SI 3662 and SI 3663 - two wedge-shaped Kharoṣṭhī documents from Niya in the Petrovsky Collection

Abstract

Two unedited wedge-shaped wooden documents SI 3662 and SI 3663 (= SI P 138/a and SI P 138/б, respectively) are deciphered in this paper. A preliminary analysis of their content and physical features shows that SI 3663 is very probably the wedge under-tablet to be bound with the wedge covering-tablet N. i. 17 found by Aurel Stein at the N. I. site (Niya, China). SI 3662, a king’s instruction issued to Soṃjaka the Cozbo, is in all probability from the Niya sites, too.

Full Text

1. Introduction SI 3662 and 3663 belong to the SI (Serindia) collection of the Institute of Oriental Manuscripts, Russian Academy of Sciences (IOM, RAS). These unedited tablets are in rather good preservation. They lack site marks, and their text is bleached to some degree. Bearing old shelf numbers SI P 138/a and SI P 138/б, respectively, after the name of Nikoloai Fyodorovich Petrovsky (1837-1908),1 their existence has been known to Russian scholars. For example, Dr. Margarita Vorobiova-Desiatovskaia’s introduction to the SI © CHING Chao-jung 慶昭蓉, Postdoctoral researcher, Centre de recherche sur les civilisations de l’Asie orientale, CNRS / International Research Fellow, Japan Society for the Promotion of Science-Kyoto University (ching.cj@gmail.com) * I would like to express my sincere gratitude to Prof. I.F. Popova and the Department of Manuscripts and Documents of the Institute of Oriental Manuscripts, Russian Academy of Sciences since 2009. The present paper is one of the results from my project “Dynamics of writing traditions on the Silk Road: A case study of Tocharian and other languages” (Mentor: Prof. Yoshida Yutaka; Host Institute: Department of Linguistics, Kyoto University), during which my consultation on the originals of SI 3662 and 3663 in autumn 2016 was supported by the Kakenhi of the JSPS. 1 On his outstanding career, see VOROBIOVA-DESIATOVSKAIA 2004; POPOVA 2008: 25. collection in 2004 - translated by Dr. Jan Nattier into English in 2006 - is extracted as follows:[2][3] (1) The N.F. Petrovsky Collection At present 582 items are registered in the holdings of this collection, for whose study S.F. Oldenburg was principally responsible. Of these, 266 are Sanskrit manuscripts in Brāhmī script on paper… Another 297 are Khotanese manuscripts on paper… Another eleven fragments are in Tokharian. There are two documents on wood written in northwestern (Gāndhārī) Prakrit in the Kharoṣṭhī script, and one document on wood with two different scripts: Brāhmī on one side (in the Tokharian B language, = Kuchean) and Kharoṣṭhī (in the Gāndhārī language) on the other. Two documents on wood are written in Old Uighur, in the Uighur script. … In addition to the bilingual-biscript tablet, namely SI P/141 (= SI 3672),[4] the “two documents on wood written in northwestern (Gāndhārī) Prakrit in the Kharoṣṭhī script” no doubt mean the Niya-Gāndhārī ones to be treated below.[5] In fact, on the occasion of the International Conference “Turfan Revisited” (8-13 September, 2002), Dr. Vorobiova-Desiatovskaia had pointed out the existence of Niya documents in the Petrovsky Collection:5 In all, the Petrovsky collection of manuscripts written in Indian scripts contains 582 items. The different languages present therein are: Sanskrit (251 items), Khotanese Saka (297 items), Tocharian B, Old Uighur, Old Tibetan and North-Western Prakrit. The majority of the manuscripts are written in Central Asian Brāhmī script of the southern type. But we also have some wooden documents in Kharoṣṭhī script originated from the region of Niya and Kroraina, and wooden documents with text on both sides - Tocharian B in Brāhmī on the recto side and Kharoṣṭhī on the verso side. There is also a unique wooden business document in the Old Uighur language. … During my visits at the IOM in 2009 and 2012, SI 3662 was still kept in a white cardboard box together with the only Kuchean document in the Malov collection, i.e. the wooden tag about monastic wealth SI M-TD/31б (= SI 3664),[6] while other wooden documents in the Petrovsky Collection had been kept in separate paper envelopes, number by number. SI 3662 and 3664 were delicately embedded into the box with silk cloth lining until their separation into new envelopes around 2014. Since SI 3664 was selected for exhibition,[7] SI 3662 must have been cherished for a long time as well. In 2015, SI 3663 drew my attention during my consultation of SI 3662. From the photographs kindly provided by the Institute in May 2017 [Fig. 1-4],[8] readers can easily recognize both their shapes as “wedge under-tablets”, in Aurel Stein’s terms.[9] Surprisingly, the content of SI 3663 matches the wedge covering-tablet N. i. 17, which is edited in Kharoṣṭhī Inscriptions Discovered by Sir Aurel Stein in Chinese Turkestan as No. 15. On the other hand, SI 3662 shows some features that imply its scribe’s hastiness or lack of experience. There is no doubt as to their genuineness and precious value for Niya studies.[10] Fig. 1. SI 3662, obverse side (Photo courtesy of the IOM, RAS) Fig. 2. SI 3662, reverse side (Photo courtesy of the IOM, RAS) Fig. 3. SI 3663, obverse side (Photo courtesy of the IOM, RAS) Fig. 4. SI 3663, reverse side (Photo courtesy of the IOM, RAS) 2. SI 3662 (l. 3.2 cm×w. 22.8 cm×th. 0.5 cm) Und. Obv. 1 [ma]hanuava maharaya lihati cozbo soṃjakas̠ a maṃtra deti s̠ aca[11] ah.[n]. ? ? 2 pug̠ o lýipeyas̠ a ca [12] viṃñav̠ e[nt]i yatha dvaraka lýipana ma[tr]eti [yatha] mṛt[a] jaṃna [l].[ip].[m]. ? go 1 taha matreti lýipatg̠ a ditag̠ a matreti puna mṛ[ta]g̠ a yahi e[da] ? ? 3 mu[dra atra eśati] praṭha yati lýipana śa[va]tha ka[v̠ ]iṣ[y]ati go [1] ṣ. v.oṣidavo Und. Rev. [FAINT TRACES] lýipanenas̠ a ca Notes [U. O. 1] [ma]hanuava maharaya lihati: A blank of 12.5 cm after this sentence. [U. O. 2] viṃñav̠ e[nt]i: Here viṃñav̠ eṃti ‘inform, report’(pres.3pl.)[13] is expected. However, the second anusvāra is not written, although the tablet surface below v̠ e is slightly scratched, perhaps during its unearthing. Moreover, the final akṣāra of this verb looks strange [Fig. 5]. It may denote nti, a ligature so far unknown to scholars, if it is not an inadvertent error of the scribe. ma[tr]eti: This verb is faded, seemingly due to surface friction. [Fig. 5] yatha: There is a dark brown speckle between ya and tha. [Fig. 6] mṛt[a]: The ink spot above ta seems to be a discontinued vocalic remark denoting i or e. [Fig. 6] [l].[ip].[m]. ?: An extremely faint passage denoting a personal name in genitive case based on context. The final akṣāra is totally bleached. It can be restored as a cramped s̠ a similar to the one in lýipeyas̠ a in the same line. [Fig. 6] lýipatg̠ a: My transcription tg̠ a follows Burrow’s tǵa with regard to current convention.[14] [U. O. 3] praṭha: The word is written rather cursively. [Fig. 7] ka[v̠ ]iṣ[y]ati: Sic! If in the scribe’s mind, kariṣyati (fut.3sg. of √kṛ) was to be written, he however distorted the tail of the r to the right, making it resemble v̠ (v́ in old convention). [Fig. 7] go [1] ṣ. v.oṣidavo: The two akṣāras after go are clumsy. The scribe may have intended to write vyoṣidavo ‘to be handed over, to be paid (that is due)’[15] immediately after go, then vyo was altered to the figure 1 and the unfinished ṣ was obliterated, before the gerundive was rewritten afterward. [Fig. 8] Fig. 5. viṃñav̠ e[nt]i yatha dvaraka lýipana ma[tr]eti Fig. 6. yatha mṛt[a] jaṃna [l].[ip].[m]. ? go 1 Fig. 7. praṭha yati lýipana śa[va]tha ka[v̠ ]iṣ[y]ati Fig. 8. go [1] ṣ. v.oṣidavo Text restoration with preliminary punctuation and translation |U.O.1 [Ma]hanuava Maharaya lihati. Cozbo Soṃjakas̠ a maṃtra deti. s̠ aca ah(*u)[n](*o iśa) |2 Pug̠ o Lýipeyas̠ a ca viṃñav̠ e[nt]i. yatha Dvaraka Lýipana ma[tr]eti. yatha mṛt[a] jaṃna [L](*ý)[ip](*a)[m](*as̠ a) go 1. taha matreti lýipatg̠ a ditag̠ a, matreti puna mṛ[ta]g̠ a. yahi e[da] |3 (*kila) mu[dra atra eśati], praṭha yati Lýipana śa[va]tha ka(*r)iṣ[y]ati, go [1] ṣ. v(*y)oṣidavo. |U.R. (*Pug̠ o Lýipeya){{s̠ a ca}} Lýipanenas̠ a ca. [Main text] His majesty the king writes. He instructs Soṃjaka the Cozbo as follows: Pug̠ o and Lýipeya inform that Lýipana the Dvaraka makes a statement: “The dead person Lýipama had one cow.” Then Lýipatg̠ a makes a statement: “It was given (to him)”, and again he makes a(nother) statement: “It has died.”[16] When this sealed wedge-tablet reaches (you) there, (then) as soon as Lýipana makes an oath, one cow is to be handed over to him. [Object] Concerning Pug̠ o, Lýipeya and Lýipana.[17] Discussion Since double-wedge documents are highly formulated, the covering-tablet to be bound with SI 3662 should bear a sentence on the obverse side such as Cozbo Soṃjakas̠ a dadavo/dadavya. Its reverse side would be merely blank or begin with typical formulae, for example atra na paribujiśatu, hastagada (kartavo/kartavya),[18] iśa vis̱ ajidavo ‘if you are not clear about it there, (the relevant people/things must be taken) in custody and sent (to me) here.’ Yet it is difficult to find a tablet of appropriate size to match from other collections. At least it is clear that all the three covering-tablets addressed to Soṃjaka, ΓA1151, 1152 and 1155 kept in the Hermitage, are to be excluded.[19] The scribe seems to be inexperienced. Possible errors or clumsy features include: (1) the strange akṣāra for the 3rd person plural ending; (2) alteration of mistakenly written *mṛte or *mṛti to mṛta; (3) incorrect writing of kariṣyati; (4) modification of the text after go ‘cow’ in order to insert its quantity. This may explain the occurrence of the syntax yatha… yatha… taha… which looks unusual among Niya documents. As remarked by Burrow, yatha with the indicative is regularly used in introducing quoted speech.[20] It is noteworthy that when only yatha is used, the speech is quoted indirectly, i.e. from the king’s point of view. For example: KI No. 52. ahono iśa Lýipeya viṃсav̠ eti. yatha eṣa iśa krasena [sic] dharmena mahi maharayas̠ a uṭa liṣita, tas̱ a nadha coritag̱ a hoati. “Lýipeya reports here now that he dispatched a camel to me the great king to the klaseṃna arrangements, and its load was stolen.” On some occasions, its coordination with taha described more complicated situations. For instance: KI No. 63 Lýipeya viṃnav̠ eti: yatha atra khakhorni stri 3 nikhalitaṃti, taha sudha edas̱ a stri maritaṃti, …“Lýipeya reports that they took out three witch-women. They killed only the woman belonging to him, …” In a few cases, however, not in double-wedge documents, but rectangular ones, as a kind of judgment, the speech is quoted directly after yatha: KI No. 318 Larsu viṃñavita. yatha mahi naṭha, taha Saṃgila ni daza Kacanoas̱ a paride nikhalida. “Larsu reported, ‘property of mine was lost and was recovered from Kacano, slave of Saṃgila.’ ” Needless to say, in KI No. 63 and other similar examples from double-wedge documents, the passage yatha… taha… delineates a whole statement from a certain person who informs/reports (vi(ṃ)ñav̠ eti) or complains (garahati) to the king.[21] In SI 3662, I assume that the text from yatha Dvaraka … until puna mṛtag̠ a is Pug̠ o and Lýipeya’s quotation of different statements including Lýipana’s. The problems to be solved are Lýipana’s concern and Lýipatg̠ a’s role in this matter. My assumption is based on the usage of the otag̠ a participle. As shown by Burrow and Jamison, this extended form of past participle is used frequently as a passive adjective with a genitive agent.[22] So, if Lýipatg̠ a were the one who gave a cow to Lýipama at an earlier time, or who gave it to someone else after Lýipama’s death, he should have been expressed in the genitive (i.e. Lýipatg̠ as̠ a). So it is easier to interpret that Lýipana did not receive - either as Lýipama’s relative, creditor or a local officer being responsible of animal husbandry - the cow left by Lýipama, while Lýipatg̠ a asserted his ownership and then stated the cow’s death. If this interpretation is correct, here we see an unusual order of VS after taha, not to mention the fact that VS is so far unfound with ma(ṃ)treti ‘he/she says’ in Niya documents. The function of the denominative verb ma(ṃ)treti is different from ma(ṃ)tra deti ‘gives an instruction’ that is exclusive to the ruling class. Unlike ma(ṃ)tra deti to be followed by the introductory particle s̠ aca, no word is placed between ma(ṃ)treti and quoted speech.[23] In fact, NPkt. ma(ṃ)tra means not only ordinary speech or official instruction, but also argument and claim.[24] Hence it seems to me that Pug̠ o and Lýipeya reported different arguments from Lýipana and Lýipatg̠ a at the same time, in particular Lýipatg̠ a’s unwillingness to give his cow away. The name Lýipama is only attested in KI Nos. 21, 78 and 345 (verso). While the latter two are just name lists, KI No. 21 involves Dhamaśriae’s inheritance concerning a cow of which the ownership was shared between her father Lýipama and a man called Kame.[25] SI 3662 reveals another problem left by Lýipama. In the corpus of Niya documents, the name Lýipana is attested more often. An Ari-official called such is mentioned in KI No. 767 dated to the 6th regnal year of Vaṣmana. Another Ari of the same name in KI No. 123 is very likely the identical person, too, of which the text is dated to the 30th year of an unnamed king (possbily Mahiri, i.e. Vaṣmana’s predecessor). If we assume the attestations in Nos. 278, 309 and 450 all indicate the same person, his business and/or official duty would then closely relate to cultivation, storage of grains and collection of land tax. In this case, his claim on a cow would not be surprising: Although his occupation Dvaraka (lit. ‘Door/Gate-man’)[26] in SI 3662 is thus far unattested, at least he seemed to be familiar with farming business. As to Pug̠ o and Lýipeya, the former is once mentioned in KI No. 322 dated to the 21st regnal year of Mahiri. As to the latter, the dates attested span from the 11th year of Mahiri to the 11th year of Vaṣmana.[27] Since Soṃjaka was known to be one of the most active officers during Mihiri’s reign,[28] and given the fact that he had served as a Cozbo as early as the 20th year of Aṃgoka (KI No. 582), the predecessor of Mahiri, it is safer to date SI 3662 to Mahiri’s reign. Nonetheless, the chronology of Ancient Niya is still an open issue,[29] hence discussion about text dating must be suspended for the moment. 3. SI 3663 (l. 5.6 cm×w. 23.8 cm×th. 0.7 cm) Und. Obv. 1 [ma]hanuava maharaya lihati cozbo [ta]ṃjakas̠ a matr[a de]ti s̠ aca ahuno iśa 2 lýipeya viṃñav̠ eti yatha edeṣa vaṃti krorayaṃmi [dharmena vaḍ̠ a]vi kabhoḍhami tatra kabh.ḍha[mi kolý]is̠ a suḡita[s̠ a[30] ca nac̄ ira] gachaṃti ? 3 me rayaka vaḍ̠ avi naṭaṃti ima var[ṣ].mi carapuruṣa [a]p[ru]ya ? [s̠ a]vida [co]ritaṃti avi ahuno caturtha varṣa lýipeyas̠ a ? 4 s̠ avida coritaṃti sudha ahuno pra[g̠ aṭa] nikh. ? ? kolýis̠ a ni [da]za sutrena cora nikasati yahi eda kila[mutra] Und. Rev. 1 c[o].bo lýipeyas̠ a [U. O. 1] [ma]hanuava maharaya lihati: A blank of 11 cm after this sentence. [U. O. 2] krorayaṃmi: Sic! An error for Krorayinaṃmi/Kroraiṃnaṃmi ‘in Kroraina’ [fig. 9]. [dharmena vaḍ̠ a]vi: A dark brown spot spans from dha to ḍ̠ a. [fig. 9] kabhoḍhami: The bh here can be classified as Type 4 in Glass’ scheme.[31] The next attestation in the same line is too bleached to classify. [Fig. 9 and 11] [U. O. 3] naṭaṃti: A newly attested pret. 3pl. ‘they were lost’ developed from naṭha (Skt. naṣṭa < √naś ‘to be lost, perish, disappear, etc.; to cause to be lost or disappear, drive away, remove, etc.’). Here it is so translated instead of ‘they perished/removed (the mares)’, cf. KI No. 122 Parcona pirovaṃmi go mahaṃta 1 naṭha ‘One large cow was lost at the fortified station of Parcona’.[32] ? [s̠ a]vida: To be restored as (*pra)[s̠ a]vida. The word is translated by Burrow as an adjective ‘granted, allowed’ as well as a noun ‘a grant’ derived from pra + √sū ‘to allow, give up, to deliver’.[33] Presently it is translated as ‘produced, released’.[34] In SI 3663, this adjective means approximately, ‘appointed, arranged, assigned (for an action or a task)’. pra[g̠ aṭa] nikh. ? ?: To be restored as pra[g̠ aṭa] nikh(*aleṃti). Having related the two words to Skt. prakṛta- and niṣkālayati respectively,35 Burrow interpreted prag̱ aṭa nikhalitaṃti in KI No. 17 as ‘(dogs and foxes) fetched out (the treasure) into the open’. The other example in KI No. 211 aṃсeṣa palýi na prag̱ a[ta] nikhales̱ i was translated by him as ‘you are not revealing the tax of other people’. In SI 3663, this phrase seems to indicate that the hunting persons revealed their misbehaviour, as clearly described in the next sentence: Kolýis̠ a’s slave drove (or: chased up; lit. ‘moved’) the stolen mares by rope. [Fig. 10] Fig. 9. krorayaṃmi [dharmena vaḍ̠ a]vi kabhoḍhami Fig. 10. kolýis̠ a ni [da]za sutrena cora nikasati Fig. 11. Left part of SI 3663, obverse side. Text restoration, preliminary punctuation and translation The word kabhoḍha (loc. kabhoḍhami), suggested by Burrow in LKD as ‘grazing-land, pasture’ and earlier in 1934 as ‘some privately owned pasture land’,[35] was attested in KI only in three documents: Nos. 13, 15 and 392. Strikingly, the content and size of SI 3663 (w. 23.8×l. 5.6 cm) and those of KI No. 15 (w. 23.5×l. 5.7 cm)[36] match each other. Although the photograph of KI No.15 is unavailable to this day, it is possible to restore the full text as follows: SI 3663 + KI No. 15 |C.O. Cozbo Taṃjakas̱ a dadavo |U.O.1 [Ma]hanuava Maharaya lihati. Cozbo [Ta]ṃjakas̠ a matr[a de]ti. s̠ aca ahuno iśa |2Lýipeya viṃñav̠ eti. yatha edeṣa vaṃti Kroray<*in>aṃmi [dharmena vaḍ̠ a]vi kabhoḍhami. tatra kabh(*o)ḍhami [Kolýi]s̠ a Suḡita[s̠ a ca nac̄ ira] gachaṃti. (*i)|3me rayaka vaḍ̠ avi naṭaṃti. ima var[ṣ](*aṃ)mi Carapuruṣa [A]p[ru]ya (*pra)[s̠ a]vida. coritaṃti. avi ahuno caturtha varṣa Lýipeyas̠ a (*pra)|4s̠ avida. [co]ritaṃti. sudha ahuno pra[g̠ aṭa] nikh(*aleṃti). Kolýis̠ a ni [da]za sutrena cora nikasati. yahi eda kila[mutra] |C.R.1 atra eśati, praṭha Kolýis̱ a Suḡitas̱ a ca varidavo, na iṃ ci kabhoḍhami nac̄ ira gaṃdavo. ghrida-coritag̱ a prace vivada śavathena sakṣ̄ iyena samuha |2anada prochidavo. avi śamuta prace samuha anada prochidavo, yatha dharmena nic̄ e kartavo. atra na paribujiśatu, hastagada iśa vis̱ ajidavo. |U.R. C[o](*z)bo Lýipeyas̠ a [Distination] To be given to Taṃjaka the Cozbo. [Main text] His majesty the king writes. He instructs Taṃjaka the Cozbo as follows: Now here Lýipeya informs that by their side,[37] according to the law (set) in Kroraina,[38] there are mares in a pasture. In the pasture, there Kolýis̠ a and Suḡita go hunting. Those royal mares disappeared (or: were lost). This year the detective Apruya was appointed (to investigate there). They were (still) stealing.[39] And recently[40] Lýipeya was appointed for the fourth year. They were (still) stealing. Only right now they expose (their misbehaviour) evidently: Kolýis̠ a’s own slave drives (or: chases up) the stolen (mares) by rope. When this sealed wedge-tablet reaches (you) there, forthwith Kolýis̠ a and Suḡita are to be prevented from going hunting in the pasture. The dispute about the stolen ghee is to be carefully investigated with sworn testimony. Also as regards to the śamuta, inquiry must be carefully made by you in person and a decision is to be made according to law; if you are not clear about it there, they must be sent here in custody. [Object] In respect to Lýipeya the Cozbo. Discussion One may question the absence of ghrida ‘ghee’ (Skt. ghṛta) and the hapax śamuta in SI 3663, since both were mentioned by the king in KI No. 15. Nevertheless, another double-wedge document KI No. 13 (N. i. 15 + 107), also concerning improper usage of pasture, speaks for an underlying connection between loss of mares and horses and that of ghee. The main text of No. 13 is extracted as follows: ... ahono iśa |U.O.2 Pug̱ o viṃñav̠ eti. yatha etas̱ a kabhoḍhami vaḍ̠ avi storaṃ ca, taha jaṃna tatra nac̄ ira gachaṃti. vaḍ̠ avi aśpa vijaṃti. avi tatra ghrida naṭha. yahi eda kila|3mudra atra eśati ... yatha dharmena nic̄ e kartavo. |4 |C.R.1 jaṃna varidavo. ma iṃ ci bhuya nac̄ ira gachaṃti. ye jaṃna tatra nac̄ ira gadaṃti: Yitaka Og̱ a ? Sucaṃma Vaṃto Opg̱ eya Cinamas̱ a ca. “Pug̱ o informs now here that in his pasture there are mares and horses. There the people go hunting. They wound the mares and horses. Also some ghee there has been lost. … a decision is to be made (by you) according to law. The people are to be prevented. They shall not go hunting anymore. The people who went hunting there were Yitaka, …, Opg̠ eya and Cinama.” The word śamuta in KI No. 15 is not translated in TKD, but in LKD Burrow indicated the possibility to identify it with another obscure word śamuḍa. The latter is mentioned once together with meat (KI No. 252) and once in contrast to felt garments (KI No. 387), so perhaps it is another product of animal husbandry. In SI 3663 + KI No. 15, the absence of ghrida and śamuta in Lýipeya’s report may be explained by the ongoing investigation. In other series of double-wedge documents, for example KI Nos. 58 and 63, the king - or his scribe - just simplified the background information in later instructions when the case had been processed for a certain period. So we may assume the existence of earlier records about this matter, too, such as the king’s initial order of investigation of the loss in the pasture. 4. Concluding words In addition to the famous Dharmapada collected from Khotan,[41] SI 3662 and 3663 are the only Kharoṣṭhī material from the southern rim of the Tarim Basin in the SI Collection. Although these two wedge tablets cannot be dated precisely, they are not to be dated to the earlier kings such as Pepiya or Tajaka from prosopographical aspects. As to the provenance, SI 3662 may have been excavated from N. V, which is known to be closely related to Soṃjaka during his service as a Cozbo officer.[42] On the other hand, if my pairing of SI 3663 and KI No. 15 (N. i. 17) is correct, SI 3663 was very likely unearthed at the N. I. site. According to Stein, when KI No. 15 was found on 28 January 1901 by himself, it had been already detached and ‘lying on the surface of the sand’.[43] Moreover, the one who discovered the wooden documents at the Niya sites was a young villager Ibrāhīm, just about one year ago.[44] If local villagers told him everything they knew, one may imagine that SI 3662 and 3663 were either collected by Ibrāhīm himself,[45] or by someone between Ibrāhīm’s discovery and Stein’s first excavation, or even by someone afterwards until Petrovsky resigned his position in Kashghar in 1903. In other words, taking SI 3992 and 3993 as holdings of the Petrovsky collection as granted, these tablets were very probably unearthed before Stein’s revisit and the arrival of other expeditions. They shall be analyzed together with the ones kept in the State Hermitage Museum in order to give a fuller view of the Russian collection of ancient documents from Chinese Turkestan. [In my paper collaborated with OGIHARA Hirotoshi, “SI 3656 and other Kuchean tablets related to the Kizil grottoes in the St. Petersburg Collection”. Written Monuments of the Orient, 2016(2), 44-67, the new shelf number of SI P 139/д (= SI 3668) is wrongly given as SI 3669 by mistake. We apologize to all the readers for our error.]