. 891 śaka chatthī roc· śrāvana sanaścaravāra nu mrateñ· ṇnirantarācāryya khloñ· vnaṁ kaṁmrateṅ· jagat· liṅgapura jau bhūmi vrai travaṅ· khsac· ṇā saṅkuti ti pūrvva Issa vrai canhvar· raṁcyak· vrai sastā chdiṅ· chok· khlā thleṅ· dau lvaḥ ta gi vrai cassa nuva chdiṅ· hara thleṅ· dau lvaḥ ta gi vrai cassa nuva gnaṅ· dhār· ti pūrvva prasap· chdiṅ· lṅyaṅ· ti jau nu bhājana ta vraḥ kaṁmrateṅ· Aña ta rājya nu vraḥ kaṁmrateṅ· Aña ta sak· vrāṁhma ṅana bhājana jyaṅa 40 nu thve pūjā vraḥ kaṁmrateṅ· Añ· suvarṇṇa. Oy yajña ta vraḥ kaṁmrateṅ· Añ· ta rājya liḥ II ta vraḥ kaṁmrateṅ· Añ· sak· vrāhma liḥ II pratidina. pādamūla ta phjauva bhūmī sastā chok· khlā dhār· pi jau. vāp· vrau vāp· vadeva vāp· hari devaparicāra vāp· Āditya grāmapāla vāp· mādhava grāmapāla vāp· śivagaṇdha vāp· sīr· vāp· se vāp· yogipāla vāp· Āt· vāp· candra vāp· pañ· vāp· Amṛta khloña vnvak· vāp· vurūna vāp· vrahma taṁmraduṅ· kaṁlūṅa vāp· nagiya cāṁ phjuḥ vāp· Ap· cāṁ phjuḥ

891 861 ṇirantarācāryya The distinction i/ī is not made in this inscription. The spiral form would invite to read ī, but this reading would occasion many errors, therefore we shall systematically restore the correct spelling. Note that cacuminal is also used in place of dental in the next line where ṇā should be corrected to nā. canhvar· canhur cyak Claude Jacques assumes that this is an original form of canhvar, "channel" (404191), but the reading canhvar seems more correct to us despite the small size of the va subscript loop. Note that the ra in canhvar has two hastings – a rather pre-Angkorian usage – while raṁcyak has only one. raṁcyak· cyak chdiṅ· cdiṅ We are tempted to read cdiṅ, as Claude Jacques does, but besides the fact that the consonant ca is more of a pre-Angkorian form (Jenner 2009, s.v.), it would then be very different from the other occurrences of this inscription. It must be recognized, however, that this cha is also dissimilar to that of the word chok that immediately follows it. Of the three occurrences of chdiṅ, only the second, at the end of line 4, has a perfectly traced cha. The other two follow another ductus that might lead us to read ca. This term is the origin of the modern ស្ទឹង; we translate it here as "river," but it may refer to relatively small streams (Headley et al. 1997, s.v.: "small river, stream"). chdiṅ· chdin chdiṅ· cdiṅ vrāṁhma ṅana vrāṁhnaṅana 40 20 The upper loop of this character is obviously an additional bar that imposes the reading 40 (Soutif 2008, p. 60). thve phle suvarṇṇa suvarṇna The subscript (na/ṇa) is difficult to decipher, but the subscript evokes a small ṇa resembling a small sinusoid that is encountered elsewhere, although quite rarely (cf., e.g., K. 1198, face A, l. 5; EFEO stamping no. 1654). vrāhma vrāhmaṇa phjauva pāñjauva The reading pāñjauva given by Claude Jacques (l. 9-10) seems to be erroneous. He notes: "the form of the ñ can only be guessed at because of the subscribed ja. This word seems to be a hapax: it is clearly derived from jau, preceded by the causative pa" (404202). While the reading of ja is certain, the other characters seem more doubtful to us. The last akṣara of line 9 is clearly a ta and the first of line 10 a pha: the right haste of this character clearly returns backwards in a downward loop. It should therefore be read "ta phjauva". Only one other occurrence of this term is known, but with a different spelling, phjov, in the unpublished inscription K. 1186 (face A, l. 18; EFEO rubbing no. 1684). In any case, it is indeed a causative derivative of jov (~ jau ~ jauv ~ jo ~ jā), meaning "to buy" (Jenner 2009, s.v.). vadeva As Claude Jacques (404204) suggests, it seems likely to us that the first syllable of this anthroponym has been omitted (śivadeva, navadeva, śravadeva?). devaparicāra harideva paricā I The reason behind this reading lies in the fact that, as in the case of canhvar (l. 3), a double-hastened ra is used for the akṣara bearing the vowel i, whereas the second one has only one haste and can therefore be confused with a daṇḍa. sīr· līr The position of the vowel above the left loop and not above the right haste imposes this reading. yogipāla yogapāla candra candu Although the subscribed da is difficult to decipher and the subscribed ra does not go up on the left – for lack of space, presumably –, the reading candra seems more plausible. pañ· vāp ña The second pa – which is subscripted – is discrete and merges with the virāma of the vāp- at the end of line 13, but it seems clear to us. vurūna Claude Jacques suggests correcting to varuṇa, although, according to him, this name is not attested in the corpus (405210). We have however found several occurrences as anthroponyms (for example si varuṇa, K. 231, l. 6 ; 888 śaka; 372). This correction therefore seems the more reasonable to us. taṁmraduṅ· kaṁmraduṅ Claude Jacques suggests correcting to kaṁmrateṅ. The ka is doubtful and looks like ta, but this correction seems correct because of the lack of interpretation of the hapaxes (k/t)aṁmra(du)ṅ and (k/t)aṁmra(tu)ṅ. Presumably, the Vāp Varuṇa and Vrahma occupied a function "within (kaṁlūṅ) [the temple enclosure]," as in the inscriptions K. 89 (khloñ vala kaṁluṅ kaṁveṅ, "chief of the population inside the enclosure"; l. 11; 924 śaka; 3166-167) or K. 134 ('nakk paṁre kaṁluṅ vraḥ, "servant in the temple"; l. 10; 703 śaka; 293-94).

891 śaka, sixième jour de la lune décroissante de la mansion lunaire śrāvana, un samedi.

Alors, le Mrateñ Nirantarācārya, chef du temple du Kaṁrateṅ Jagat de Liṅgapura, a acheté le domaine forestier bhūmi vrai de Travang Khsac « le bassin sableux » à Saṅkuti. À l’est, il se termine à Vrai Canhvar Raṁcyak « la forêt du ruisseau des pandanus ». Il comprend la forêt Sastā, la rivière chdiṅ de Chok Khlā « bosquet des tigres » en montant jusqu’à l’ancienne forêt vrai cassa et la rivière Hara, en montant jusqu’à l’ancienne forêt et le gnaṅ de Dhāra, bordé à l’est par la rivière Lṅyaṅ. Tout cela a été acheté avec un plat, pour le Vraḥ Kamrateṅ ’Añ ta Rājya et le Vraḥ Kamrateṅ ’Añ ta Sak Brāhma ; le plat pesait 40 jyaṅ et il a fait la pūjā du Vraḥ Kamrateṅ ’Añ Suvarṇa. Il donne en sacrifice chaque jour 2 liḥ au Vraḥ Kamrateṅ ’Añ ta Rājya et 2 liḥ, au Vraḥ Kamrateṅ ’Añ Sak Brāhma.

Le vénérable a fait acheter les terres de Sastā, Chok Khlā et Dhār. Pour les acquérir : Vāp Vrau, Vāp Vadeva, Vāp Harī, serviteurs du dieu devaparicāra, Vāp ’Āditya, gardien de village grāmapāla, Vāp Mādhava, gardien de village, Vāp Śivagaṇdha, Vāp Sīr, Vāp Se, Vāp Yogipāla, Vāp ’Āt, Vāp Canda, Vāp Pañ, Vāp ’Amṛta, chefs d’équipe khloñ vnvak, Vāp Varuṇa, Vāp Vrahma, Kamrateṅ à l’intérieur du temple, Vāp Nagiya, cāṁ phjuḥ et Vāp Kap, cāṁ phjuḥ.

Line 9: It is not specified who this pādamūla is. One may wonder whether it is the Mrateñ Nirantarācārya or the venerable of the temple receiver of the donation.

gnaṅ, l. 5: Philip N. Jenner conjectures the meaning of "place to live, dwelling-place, residence" (2009, s.v.), but until a more precise context is found, Saveros Pou's proposed translation, "barreau de clôture. Clôture" (2004, s.v.), which refers to the modern ឃ្នង , "movable rail of a fence that can be raised or lowered to allow animals to enter or exit" (Headley et al. 1997, s.v.) seems more satisfactory to us; thus, it could be an enclosure here.

cāṁ phjuḥ, l. 14: This function, which also appears in the K. 391 inscription (O, l. 21; 1004 śaka; 6298-2992), has not yet been clearly clarified. As George Cœdès pointed out, pjuḥ (~ phjuḥ), and its derivative paṁjuḥ, are regularly applied to religious people (5843). Philip N. Jenner conjectures the meaning of pjuḥ "to serve" and by extension that of cāṁ phjuḥ (2009, s.v.), "[a person who] supervises service," which explains the interpretation proposed by Claude Jacques, "guardian of services" (55).

Chok Khlā et Dhār, l. 10: Claude Jacques combines the last two names into a single toponym (404).

First published by Claude Jacques (402-405). Edited later by Dominique Soutif and Julia Estève (), using photographs and stamps made in January 2013 during a survey organized by the them and Ang Choulean, in collaboration with the epigraphy team of the Authority for the Protection of the Site and Development of the Angkor Region (APSARA).

402-403