No More Learning

Let us here stop this discussion of these infantile theories.
***
What is the upper limit of the destructions?
lOOc-lOld Three Dhyanas, beginning with the second, are, in order, the top of the destructions; by reason of the community of the nature of destructions and the vices of the first three Dhyanas. As for the Fourth Dhyana, no destruaion, by reason of its non-agitation: this is not that it is eternal, for its mansions are produced and perish along with the beings who live therein.
1. The Second Dhyana is the limit of the destruaion by fire: everything below it is burned; the Third Dhyana is the limit of the destruaion by water: everything below it is dissolved; the Fourth Dhyana is the limit of the destruaion by wind: everything below it is dispersed. "The top of the destruaion" is the name of that which remains when the destruaion is finished.
The heaven of the First Dhyana then perishes by fire: in faa, the vice or imperfeaion of the First Dhyana is vitarka-vtcdra\ these burn the mind and are thus similar to fire. The Second Dhyana perishes by water: in faa, it has joy for its vice. This, being associated with physical well-being, rends the body soft and flabby: it is similar to water. This is why the Sutra teaches that the sensation of suffering is destroyed by the
568
suppression of all solidity of the body. The Third Dhyana perishes by
wind: in faa, it has inbreathing and outbreathing, which are wind, for its vice.
You say that, when an unfired clay
? The external vices of a Dhyana (that is, the calamities which destroy
a Dhyana heaven) are of the same order as the internal vices which
affect the person who has entered into this Dhyana. (See viii. French
569 trans, p. 126).
2. Why is there no destruction by the earth element as by the fire element, etc?
What one calls the physical world is earth, and as a consequence can be opposed by fire, water, and wind, but not by earth itself.
3. The Fourth Dhyana is not subject to destruction, because it is free
from agitation. The Buddha said in fact that this Dhyana, being free 57
from internal vices, is non-movable (dnejya). ? The external vices have thus no hold on it and, as a consequence, it is not subject to destruction.
According to another opinion, the non-destruction of the Fourth Dhyana is explained by the force of the Suddhavasakayika gods whose abode it is. These gods are incapable of entering into Arapyadhatu, and
571 are also incapable of going elsewhere [to a lower sphere].
The receptacle world of the Fourth Dhyana is not eternal, but by this fact, it does not constitute a "sphere;" like the stars, it is divided into diverse residences; these different mansions, the abodes of beings, arise and perish with these beings. (Vibbdsd, TD 27, p. 692bl9).
***
What is the order of succession of these three destructions?
102. Seven by fire, one by water; and when seven destructions by
water have thus taken place, seven by fire followed by the
572 destruction by wind.
After seven series of seven successive destructions by fire, de-
struction is then done seven times by water; the eighth series of seven
destructions by fire is followed by one destruction by wind. The physical
worlds, in fact, last by reason of the duration of the gods who are reborn
therein by the force of their absorptions. There are fifty-six destructions
by fire, seven by water, and one by wind: this is what justifies the
575
declaration of the Prajnapti kalpas. (iii. 80d).
that the Subhakrtsnagods live sixty-four
***
The World 495
? 1. According to Buddhaghosa (AtthasdUni, 62), Kamadhatu is made up of four bad realms of rebirth (hellish beings, animals, Pretas and Asuras, see iii. 4a), humans and the six classes of gods: in all eleven divisions (padesa).
The six gods of Kamadhatu (Burnouf, Introduction, 603,608; Hastings, article, "Cosmogony and Cosmology"; names explained in the Vydkhya) are an old list, sometimes reduced to five names (Trayastrimsas . . . Paranirmkavaiavartins subject to Mara, Samyutta, i. 133). [Each category has chiefs or a king, Anguttara, iv. 242. ] Above there are, in Digka, i. 215, the Brahmakayikas and Mahabrahma; in Anguttara, i. 210, the Brahmakayikas and the higher gods; in Mahdniddesa, 44, the Brahmakayikas. (See below, note 4).
The Vydkhya explains the words naraka, etc First opinion: from the root nr (naye, Dhdtupdfha, i. 847), "Beings are led there by transgression"; second opinion: from the root r (gatiprdpanayoh or gatmhsanayos [thereadingoftheVydkhyd]tL9&3,vi. lll)preceededbyanegation;thirdopinion: from ram, ranj, "beings do not rejoice therein"; fourth opinion, that of Sarhghabhadra, "Beings do not obtain (f = prop) protection there. "
2. Vydkhya: All of the Dhyanas belong to the three realms, corresponding to weak, medium, and strong absorptions: thus the Fourth Dhyana contains the stage of Anabhrakas, Punyaprasavas, Vrhatphalas. But a strong absorption of the Fourth Dhyana can be mixed with an andsrava absorption (as explained in vi. 43); from whence there are five new places: Avrhas, etc The Fourth Dhyana is thus made up of eight stages.
3. We are reminded of the four bhavOppattis of Majjhima, iii147: the Parittabhas gods, the Appamanabhas gods, the Samkilitthabhas gods, and the Parisuddhabhas gods.
4. ThisistheopinionoftheBahirde^akas(F^^)orthePakatyas(F^4r^ TD27,p. 509a22), "the Westerners. " This refers to the masters in Gandhara. In this country, there are some Sautrantikas, but when the Vibhdsd speaks of Pascatyas, it means the extra-Kashmirian Sarvastivadins, or those of Gandhara.
The Mahdvyutpatti has four names for the First Dhyana: Brahmakayikas, Brahmaparisadyas, Brahmapurohitas, and Mahabrahmanas. From whence, according to Georgy, and Hodgson, there are four distinct heavens; Remusat and Burnouf {Introduction, 608) have discussed their various sources. In the Kola the Parisadyas are omitted, and Brahmakayikas are the inferior class. Elsewhere (iii. 5a, etc) Brahmakayikas is the generic name for all the gods of the First Dhyana, of all of gods of the world of Brahma.
Vydkhya: He is called Brahma, because he is produced by broad, large (brhat) roots of good Who is he? He is called Mahabrahma. He is great, mahdn, because he has attained the intermediate dhyana (viii. 23); because he is born and dies before and after all others (iii. 6a, Para. 2); and because he is distinguished by his stature, etc. They are called the Brahmakayikas, because the kayo, that is to say, the nivdsa, or dwelling, of Mahabrahma is theirs (tasya kayo nivdsa esam vidyate). Some are called Brahmapurohitas, because Brahma is placed {dhtyate) at their head ipura esam). And some are called Mahabrahmanas, because dyurvarnadibhir viiesair mahdn brahmd esam.
The Vydkhya explains the names of the gods in very clear language (Cosmologie bouddhique, p. 119). Samgrabhadraisinagreement. Fordetailsonallthegods,seeiii. 64.
5. Vibhdsd, TD 27, p. 509a22.
a. Pali sources. Rupadhatu (or world of Brahma, see above, note 1) contains sixteen places: 1. -9.
three places for each one of the first three Dhyanas; 10. -11. the Vehapphalas and Asafinasattas for the Fourth Dhyana; 12. -16. five Suddhavasas, reserved for the Anagamins. Variants: Majjhima, i. 329; iii. 147, etc
b. The "correct opinion of the Sarvastivadins" admits sixteen places, by assigning only two places to the First Dhyana.
c The masters of the West (the Vaibhasikas of Gandhara, Piscatyas, and Bahirdes'akas) admit seventeen places, by assigning three places to the First Dhyana (a special place for the Mahabrahmanas).
Footnotes 491
? 498 Chapter Three
d. Some other Bahirde? akas (Kola, ii. 4ld) admit seventeen places, by assigning two places to the First Dhyana, and a special place to the Asarhjnisattvas in the Fourth Dhyana.
e. Eighteen places, by assigning three places to the First Dhyana (a special place for the Mahabrahmanas) and a special place to the Asarhjnisattvas.
This is the opinion of &rilabdha 0r$-h-to, Watters, i. 355, the author of a Sautrantika Vibhdsdfastra), according to P'u-kuang and Fa-pao who quote Saihghabhadra according to whom: "the Sthavira admits eighteen places". This is the opinion of the Mahayana Sthiramati or Saramati (TD, 31, number 1606, Grand Vehicle).
f. The Yogacaras also admit the number of eighteen, but they place the Asarhjnisattvas in the heaven of the Vrhatphalas, and obtain the number of eighteen by recognizing the MahesVaradevas (compare Mabdvyutpatti, 162. 7, Mahamahesvarayatana).
Yuan-hui and other commentators say that the School (nikaya) of the Sthaviras admit eighteen places, wrongly interpreting the expression "the Sthavira" of Saihghabhadra, (a) because, among the twenty schools, there is a School of the Sthaviras (but Ts'e-yuan, P'u-kuang and Fa-pao say that "the Sthavira" of Saihghabhadra is Srllabdha); (b) because they do not know that eighteen is the number of the Sautrantika &rilabdha, whereas seventeen is the established number in the Sautrantika system. The Vibhdsd attributes the opinion of seventeen places to the Western Masters, the Gandharans. Among these masters, there are some Sautrantikas, but many are Sarvastivadins. The Vibhdsd designates a divergent opinion of the Sarvastivadin school by the expression "Western Masters", and not the Sautrantikas; this is why Saihghabhadra says only, "There are others who say (that the places are seventeen in number). . . " and does not say "another School" . . .
Saeki gives, in resume, "Sixteen places: the correct opinion of the Sarvastivadins. Seventeen places: a. Westerners (separate Mahabrahmas) and b. some other masters (separate Asamjfiins). Eighteen places (by distinguishing the place of Mahabrahma from that of the Asamjfiins): and the Sthavira (a) ? rllabdha, a Sthavira not included among the twenty schools, who is also termed a Mulasautrantika; and (b) the Sthavira, who is included among the twenty schools, and who is termed the Mulasthavira. Further, Sthiramati (Saramati) (and the Yogacarins). "
6. This "dwelling" is dhydndntarikd, Koia, ii. 41d
7. I believe that I correctly understand Hsiian-tsang and Paramartha (i-sheng so chU ~"^ff\Jifwhichshouldbecorrectedtoreadi-cbu^'H); buttheKosa,inmanyplaces,speaksof some "Great Brahmas", Mahabrahmanas: these are the Mahabrahmadevas of the Pali texts, the followers or the court of the one, sole king (see Ko/a, vi. 38b); the Vydkhyd explains their name:
"They are Mahabrahmanas because Brahma is the greatest among them by his duration of life, his color, etc"
8. Beal, Catena, 94, "Following the Kosha Shaster, it is said that Brahma has no distinct abode, only in the midst of the Brahmapurohita Heaven there is a high-storied tower, and this is (the abode of Brahma). "
9. The problem as to whether there is r&pa in this dhdtu or not, is discussed viii. 3c
10. As we will see iii. 41, the mind and mental states, in the first two Dhatus, are an dfrita (a thing supported) which have for their dfraya (support) a body endowed with organs. The mind "falls", or dies, when the organs are destroyed On nikaya - nikdyasabhdga, ii. 41, see below iiL7c, nikdyasabhdga = upapattydyatana.
11. Hsiian-tsang corrects: The mental series of material beings is not supported on these two, because they lack any force: they are strong among non-material beings because they proceed from an absorption from whence the notion of matter has been eliminated But why not admit that the mental series of non-material beings is directly supported on this absorption? Why is there an intermediate support? Furthermore, the "genus" and vital organ of non-material beings is supported by matter. . .
12. See ii. 14, viii. 3c
? 13. Vibhasd, TD 27, p. 387b26-27: As we have prthivUhdtu, etc
14. This explanation holds for the expression drupyadhdtu: but when one speaks of the non- material absorptions (viii. 2), drupya should be understood as equivalent to rupa, or rather as signifying "favorable to Arupya[dhatu]" (Vydkhyd).
15. Samyukta, TD 2, p. 199a6. The first stanza is quoted in Anguttara, iii. 4l 1, where it is attributed to a deity (devatd). In Sanskrit as in Pali it begins with na te kdmdydnidtrdniloke/ samkalpardgah purusasya kdmah. The two other stanzas, to my knowledge, are absent in the Pali. The Vydkhyd gives the last two lines.
The argument of the Ajivika does not hold He thinks that a Bhiksu ceases to be a Bhiksu if he is a kdmopabhogin; and, if kdma = desire, then the Bhiksu would be kdmopabhogin, and would lose the quality of Bhiksu, even when he does not enjoy these objects. But the Bhiksu, through desire, corrupts his morality and his precepts, without losing the quality of Bhiksu: by this he transgresses by his body or speech the boundary of the rules of the Tathagata.
In Kdthdvatthu, viii. 3-4, the Theravadins establish against the Pubbaseliyas that kdma, in the expression Kdmadhdtu, does not signify "object of desire", as rupdyatana or "visible matter",etc, but rather "desire. " They quote the stanza of the Anguttara, iii. 411 and Samyutta, i. 22, as does Vasubandhu.
Compare Vibhanga, 256 quoted in Atthasdlint, 164-5; the distinction between the vatthukdmas and the kilesakdmas, Mahdniddesa to Suttanipdta, 766; Compendium, p. 81, n. 2.
16. Some dharmas are produced (samuddcarantt) in Kamadhatu which are of the sphere of Rupadhatu and Arupyadhatu, for example the different absorptions (viii. l9e); a person can experience "desire" (rdga) with respect to these absorptions: but as this desire has a dharma of a higher sphere for its object, it does not lodge in him: like a foot on a burning paving-stone (v. 2,39). So too among the beings of Kamadhatu some andsravas, or "pure" dharmas are produced, namely the mental states which constitute the Path: these dharmas are not the object of "desire" {rdga) in any Dhatu (v. 16, viii20c): they are thus adhdtupatita or adhdtvdpia, foreign to the Dhatus.
From whence the principle that it is "thirst" which determines the spheres (Kamadhatu = one sphere, Rupadhatu = four spheres); viii20c
All dharma with respect to which the "thirst" of a being in Kamadhatu develops is also of the sphere of Kamadhatu.
17. Seevii. 49-51.
18. There are four anantas: dkdso ananto, cakkavdldni anantdni, sattakdyo ananto, buddhandnam
anantam (Atthasdlint, 160).
19. P'u-kuang (quoted by Saeki, viii. 5b) says, "If one follows the opinion of the Mahilasakas, there is an arising of new beings (yu shih ch'iyu ch'ing ^#&fe^frfpf = asty ddyutpannah sattvah) which do not arise from action-defilement (karmakleia)".
20. Vibhasd, TD 27, p. 966bll-27, mentions two opinions, a horizontal disposition, and a horizontal and vertical disposition of the universe, and the difficulties that they present. Compare Digha, i. 33.
21. This Sutra is Samyukta, 34.
7; it is used in the Lokaprajridpti, beginning (trans, in Cosmologie bouddhique, p. 196).
22. The Isadhara, "the cloud the drops of which have the dimension of a pole," is one of the four clouds at the beginning of the cosmic age, Siksdsamuccaya, 247', Koia, iii. 90c (A "rain" Naga of Waddell,JRAS, 1894,98. )
23. The opinion of the Dharmaguptakas according to P'u-kuang. The opinion of Sthiramati, Tsa-chi (TD 31, number 1606), etc
Footnotes 499
? 500 Chapter Three
24 For example, Samyukta, TD 2, p. 120b26, Madhyama, TD 1, p. 493b8.
25. On the disposition of the universe, see the references in Hastings, article "Cosmology", 137b
(Mahavastu, i. 122; Lotus Sutra, Chap, xi; Avatamsaka). See below iii. 45,73.
26. Are there five or six gatis?
Kathdvattbu, villi. RegardlessofMajjhima, 173 (pafcakbopan'imasariputtagatayo. . . ), the
Andhakas and the Uttaripathakas maintain that the Asuras constitute a separate gatihut the Kalakafijakas are placed among the Pretas, and the Vepacitti (Samyutta, 1221, Dialogues, 11280, Brethren, 749) among the Devas. (According to the version and the note of Aung-Rhys Davids. )
Furthermore Majjhima, L73; Digha, HL234, Anguttara, iv. 459, Samyutta, vi. 474.
Yet there are four apdyas or painful realms of rebirth: the damned, animals, Pretas and Asuras (References: Rhys Davids-Stede). These are the aksanagafis of the Siksasamuceaya, 147, the akkhanas of the Digha, iii. 264, and the duggatis of the Petavatthu, which are lacking in the paradise of Amitibha (SukhaSvati, 11).
In the Lotus Sutra, we sometimes have six gatis (Burnouf, 309), sometimes five (Burnouf, 377).
Nagarjuna's Letter to a Friend gives five gatis; the same for the inscription of Bodh-Gaya (Fujishima, Mr. 1888, ii. 423; Chavannes, RHR. 1896, ii).
But six gatis In the Dharmasamgraha, 57, and other sources named ibidem.
Notes of Kyokuga Saeki i. The Asuras 1. are included among the Pretas {Vibhasa and Sambhinnahrdaya, TD28, number 1552); 2. are not Included among the gatis (Buddhabhumi,etc); 3. are a sixth gait (Mahasamghikas, etc); 4. are Included among the Pretas and the animals (SaddharmasmrtyupasthanasMra); and 5. are included among the Pretas, the animals and the gods (Sagathasutra). H. TheSutrasaysthattherearefivegatisMowcanonemaintainthattherearesix? Some five centuries after the Buddha's demise, there were many schools; the schools were not in agreement; some held to five gatis, others to six. The first make the Sutra say that there are five gatis,theothersthattherearesix. iii. IntheMahayana,theAvatamsakasaysthattherearesixgatis.
27. See below note 37.
28. Definition of gati in Karanaprajflapti, translated in Cosmologie bouddhique, p. 345. We see there that ndrakanikdyasabhagatd, narakdyatanasamanvagama, and ankrtavya^rtanarakarupadmi as well as ndrakesu pratisamdhih are narakagati.
29. This Sutra, discussed in Vibhasa, TD 27, p. 309bl5, is the Saptabhavasutra. The schools which deny intermediate existence contest its authenticity, see Samghabhadra, translated in Nirvana, 1925, p. 23, note.
30. Vasubandhu means: Only the Ka? mirians read this Sutra; this Sutra is muktaka, that is, it does not form part of the Agamas.
According to Vibhasa) TD 27, p. 865al8.
31. If the gatis are undefUed-neutral dharmas, the Prakarana should say how these dharmas are abandoned through meditation, 'The anuiayas abandoned through meditation and the universal anuUyas are active and develop with respect to the gatis" (see L40c and Chapter V); it should not say, "all the anuiayas", for among them there are those that are not active in undefiled-neutral dharmas.
32. This is the explanation of the Vibhdfd, TD 27,371bl6.
Samdhicitta =pratisamdhiatta=upapattibbavaoftheKo/a,iii. 13c,38. Wehavepratisamdbi=
vitffiana, the element of "reincarnation", Pafisambhiddmagga, i, p. 52; pratisamdhickta and virffiana, Visuddhimagga, 548,659.
33. According to iii. 38, ii. 14.
34. Paramartha (TD 29,p. 199c7) differs: "You say that karmabhava is mentioned here because it is the cause of the realms of rebirth: it would be fitting also to mention the skandhas which are also the cause of the realms of rebirth. "
? 35. One could say that intermediary existence is a realm of rebirth, even though it is mentioned separately in order to indicate that it is the access to other realms of rebirth.
36. drtipya na gatih syui cyutideia evotpdddt / drupyagd biyatra cyavante vihare vd vfksamUle vd ydvac catmthydrh dhydnabhOman tatraivotpadyante: "In whichever place beings die, beings 'who should go to the arupyas', (artipyaga), either in a monastery, at the foot of a tree, or in the sphere of
the Fourth Dhyana, are reborn in this place from an dkdidnantya existence, etc" (See above iii. 3b).
37. Theundefiled-neutrai^rffWfctf canbeeithervipdkajaoraupacayika(137,ii. 57). Samghabhadra adopts the opinion of the second masters.
We see ii. 10 that tbefivitendnya is exclusively retribution, but that the five material organs,-- the mental organ, and the four sensations--are sometimes retribution, sometimes not retribution.
38. Vibhdsd, TD 27, p. 707a23-25.
The Sutra (Mahdvyutpatti, 119. 1-7) has: \. r&pi^ santi sattvd ndndtvakdya ndnMvasamjfihw
tadyathd manusyd ekatydi ca devdh, 2. . . . ndndtvakdya ekatvasamjfisms tadyathd devd brahmakdyikdh prathamdbhintrvrttdh, 3. . . . ekatvakdyd ndndtvasamplinas tadyathd devd dbhdsvardh,4. . . . ekatvakdydekatvasamjn^nastadyathddevdSubhakrtsnah, 5. dkdfdnantydyatanam (better: -dyatanopagdb), 6. vi/ttdndnantydyatanam, 1. dkimcanydyatanam. .
Digha, ii. 68 (iii. 253, 282, Anguttara, iv. 39, v. 53), "There are seven vnttidnatphitis and two dyatanas (which are joined to the vijndnasthitis in order to make nine sattvdvdsas, Kofa, iil 6c): 1. santi Ananda sattd ndnattakdyd ndnattasaMino seyyathdpi manussd ekacce ca devd ehacce ca vinipdtika. . . 2. sattd ndnattakdyd ekattasaMino. . . and as above to subhakitmd; 5. santi Ananda sattd sabbaso rJipasafinanamsamrtikkamapatighasafiTianam atthagamdndnattasaMdnam amana sikdrd ananto dkdso ti dkdsdna^cdyatanupagd; 6. . . . v&fiandnaricayatana&paga; 7. . . dkhncoMdyotanupagd.
39. The gods of the First Dhyana are, according to the system of the Foreign Masters (bahifdefakanaya): 1. the Brahmakayikas, 2. the Brahmapurohitas, 3. the Mahabrahmans (mahabrahmanaf ca). For the Masters of Kas'mir, the Mahabrahmans do not form a "place" (sthdna) distinct from the Brahmapurohitas (as we have seen, p. 366).
The vimpdtikas of the Pali text are missing.
40. The Vibhdsd (707a) omits the "gods of the First Dhyana," because these gods are not necessarily different in ideas (note of Kyokuga Saeki).
Hsiian-tsang and Paramartha translate prathamdbhinirvftta as those who arise at the beginning of the kalpa; the Lotsava has, literally, dan pot byun ba.
41. We follow the version of Paramartha. Vasubandhu (as the Lotsava and Hsiian-tsang) explains the expression ndndtvakdya as "difference of ideas because their ideas are different; endowed with this difference, they have different ideas. " On ndnMvasamjfia, the references of Rhys Davids-Stede and of Franke, Digha, p. 34, a8.
42. The Brahmakayikas are all the gods of the First Dhyana; by naming the first ones, one names the others.
43. Compare Digha, i l 8 , iiL29. Brahma thinks, "maydime sattd nhnmitd. . . "; theother gods think, "imind mayath bhotd brahmund nimmad. "
The Tibetan can be translated, "The aspect of the idea not being different, they are of the same idea. " Paramartha is very dear: "Because they have the similar idea that Brahma is their sole cause. " Samghabhadra refutes an objection: 'Their ideas are different, since the Brahmakayikas think that they are created, whereas Mahabrahma tliinks that he creates. " In faa, he says, both have the idea of a single cause, the idea of creation (nirmdna).
44. Height = aroha (uttaratd); greatness, breadth = parindha (sthantyapramdna); body = dkftivigraha, that is to say vigraha consisting of dkrti, or "figure", and as a consequence equivalent to fartra. [There is, furthermore, vedandvigraha, "mass of sensation", etc]; speech = vdgbhdsd (vdguccdrana): Hsiian-tsang and Paramartha translate: voke^ound (~vdg bhdsah).
Footnotes 501
? 502 Chapter Three
45. The Digha differs. We have the formula for Brahma's desire, " . . . oho vata atffie pi sattd itthattam dgaccheyyun ti," and the thought of the gods, "imam hi mayam addasdma idha pafhamam upapannam/ mayampan'amhdpacchdupapannd. ""forwehaveseenhimariseherebeforeus,and as for us, we have arisen after him. "
46. P'u-kuang (TD4l,p. 153a20) says that there are three answers to this question; Samghabhadra (TD 29, p. 462c20) mentions six responses; and the Vibhdsd (TD 27, p. 508c3) gives five answers of which Vasubandhu reproduces the first three.
47. Intermediate existence can be prolonged only when the intermediate being should be reincarnated into Kamadhatu, iii. l4d
48. According to the Vydkhyd
Paramartha understands: "The gods remember the past in the world of Brahma; they formerly
saw [=in a preceding existence in the world of Brahma] Brahma of long life and lasting for a long time; later, they see him anew; and as a consequence they s a y . . . "
Hsiian-tsang: "The gods remember the past of this being in this world itself; they have seen him previously. . . "
49. Destruction of the universe by fire, iil90a-b, 100c-d.
50. See viii. 4.
51. The Vydkhyd: paribhidyate'neneti paribhedah.
52. Vasubandhu reproduces the seventh of the eight explanations of the Vibhdsa, TD 27, p. 708al3.
53. Addition of Hsiian-tsang who follows Samghabhadra (whom the Vydkhyd quotes ad iii. 7a), "The Aryans who are in the first three heavens of the Fourth Dhyana, desire to enter among the Suddhavasas (last five heavens of the Fourth Dhyana) or into Arupya; and the Suddhavasas desire Nirvana.