**1. Introduction**

According to its colophon, the full title of the *D. ak¯ ar¯ n. ava* ("Ocean of D. akas") ¯ is *Sr´ ¯ıd. ak¯ ar¯ n. ava-nama-mah ¯ ayogin ¯ ¯ıtantraraja ¯* (the great king of Yogin¯ıtantras named "Glorious Ocean of D. akas"). ¯ <sup>1</sup> Exegetically, the term "*D. ak¯ ar¯ n. ava*" has multiple meanings. The literal, and possibly basic meaning, is "Ocean of D. akas". In the ¯ closing verses of the chapters in the *D. ak¯ ar¯ n. ava*, the Lord or Blessed One is described as having the nature of "the complete union of all heroes" (*sarvav¯ırasamayog ¯ ad¯* ),<sup>2</sup> which is equivalent in meaning to the "Ocean of D. akas". From another perspective, ¯ in its comment on the scripture's colophon, the *Bohita¯* ("ship" to cross over the waves of reincarnation),<sup>3</sup> an extensive commentary on the *D. ak¯ ar¯ n. ava*, whose authorship is

<sup>1</sup> *D. ak¯ ar¯ n. ava*, Skt ed. (my unpublished edition), colophon, 11: *iti ´sr¯ıd. ak¯ ar¯ n. avam. nama ¯ mahayogin ¯ ¯ıtantraraja ¯ m. s. ad. arvan. adya ¯ m. yogajñanasamvarakriy ¯ atattv ¯ ar¯ n. ava\*s ¯ ¯ıta* (for *-s¯ına-* or *-sita-*) *d. ak¯ ar¯ n. avadima ¯ m. nama tulya ¯ m. tantraraja ¯ m. samaptam / ¯* ("This ends the great king of Yogin¯ıtantras named 'Glorious Ocean of D. akas', the king of Tantras, [which is] the same as the so-called 'the First of the ¯ Six Oceans', [namely] the first, the 'Ocean of D. akas', residing on the 'Ocean of Yoga', '[Ocean of] ¯ Gnosis', '[Ocean of] Restraint', '[Ocean of] Actions', and '[Ocean of] Truth'.). The *D. ak¯ ar¯ n. ava* is the first and the essence of the Six *Arn. ava* (Ocean) scriptures, viz., *D. aka- ¯* , *Yoga-*, *Jñana- ¯* . *Sam. vara-*, *Kriya-¯* , and *Tattva-arn. ava*s. Most likely, the Six Oceans are imaginary root scriptures. They are also treated as representing aspects of the *D. ak¯ ar¯ n. ava* in Padmavajra's *Bohita¯* commentary. For example, the *Bohita¯* (D 1419, 5r2–r3) states thus: *'dir yang mkha' 'gro rgya mtsho ni rjod par byed pa'o// brjod par bya ba ni rnal 'byor rgya mtsho'o// 'brel pa ni ye shes kyi rgya mtsho'o// slar yang dgos pa ni sdom pa rgya mtsho'o// de bzhin du dgos pa'i dgos pa ni bya ba rgya mtsho'o// rnal 'byor ma'i bha ga rnams ni de kho na nyid kyi rgya mtsho'o// thams cad rgya mtsho ni rtsen pa dang ldan par gyur to zhes bya ba'i don to//* ("Regarding that, (1) the Ocean of D. akas is the title of the scripture ( ¯ *rjod par byed pa*, \**abhidhana ¯* ); (2) the subject matter (*brjod par bya ba*, \**abhidheya*) is the Ocean of Yoga; (3) the connection (*'brel pa*, \**sambandha*) is the Ocean of Gnosis; (4) furthermore, the purpose of composition (*dgos pa*, \**prayojana*) is the Ocean of Restraint; (5) likewise, the purpose of the purpose (*dgos pa'i dgos pa*, \**prayojanaprayojana*) is the Ocean of Actions; and (6) the *bhaga* or wombs of yogin¯ıs constitute the Ocean of Truth. All are oceans endowed with plays. This is the meaning.).

<sup>2</sup> The expression *sarvav¯ırasamayoga ¯* is the title of one of the Saiva Vidy ´ ap¯ ¯ıt.ha scriptures (Sanderson 2009, p. 156). This is the ultimate origin of that expression.

<sup>3</sup> \**Sr´ ¯ıd. ak¯ ar¯ n. avamahayogin ¯ ¯ıtantrarajasya ¯ t. ¯ıka Bohit ¯ a n¯ ama ¯* (*Dpal mkha' 'gro rgya mtsho rnal 'byor ma'i rgyud kyi rgyal po chen po'i 'grel pa gru gzings zhes bya ba*), D 1419. In the Tibetan translation, the title of this text is *gru gzings* ("ship"), and its Indian word is transcribed as *bo hi ta¯*. A fragment of its Sanskrit manuscript (NGMPP A48/9, three damaged palm leaves, date unknown; for details, see footnote 53 in this monograph) contains a passage that mentions the title of Chapter 22, according to which the title of this work is also *bohita¯* (*iti ´sr¯ıd. ak¯ ar¯ n. n. ave mahayogin ¯ ¯ıtam. trarajye Bohit ¯ ay¯ a¯m. t. ¯ıkay¯ a¯m.* . . . , 133r3, the first leaf, of NGMPP A48/9).

The letters *b* and *v* are indiscernible, but the title is more likely to be *Bohita¯* than *Vohita¯*. Padmavajra teaches an etymology of the word *bohita¯*, where the initial sound *b* is related to \**bhadra* (*bzang po*). Although it is difficult to precisely translate the verse in Tibetan, I translate the whole etymology as follows (*Bohita¯*, D 1419, 317r6–v1): "The *Bohita¯* (ship) of that [scripture, viz., *D. ak¯ ar¯ n. ava*] is a carrier to the opposite side. The letter *ba* is to be always mindful of being in the face of good ones (*ba*=*bha* of \**bhadra*: *bzang po*). The letter *om.* (for *o*) is to absorb yourself into the sound (*na da ¯* , which in this passage may indicate the sound *om.* ) through the path of meditation. The letter *hi* is to bring benefit (*hi* of \**hita*: *phan pa*), with the purpose of benefiting both yourself and others equally. The letter *ta* (for

attributed to Padmavajra (also called Saroruha),<sup>4</sup> states that the words "D. akas" and ¯ "Ocean" indicate the same number of "heroes (Tantric males)" and "yogin¯ıs (Tantric females)" as the atoms that constitute Mt. Sumeru, respectively.<sup>5</sup> As mentioned in the Preface of this monograph, the *D. ak¯ ar¯ n. ava* belongs to the Buddhist scriptural tradition called Sam. vara or Cakrasam. vara.<sup>6</sup> The oldest scripture in the Sam. vara

<sup>6</sup> The words *sam. vara* and *cakrasam. vara* are native terms that can be found in the texts belonging to this scriptural tradition and are used to indicate aspects of this scriptural tradition. The word *sam. vara* in this scriptural tradition originally comes from the Saiva concept of ´ *´sam. vara*, "the supreme pleasure," as Sanderson pointed out (Sanderson 2009, p. 156). In that scriptural tradition in Buddhism, the word *sam. vara* has multiple meanings. Roughly, it has both meanings of *sam. vara* (from *sam. -* √ *vr. or sam. -* √ some verbal root related to it: "concealing," "protection," etc.) and *´sam. vara* (*´sam*+*vara*: "supreme pleasure" or *´sam.* √ *vr.* : "concealing pleasure"). Even when the word is used in the latter sense, the word is *sam. vara* and hardly ever *´sam. vara* in the available Sanskrit manuscripts of this scriptural tradition. (A retainer deity named Samvara or ´ Sambara can be found in this scriptural tradition, but it ´ seems rather derived from Sabara.) This is partially because ´ *´s* and *s* are phonetically equivalent in east India and Nepal, and partially because by doing so that word can have multiple meanings. For *sam* conventionally used in the sense of *´sam*, see also Bhavabhat. t.a's *Cakrasam. varavivr. ti* (Skt ed. (Bang 2019, p. 160, 3.4.2): . . . *sukham. sam iti cakhy ¯ atam iti vacan ¯ at / yo 'ya ¯ m. laukikah. sam. ´sabdah. sa talavya eva / ¯* ; " . . .

because the scripture tells thus: 'And *sam* is known as pleasure' (*Sarvabuddhasamayoga ¯* [see below] 1.10a: *sukham. sam iti vikhyata ¯ m.* ). That which is this, conventionally used word *sam*, is exactly the palatal [*´sam*].") For this, see also Pandey 2002, p. 6, l. 10–l. 11 ( . . . *yo 'yam. laukikah. ´sam. ´sabdah. sa talavya ¯ eva /*, "That which is this, conventionally used [word *sam*], is exactly the palatal word *´sam*."). For the terms *sam. vara* and *´sam. vara*, see also (Cicuzza 2001, p. 15; English 2002, p. 389, note 27; Sanderson 2009, pp. 166–69; Szántó 2015, pp. 369–70; Sugiki 2015, pp. 360–61).

Tsuda, a pioneer of this field of research, as well as some older scholars, have called this scriptural tradition Sam. vara, perhaps following historical scholarly monks in Tibet such as Bu ston, who called it *Bde mchog* or "Supreme Pleasure" (Tsuda 1974, p. 28). Tsuda and other old scholars used the word *sam. vara* and not *´sam. vara* (literally, "supreme pleasure"). This is probably because, as mentioned above, the word is normally *sam. vara* in the available Sanskrit manuscripts of this scriptural tradition.

However, in Buddhist tantras, including the *D. ak¯ ar¯ n. ava*, the word *sam. vara* is also widely used as a designation for the earlier scripture *Sarvabuddhasamayoga ¯ d. akin ¯ ¯ıjalasa ¯ m. varatantra* (abbreviated to *Sarvabuddhasamayoga ¯* , already present in some form in the early 8th century CE (Szántó and Griffiths 2015, p. 367)). In order to avoid confusion, in this monograph, I call this early scripture "*Sarvabuddhasamayoga ¯* ," and the scriptural tradition in question "Sam. vara". This is perhaps the way many modern scholars and practitioners of Tibetan Buddhism call and distinguish them. It is also

*ta¯*) is to dissolve all things into emptiness (perhaps *ta¯* of \**´sunyat ¯ a¯*: *stong pa nyid*). Protecting sentient beings out of compassion and being devoid of immaturity, the *Bohita¯* (ship) containing a commentary [to the *D. ak¯ ar¯ n. ava*] crosses over the waves of reincarnation." (*de'i gru gzings ni pha rol du 'gro bar byed pa ste/ /ba yig bzang po'i gdong rnams la/ /dus thams cad du dran pa ste/ /\*om. (*for *o) yig bsgom pa'i lam du ni/ /bdag nyid na da'i nang du thim// /hi yig phan pa skye ba ste/ /rang gzhan don ni mnyam don can/ /\*ta ( ¯* for *ta)¯ yig dngos po thams cad rnams/ /stong pa nyid du 'chad pa'o// /sems can rjes su 'dzin pa legs/ /byis pa nyid ni spong ba ste/ /gru gzings 'grel pa gang ldan pa/ /'khor ba'i dba' rlabs las brgal lo//*).

Historically, the word *bohita¯* is probably derived from the Sanskrit *vahitra* or Prakrit *bohittha* or *vohitta* (cf. PH) and appears closer to the New Indo-Aryan form of that term in the eastern areas of the Indian subcontinent, i.e., *bohit* in Hindi (meaning "ship").

<sup>4</sup> There were multiple Tantric masters who were named Padmavajra, Saroruhavajra, and Saroruha. It seems unlikely that the author of the *Bohita¯*, named Padmavajra, was identical to the famous Siddha Padmavajra or Padmavajra of the Hevajra scriptural tradition.

<sup>5</sup> *Bohita¯*, D 1419, 316v4: *mkha' 'gro zhes bya ba ni ri rab kyi rdul dang mnyam pa'i dpa' bo rnams te*/ *rgya mtsho zhes bya ba ni de'i tshad kyi rnal 'byor ma rnams so*// ("'D. aka' refers to the same number of heroes ¯ as the atoms of Mt. Sumeru, and 'ocean', to the same number of yogin¯ıs as those.").

scriptural tradition is the *Cakrasam. varatantra* (abbreviated to *Cakrasam. vara*), also called *Herukabhidh ¯ ana ¯* and *Laghusam. vara*. It was gradually composed between the late 8th and 10th century CE or between the late 9th and 10th century CE.<sup>7</sup>

There are some previous studies of the *D. ak¯ ar¯ n. ava*. S´astr ¯ ¯ı and Chaudhuri published critical editions of all Apabhram. ´sa verses in the *D. ak¯ ar¯ n. ava*. 8 I have published a Sanskrit edition and English translation of Sections 3 and 8 of Chapter 50 in the *D. ak¯ ar¯ n. ava*. <sup>9</sup> Maeda's and Tanaka's studies mentioned below are also important previous studies.

The main topic of Chapter 15 of the *D. ak¯ ar¯ n. ava* (abbreviated to the *D. ak¯ ar¯ n. ava* 15) is the practice of visualizing a large-scale man. d. ala of Lord Heruka, comprising 986 major deities. The chapter includes the following discourses:


valid to call that early scripture "*Sam. vara*" and the scriptural tradition in question "Cakrasam. vara," which is the approach I took in some of my previous papers.

<sup>7</sup> The date of compilation of the *Cakrasam. vara* is controversial. In his *Namamantr ¯ arth ¯ avalokin ¯ ¯ı*, Vilasavajra ¯ mentions an early version of the *Cakrasam. vara* or another text that he calls *Cakrasam. vara*. Davidson analyzed the appearance of the *Cakrasam. vara* to be in the late 8th century (Davidson 1981, pp. 7–8; Davidson 2004, p. 40, also Gray 2007, pp. 11–14) and Sanderson, to be around the relatively late 9th century (Gray 2012, pp. 158–65, also Gray 2012, pp. 3–19). The *Cakrasam. vara* is also mentioned in Dhammakitti's Singhalese chronicle, *Nikayasa ¯ m. graha* (14th to early 15th century), as a scripture introduced into Sri Lanka during the reign of Matvalasen (Sena I, reign 846–866) (Szántó 2012a, pp. 38–39). It is also controversial when Jayabhadra, who wrote the earliest commentary on the *Cakrasam. vara*, was active: Gray analyzed his date to be in the early-to-mid-9th century (Gray 2007) and Sanderson, to be sometime the early part of the 10th century (Sanderson 2009, p. 161). Jayabhadra, who was born in Sri Lanka (Skilling 1993, p. 179), wrote the commentary near the Mahabimbast ¯ upa ¯ on the Konkan coast (south-west India) (Szántó 2012a, pp. 40–41; Kano 2020, pp. 32–35).

The extant version of the *Cakrasam. vara* consists of 51 chapters, and it is possible to divide them roughly into two parts: (1) Chapters 1 to 49 and the first half of Chapter 50 (–50.18); and (2) the second half of Chapter 50 (50.19–) and Chapter 51. The former part constitutes the old part of that scripture. Jayabhadra, the author of the earliest commentary, as well as Bhavyak¯ırti, used some forms of that old version. The latter part was developed and added to the old part by the 10th century CE at the latest. For this, see (Sugiki 2001, pp. 91–99), which is summarized in English in (Sugiki 2010c, pp. 509–10), and (Sanderson 2009, pp. 159–60).

<sup>8</sup> (S´astr ¯ ¯ı 1916; Chaudhuri 1935).

<sup>9</sup> (Sugiki 2018a, 2018b).

<sup>10</sup> The *Hevajra* (Skt ed. (Snellgrove 1959), I.7.27) is the first to teach the meaning of the four letters *´sr¯ıheruka* or etymology of that word. The *D. ak¯ arn ¯ . ava*'s explanation of them is more extensive.


Maeda published a Sanskrit edition of about one-third of the *D. ak¯ ar¯ n. ava* 15 (which corresponds to passages 15.1–85 of the entire chapter (15.1–290) of my edition).<sup>12</sup> He did not provide an English or Japanese translation. He used only Sanskrit manuscripts that are relatively new. He did not elucidate the contents of the Heruka man. d. ala. Tanaka pointed out that the Heruka man. d. ala in the *D. ak¯ ar¯ n. ava* comprises deities from the Guhyasamaja, the Hevajra, and the Sa ¯ m. vara traditions. However, his description and analysis of the Heruka man. d. ala in the *D. ak¯ ar¯ n. ava* are brief. He did not clarify the names of most deities constituting the man. d. ala.<sup>13</sup> Thus, the Heruka man. d. ala in the *D. ak¯ arn ¯ . ava* 15 has not been studied sufficiently.

As I mentioned in the Preface, this monograph provides the first critical edition and an English translation of the Sanskrit *D. ak¯ ar¯ n. ava* 15 and elucidates its structure, form, and significance. The text of its Tibetan translation (translated by Jayasena and Dharma yon tan) is also provided here as supporting material. I also present the first critical edition and English translation of the Tibetan text of Jayasena's *Ratnapadmaraganidhi ¯* , the oldest and extensive manual for visualizing the Heruka man. d. ala taught in the *D. ak¯ arn ¯ . ava* 15.

The rest of this monograph is organized as follows. Chapter 2 investigates when and where the *D. ak¯ ar¯ n. ava* was compiled. Chapter 3 explains the materials used for the edition and translation, peculiarities of the language and meter found in the edited chapter, and editorial policy. Chapter 4 analyzes the structure, form, and meanings of the Heruka man. d. ala and clarifies its significance in the history of Buddhism. I provide a detailed and organized description of the man. d. ala's structure and form because it serves as useful material for the iconographic study of extant paintings of this man. d. ala (which is a task for my future research). These chapters constitute Part 1. Part 2 consists of Chapters 5–7. Chapters 5 and 6 present a critical edition of the Sanskrit *D. ak¯ ar¯ n. ava* 15 and its English translation, respectively. Both the critical edition and English translation are annotated. Chapter 7 presents a text of

<sup>11</sup> For the various discourses on holy sites in the Sam. vara tradition, including the discourse in the *D. ak¯ ar¯ n. ava*, see (Sugiki 2009). See also (Sanderson 1995), which is an epoch-making paper on the same topic.

<sup>12</sup> (Maeda 1995a, 1995b) presents a description of the constituents (circles and deities) of the part of the man. d. ala taught in the text that he edited in (Maeda 1995a).

<sup>13</sup> (Tanaka 1987, pp. 229–31; Tanaka 2003, pp. 378–79). He analyzed that the Heruka man. d. ala comprised three layers. However, as clarified below, it comprises four layers.

its Tibetan translation by Jayasena and Dharma yon tan. Part 3 consists of Chapters 8 and 9. Chapter 8 explains the materials used for the edition and translation of the *Ratnapadmaraganidhi ¯* and summarizes the contents of that text. Chapter 9, the final one of this monograph, provides a critical edition, English translation, and annotations of the Tibetan *Ratnapadmaraganidhi ¯* .
