
Hackin, Joseph
1924 Formulaire Sanscrit-Tibtain du XE Sicle. Paris
Kapstein, Matthew.
2000 The Tibetan Assimilation of Buddhism. Conversion, Contestation and Memory.
New York.
Lalou, M.
1939, 1950, 1961. Inventaire des Manuscrits Tibtains de Touen houang conservs la
Bibliothque Nationale. Paris
Matsunaga, Yukei (ed.)
1978
The Guhyasamja Tantra. Osaka
Mayer, Robert
1996 A Scripture of The Ancient Tantra Collection: The Phur-pa bcu-gnyis. Oxford
1991 Observations on the Tibetan Phur-pa and the Indian Kla. The Buddhist Forum II.
London
Mayer, Robert & Cathy Cantwell
1994 A Dunhuang Manuscript on Vajraklaya [IOL MSS TIB J 754, 81-82]. (Reprint)
Tibet Journal XIX.1, pp. 54-67
Mimaki, Katsumi
1994
Pacakrama: Sanskrit and Tibetan Texts Critically Edited with Verse Index and
Facsimile Edition of the Sanskrit Manuscripts. Tokyo
Roerich, George N.
1976
The Blue Annals Delhi.

23
Stein, Rolf A.
1978
A propos des documents anciens relatifs au phur-bu (kla), Proceedings of the
Csoma de Krs memorial Symposium, ed. L. Ligeti, Budapest. Pp. 427-44
Scherrer-Schaub, Cristina
1999-2000 'Prire pour un apostat - Fragments d'histoire tibtaine'. In Cahiers d'ExtrmeAsie 11 (1999-2000), pages 217-246.
Takeuchi, Tsuguhito
1998
Old Tibetan Manuscripts from East Turkestan in The Stein Collection of the
British Library Volume II. Descriptive Catalogue. Tokyo and London
Valle Poussin, Louis de la
1896
tudes et Texts Tantriques, Pacakrama. Gand & Louvain
1962
Tibetan Manuscripts from Dunhuang in the India Office Library. London.
Jiang, Zhongxin & Toru Tomabechi (eds).
1996. The Pacakramaippani of Munirbhadra (Yogimanohar). Bern.

R.Mayer, April 19, 2003.