¹ The foliation marked in the text of this article is that of the Haidarābād Codex of the Bābar-nāma.

² In the Hai. and Elphinstone MSS. the text begins here; in Kehr's MS. an invocation precedes.

³ Pādshāh. To translate pādshāh by “king” or “emperor”, as if part of the style of any Tīmūrid, previous to 913 a.h. (1507), is an anachronism, because till that date even a ruling Tīmūrid was styled Mīrzā (fol. 215), and then first did Bābar change his title. The word pādshdh (it is hardly necessary to say) occurs frequently as a common noun in the writings of Bābar's circle. He himself says, e.g., that his father was an ambitious pādahdh, i.e. ruler (fol. 5b); it was proposed to make Jahāngīr Mirzā pādshāh (ruler) in Farghāna (fol. 24b); Ḥaidar Mīrzā writes of Yūnas Khān as pādshāh in Mughūlistān, i.e. having chief authority (Tārlkh-i-rashidī, Elias & Ross, p. 74). Gul-badan Begam writes of an amir who was pādshāh, i.e. commandant, in Bhakkar, (Humāyūn-ndma, trans., p. 148)Google Scholar. I have seen an instance of its use for a chief boatman. In the Taẕkīrtu'l-bughrd the word Pādshdh is part of the style of a Mughūl nomad, Satuq-bughra Khan Ohāzl Pādshāh and, it would seem, implies his supremacy amongst the Mughūl Khāns. Perhaps Bābar's assumption of it as a title in 913 a. h. asserted his then supremacy amongst living Tīmūrid Mīrzās.

⁴ Bābar was born on Saturday, February 15, 1483 (Muharram 6, 888 a.h.), and died December 26, 1530 (Jumāda i, 6, 937 a.h.). His father, 'Umar Shaikh whom he succeeded in Farghāna, died on June 4, 1494 (Ramẓān 4, 899 a.h.), “the year of Charles VIII's expedition to Naples” (Erskine). Bābar was born nine months before Luther (b. November 10, 1483).

⁵ See Āīn-i-akbarī, Jarrett, , pp. 44 ffGoogle Scholar.

⁶ Shahrlār bār ikān dūr. The modern term suiting Central Asian towns is “Garden Cities”. Ālmālīlgh (lit. “apple-like”) was the old capital of Kulja; Ālmātū (var. Ālmātī, named also from the apple) is the Russian Vierny; the now ruined Ūṭrār is on the Sīr, somewhat below its intake of the Aris (var. Urus). “In the days of Tīmūr, Otrār was a place of great note; he died there” (807 a. h., 1405 a. d.) “while preparing for his expedition to China” (Erskine).

⁷ Of the clause here noted, there have been the following translations:— Hai., Elph., and Kehr's MSS., “Mughūl u Ūzbeg jihat dīn;” Wāqi'āt-i-bābar (i.e. Pers. trans.), I.O. MS. 217, “az jihat 'ubūr Ūzbeg;” Erskine Memoirs of Babar, p. 1), “In consequence of the incursions of the Uzbegs;” De Courteille (Mémoires de Baber, i, 1), “Grace aux ravages commis par les Mongols et Ūzbegs.” The Persian 'ubūr may be thought to improve on Bābar, since the towns mentioned lay in the tide-way of nomad passage between east and west, but they are a departure from his words. The Persian text, here as elsewhere, has caused Mr. Erskine to diverge from Bābar. It may be said (though not in this instance) that some part of the deviation found in the French translation, deviation both from the true Turki text and from Erskine's, is the sequel of defect in Kehr's earlier and Persified pages. (Cf. JRAS., January, 1908, art. Bābar-nāma, for specimens of this Persification. For Erskine's comments on the peculiarities of the Persian text see his Preface, p. viii.)

⁸ kim (Samarqand u Khujend) būlghāl. This frequent phrase of Bābar I do not find mentioned in the Turkī grammars; it always, I think, expresses apposition; “that is to say” may be its meaning.

⁹ Following the Persian trans. Abūl-faẓl and Erskine omit Bābar's seasonal limitation here (Akbar-nāna, Bib. Ind. ed., i, 85, and trans. H. Beveridge, i, 221). For a description of the passes into Farghana see Kostenko's Turkistān Region, trans. Simla, 1882, vol. i, sect, i, cap. 2 and 3.

¹⁰ Wilāyat nīng īchkārī bīla; perhaps “through the trough of the country” (de Meynard, īchkār, creux).

¹¹ “A town in Māwarā'u'n-nahr, also called Shāsh, and in modern times Tāshkend” (Rieu, i, 79). Bābar does not identify Fanākat (var. Benākat, Fiākat) with Tāshkent; he does so with Shāhrukhiā. As he distinguishes between Tāshkent, i.e. Shāsh, and Fanākat, i.e. Shāhrukhīa while Rieu identifies the two, it may be that Rieu's statement applies not to “modern” but to old Tāshkent which stood some 14 miles nearer to the Sir than the newer town does. (Is its first syllable Ar. fanā, expressive of its byegone status ?) Fanākat (Shāhrukhīa) is located by Bābar's and by Haidar's narratives near the Sīr, perhaps near modern Chināz. For a discussion on the origin of the name Tāshkent see Von Schwarz's Turkistān, index s. n. Tāshkent; see too Kostenko, i, 320; Parker, Asiatic Quarterly, 1909, art. Samarqand, pp. 2, 74; JRAS., April, 1909, art. Bābar-nāma. Also Raverty's Ṭabaqāt-i-nāṣirl, index s.nn. Tāshkent, Fanākafc, Shāhrukhīa.

¹² Hech dary gha qātīlmās. Pers. trans. (I.O. 217, fol. 1b), hech daryā'i dīgar ham-rāh na shuda. E. and de C. have understood Bābar to say that below Turkistān the Sīr is not tributary to any other river, but, although this is the fact, there is room for doubting if this is what he meant. He may preface his clear (but erroneous) statement that the whole Sīr sinks (sīnkar) into the sand by one denying an alternative end of its course, i.e. fall into a daryd, a larger body of water, presumably the Sea of Aral. His preposition is gha (to), and E.'s “other” is the translation of the gloss dīgar of his Persian source. Bābar, it is evident, did not know the whole course of the Sīr. (See Schuyler, i, 550 ff., and Kostenko, i, 198, 218, amongst modern writers about it.)

¹³ Qaṣbalār. Bābar's geographical unit is the township, or, more exactly, the village, the inhabited and cultivated oasis. Of frontiers he says nothing.

¹⁴ “i.e. passengers eat them gratuitously” (Leyden). Klaproth, “all ein es ist streng verboten sie zu verkaufen ehe sie reif geworden sind” (Archiv für Asiatische Litteratur, pp. 101 ff.)Google Scholar; cf. T.R. trans., p. 425. See Timkowski's, Travels of the Russian Mission, i, 419Google Scholar.

From this point there is a gap of two folios in the Elph. MS.

¹⁵ One kind of melon is called the nāshpātī, but as Bābar has not mentioned the pear, nāshpātī, here may mean this fruit. See Āīn-i-akbarī, Blochmann, p. 6; Kostenko, i, 251; von Schwarz, p. 361.

¹⁶ Tūuqaz tar nan sū kīrār, bū 'ajab tūr kīm bīr yīr dīn ham chīqmās. Pers. trans., I.O. 217, fol. 2: nuh jūy āb dar qila' dar mī āyid u īn 'ajab ast kah hania az yak jā ham na mī bar āyid. Erskine (p. 2, using Mr. Metcalfe's MS., see Rieu, p. 244), “The water-courses of the mills by which the water enters the city are nine, and it is singular that they all issue from the same place;” Erskine (p. 2 n., using his own MS., see Rieu, l.c.), “Nine streams of water enter the fort, and it is singular that they do not all come out at the same place;” de Courteille (i, 2), “Neuf canaux entrent dans la ville, et il y à cela d'étonnant qu'ils ne sortent par aucune issue.” Mr. Erskine had here only the Persian translation to guide him, there being still a gap in the Elph. MS. As he translated in India, the words tar nau took on their technical Indian meaning of channels or pipes serving mills. Bābar's meaning is, I think, that all the water brought into the town of Andijan by nine artificial channels was consumed there, leaving no surplus to come out at even one place.

¹⁷ Khandaq nīng tāsh yānī. Pers. trans. (I.O. 217, fol. 2), dar kinār sang bast khandaq. E. (p. 2), “On the edge of the stone-faced ditch;” de C. (i, 2), “sur le bord extérieur du fossé.” There can be little doubt that the Persian trans, is wrong in its sang bast, both on the ground of the Turk wording and because Bābar's point is the unusual circumstance of a road round a ditch; also because Andijan is built on loess and of loess.

¹⁸ Qīrghāwal āshkina sī bīla. Āshkina is allowed by dictionaries to be the rice and vegetables commonly served with the bird. Erskine (p. 2) writes “broth” and adds, in a note, “a sort of stew, or rather, jellybroth.” Ilminsky prints āiskana, whence de Courteille (i, 3), “quatre personnes ne peuvent venir à bout d'en manger une cuisse.” Klaproth (p. 104), “so fette Adler dass vier Menschen von einem ausgewachsenen satt werden können.” For a recipe likely to be āshkina see Kostenko, i, 287.

¹⁹ b. 1440; d. 1500.

²⁰ Herīdā nashu u namd tdbīb tur.M. de Courteille applies these words to Nawā'ī's writings: “quoique publiées à Herat, sont concues dans cet idiom” (i, 3).

²¹ See Shāh's, DaulatMemoirs of the Poets, Browne, E. G., pp. 350, 351Google Scholar. Yūsuf was with Bāysanghar Mīrzā he may be one with Yūsuf Badī'ī of Farghāna (fol. 181).

²² Gūzlār īl bīzkālc kūb būlūr. The Pers. translator has read Turkī gūz, autumn, to be Turkī goz, eye, and adds (1.0. 217, fol. 2), ashūh chashm u waram ān bisydr ml shud u itibbā ānrā qirrat mī gūyand. There is no Turki basis for the above gloss. For statistics of autumn fever in Turkistan and for a novel febrifuge, see von Schwarz, index s. n. Fieber, and also Kostenko, i and iii, Table of Contents.

²³ Pers. trans, farsang. Ujfalvy (Expédition Scientifique, ii, 179), “L'igadj ou le farsang vaut environ 6 kilomètres.” Cf. von Schwarz, p. 124. From de C.'s Diet. s. n. yīghāch, may be quoted what shows the variable length of this measure:” Trois fois la distance à laquelle un homme, placé entre deux autres, peut se faire entendre d'eux, soit un farsang, soit un mille.” I cannot bring Babar's statements of distance in yīghāch to agree with the farsang of about 4 miles. They work out more nearly to 8 miles per yīghāch. Here if the yīghāch equal the farsang of 4 miles, the distance from Ūsh to Andijān would be 16 miles, but Kostenko gives it (ii, 33) as 50 versts, i.e. 33 m. ¼ fur.

²⁴ Āqār sī, the irrigation channels on which in Turkistan all cultivation depends. Major-GeneralGérard, writes (Report of the Pamir Boundary Commission, p. 6)Google Scholar, “Osh is a charming little town, resembling Islāmabād in Kashmīr,—everywhere the same mass of running water, in small canals, bordered with willow, poplar and mulberry.” He saw the Āq Bīra, mother of all these running waters, as a “bright, stony, trout stream”; DrStein, saw it as a “broad tossing river” (Buried Cities of Khotan, p. 45)Google Scholar. Cf. Réclus, vi, cap. Farghāna; Kostenko, i, 104; von Schwarz, index under related names.

²⁵ Ūsh ning faẓīlatīdā khailī aḥdiṣ wārid dūr. Pers. trans. (I.O. 217, fol. 2), Faẓilat Ūsh aḥddiṣ dar warid ast; E. (p. 3), “The excellencies of Ūsh are celebrated even in the sacred traditions;” de C. (i, 2), “On cite beaucoup de traditions qui célèbrent l'excellence de ce climat.” Many and various legends have gathered round Ūsh; cf. e.g. Ujfalvy, ii, 172. It may be celebrated, as Mr. Erskine says, in the Sacred Traditions, because of places near it honoured of Musalmāns; it is open to question if Bābar's faẓīlat should be restricted, as M. de Courteille restricts it, to climate only. Ūsh has been distinguished for many centuries by its traditions, is a place of pilgrimage still and has revered objects of presumed curative power.

²⁶ A good deal has been written about the position of the Bara Koh (e.g. Ritter, v, 432, 732; Réclus, vi, 540; Schuyler, ii, 43; and the references of the first and second. Also, Timkowski, ii, 49). It seems safe to identify it with the Takht-i Sulaimān Ridge, as e.g. Ujfalvy and Schuyler's personal observations led them to do; but some considerations lead me to suggest that by Barā Koh Bābar does not mean the whole ridge, but one only of its four marked summits, i.e. the one shown in Madame Ujfalvy's sketch of it as the highest and as being symmetrical (Bābar's mauzīn). “Il y a quatre sommets dont le plus élevé est le troisiéme comptant par le nord” (Ujfalvy, i, 96). Madame Ujfalvy's sketch would seem to be taken from the north, because its third summit is the highest (De Paris à Samarqand, p. 330). A permissible meaning of the words Barā Koh is Pointed Hill; this meaning suits her sketch and Bābar's mauzīn; it also helps out the identification of her third summit as the Bara Koh, since only this third is well-shaped and definite. There is this in favour of limiting the name Bara Koh; Babar must have known that Takht-i Sulaiman was the name of the whole isolated rocky ridge. It would clear up a good deal of confusion about names and location, written of by Ritter, Reclus, Schuyler and others, if the name Barā Koh be taken as limited in the way I mention. (A suggestion made (i, 3 n.) by M. de Courteille that Barā Koh should be Bālā (high) Koh has no support in the MSS.)

²⁷ Rūd, a precise word, since the Āq Būra, issuing as the Tūrūq from the Kordun Pass (13,400 feet), falls, after creating the Little Ālāī Valley, to Ūsh (3040 feet) through a canyon 1000 to 2000 feet deep; and thence again to Andijān (1380 feet). Kostenko, i, 104; Huntingdon, in Pumpelly's, Explorations in Turlcistan, p. 179Google Scholar; French Military Map of 1904.

²⁸ When Bābar uses a word twice, once with the Arabic plural āt, once with the Turkī lār, as here, or as elsewhere, begāt u beylār, he seems to mean “all, of every degree”. Hence I translate bālghāt here by “garden-plots”, not intending, however, to give it when it stands alone the meaning of bāghcha, small garden, but taking it as the complement of the closely following bāghlār, with the meaning of “gardens of all sorts”. The point is small, but one does not follow Babar's words without receiving the impression that it is safest to give each weight. He wastes none. Ujfalvy mentions that Ūsh “est situé sur le versant d'une montagne; presque toutes les rues sont en pente” (i, 96). Perhaps this explains why all the gardens were on the torrent and why Babar mentions that they were so.

²⁹ Madame Ujfalvy has sketched its probable successor. Schuyler found two mosques at the foot of the Takht-i Sulaimān, perhaps Bābar's Jauza (Twin) Mosques. (Klaproth takes Jauza Masjid to mean “Nusstempel.”)

³⁰ Aūl shāh jūy din sū qūydrlār, Pers. trans. (I. O. 217, fol. 2b), az īn shāh (var. shah, sih) jūy āb miguzārand; Erskine (p. 3) tentatively, “carry across three streams;” de C. (i, 3), “verser de l'eau du torrent sur quiconque,” etc.

³¹ Ribbon jasper, presumably.

³² Kostenko (ii, 30), 71¾ versts, i.e. 47 m. 4½fur., Postal Road.

³³ “A town between Khurāsān and 'Irāq, near Damghān” (Erskine).

³⁴ The Persian translator inserts mayhz-i bādām, almonds, in the apricots, a fashion well known in khubānī, bought in India, but the Turkī words allow the return to the fruit of its own stoned kernel. Mr. Rickmers tells me that in the Zar-afshān Valley he has often met with apricots so stuffed. Steingass gives” jauz-āghand, a peach stuffed with walnut-kernels”. My husband has shown me that Niẓāmī seems to allude in the following passage from the Haft Paikar, to the practice of inserting almonds in fruits:—

“I gave thee fruits from the garden of my heart,

Plump and sweet as honey in milk;

Their substance gave the lusciousness of figs,

In their hearts were the kernels of almonds.”

³⁵ Pers. trans. (I.O. 217, fol. 2b), ahī-i warāq, “said to be the arkali described in many books of Natural History. See Voyages de Pallas, iv, 325” (Erskine). If, however, as is done by some travellers, the arkali (arkhara) be identified with Ovis poli, it cannot be Bābar's āqkīyīk (white or light-coloured deer or sheep) found at the level of the Sir, circa 2000 feet (cf. fol. 5), unless, indeed, the habitat of Ovis poli has changed. Parts of the Marghlnan and Khujend wilāyat are high enough for the present limit (10,000 feet) of Ovis poli, running back as they do up the northern face of the Kok Sū and Khūtūr which, moreover, have their southern slope to the Pamirs, a haunt of the great sheep. Perhaps the q kīyīk found at Akhsī were Ovis Karelini; the āq of the name not needing to be taken as pure white, light and whitish being common meanings of the word. Cf. Curzon's Pamirs and the Source of the Oxus, p. 26; Shaw's Voc. s.n. kīyīk; Atkinson's Amur, index under related names.

³⁶ Pers. trans. Tājik. Bābar describes the Asfara people as Persianspeaking Sārts. Modern opinion distinguishes the Sārt as a settled resident, usually of mixed descent. This modern view would allow Bābar's Marghīnanī Sārts to be Turki-speaking, settled Turks, and his Asfara Sārts to be Persian-speaking Tājiks. Cf. Shaw's Voc. s.n. Sārt; Schuyler, i, 104 and note; Nalivkine's, Histoire du Khanat de Khokand, p. 45 n.Google Scholar; von Schwarz, index s.n.

³⁷ Shaikh Burhānu'd-dīn 'Alī Qīlīch, b. circa 1135, d. 1197 (b. 530 a.h., d. 593 a.h.). See Hamilton's Hiādyat.

³⁸ Asfara town is in the foot-hills of the Turkistān Range; Asfara wilāyat runs back too far upon this for “foot-hills” to apply. Wārukh (4470 feet) lies 34 miles back from Asfara town, Hushyār (Curzon, Ushiyār; French Map, Outchyār) about as far. “Hill-country” suits for both Sūkh and Hushyār.

³⁹ Measured on the French Military Map, the direct distance may work out at some 65 miles, but the road makes a détour round mountain spurs. To the word farsang of his source, Mr. Erskine here attaches an elaborate note concerning Indian measurements which, valuable as it is in itself, is made the less applicable here by the uncertain length of the yīghāch.

⁴⁰ Bāghcha. Cf. n. 28.

⁴¹ Ḥai. MS. Fārsī, gūy, the word Fārsī being entered, apparently by the scribe of the MS., over the line, as if at first omitted. [The lacuna of the Elph. MS. still continues.] Kehr's MS. has kohī, but its earlier pages are Persified; the Pers. trans. (I.O. 217) has also kohī, hence the “mountaineers” and “montagnards” of E. and de C. The Fārsī of the H. MS. would have been useful to Ritter (vii, 733–4) and to Ujfalvy (ii, 175).

⁴² Of this stone neither Fedtchenko nor Ujfalvy could get news.

⁴³ Here Bābar distinguishes between Tāshkent and Shāhrukhīa.Cf. fol. 2, n. 11.

⁴⁴ In 908 a.h. (first half of 1503). He left the hill-country above Sūkh in Muḥarram, 910 a.h. (mid-June, 1504).

⁴⁵ For an interesting account of Khujend see Kostenko, i, 346.

⁴⁶ Kostenko, ii, 29–31. Andijān to Marghīnān, 47 m. 4½ fur.; Marghīnān to Khokand, 56 m. 2¾ fur.; Khokand to Khujend, 83 m. 2¾ fur.; total, 187 m. 2 fur. from Khujend to Andijān. By help of the time-table of the Transcaspian Railway, the distance by rail from Khujend to Samarqand can be pieced out as 154 m. 5¼ fur.

⁴⁷ Both are still honoured in Khujend. See Kostenko, i, 348. For Khwāja Kamāl's Life and dīwān see Rieu, ii, 632, and Ouseley's, Persian Poets, p. 192Google Scholar.

⁴⁸ Kūb ārtūq dūr. Perhaps this means that the fruit was the more taken to India where Bābar wrote. Pers. trans., bisyār bihtar; Erskine, “greatly excelled;” de Courteille, “beaucoup plus en vogue.”

⁴⁹ Hai. MS., M( )nūgh( )1; Pers. trans, and Erskine, Myoghil; Ilminsky, M()tūgh()l; de C., Mtoughuil; Réclus and Schuyler, Mogul Tau; Nalivkine, “d'apres Fedtchenko,” Mont Mogol; French Map of 1904, M. Muzbek; Kostenko, Mogol Tau. This is, says Kostenko (i, 101), the western end of the Kurama Range (Kendir Tau) which comes out to the bed of the Sir. It is 26⅔ miles long and rises to 4000 feet. Von Schwarz says it is quite bare; various writers ascribe climatic evils to it.

⁵⁰ Pers. trans, ahū-i safid, a variation of its rendering (fol. 3b) by ahū-i warāq.

⁵¹ The marāl is frequently mentioned by Atkinson who takes it to be the red or fallow deer. Von Schwarz mentions it (index s.n.), and Kostenko (i, 57, and iii, 70) writes of the export of its fresh horns to China and of the value of its skin. Under the word būghū there stands in the Hai. MS. (fol. 4) gazawan-kohī and (fol. 5) tika-kohī. De Courteille (i, 7) takes būghū marāl to mean “cerf et biche”, and this they could do if it were not open to give them the fuller meaning of two kinds of game. A precise parallel of the double meanings of these two words is found in von Schwarz's list of Turkistān game, where stand together Hirsch Damhirsch, stag and hind, or two varieties of deer.

⁵² Here in the Pers. trans, recurs the misreading of “eye” for “autumn” noticed in n. 22.

⁵³ “The Village of the Almond.” See Schuyler, ii, 3, and note.

⁵⁴ Schuyler (ii, 3), 18 miles.

⁵⁵ Hai. MS. Hamesha bū deslit td yīl bar dūr. Marghīnān ghā kim sharqī dīr, hamesha mūndīn yīl bārūr; Khujend gha kīm gharībī dīr, dā'im mīndīn yīl kīlīr. Bābar seems to say that the wind goes always east and west from the steppe as from a central generating point. E. and de C. have given it alternative directions, but in saying that wind goes east or west in a valley hemmed in on north and south there is little point. Bābar's statement is limited by him to the steppe in the contracted mouth of the Farghana Valley (pace Schuyler, ii, 51) where special climatic conditions rule. Of these, roughly put, are difference of temperature on either side of the Khujend narrows, draughts resulting from this difference, the heating of the narrows by reflected sun-heat from the Mogol Tau and inrush of north-west wind through the pass near Mīrzā Rabāṭ. Bābar calls the wind of Hā Darwesh a whirling wind and so modern travellers have found it. Thinkable at least it is that a strong westerly current (the prevailing wind of Farghana) entering over Mirza Rabat and becoming, on the hemmedin steppe, the whirlwind it does become—perhaps by conflict with the hotter indraught from the Khujend narrows—might force that indraught back into the narrows, in the way e.g. that one Nile forces back the other. Local observation only can guide the translator; the directness of Babar's words compels belief in their significance and this most so when what he says is unexpected. The manuscript sources agree in having “to (gha) Marghīnān” and “to (gha) Khujend”. It is somewhat strange that Bābar should take for his eastern wind-objective a place so remote and sheltered as Marghīnān. Makhrām, where, moreover, there is a “cleft” to which evil climatic influence is attributed would suit his context better, but it finds no mention in the Bdbarnama. Cf. Réclus, vi, 547; Schuyler, ii, 51; Cahun, , Histoire dn Khanat de Khokand, p. 28Google Scholar; Sven Hedin's Durch Asiens Wüsten, index s.n. burān.

⁵⁶ i.e. Akhsī Village. Kehr, Akhsīkīt; Ilminsky, Akhsīkiṣ. Dr. Ethémentions that in I.O. 1909, the dīwān of Aṣīru'd-dīn, the place-name is written clearly Akhsikeṣ, the form to which Ilminsky has departed from Kehr. The ancient name of Akhsī was Akhsī-kint; the three dots which have been taken as those of ṣā'ī maṣllaṣa might be those of the nūn and the tā in kīnt.

⁵⁷ See Rieu, ii, 563; Daulat Shāh, I.c., p. 131; Ethé, I.O. 1909.

⁵⁸ By measurement on the map the distance seems to be about 80 kilometres, i.e. 50 miles.

⁵⁹ Modern information about the oasis towns of Turkistān allows Bābar's description of Akhsī to be better understood than it has been either by earlier translators or by the numerous writers who have drawn inferences from their words.

1.—The Turkī passage is as follows: H., Elph., Kehr's MSS. (Ilminsky, p. 6), Saihūn daryā sī qūrghānī astīdīn āqār. Qūrghān baland jar austīdā wāqi' būlūb tūr. Khandaqī nīng ūrunīgha 'umīq jārlār dūr. 'Umar Shaikh M. kīm, mūnī pāy-takht qīldī, bīr īkī martaba tāshrāq dīn yana jarlār sāldī.

Of this the translations are as follows:—

(a) Pers. trans. (I.O. 217, fol. 3b), Daryā-i Saiḥūn az pāyhā qila'-i o mī rezad u qila'-i o bar jar balandī wāqi' shuda ba jāy khandaq jarhā-i 'umīq uftāda. 'U. S. M. kah ānrā pāy-takht sākhta, yak du martaba az bīrūn ham bāz jarhā andākht.

(b)Erskine (p. 5, translating from the Persian), “The river Saihun flows under the walls of the castle. The castle is situated on a high precipice, and the steep ravines around serve instead of a moat. When 'U. S. M. made it his capital he, in one or two instances, scarped the ravines outside the fort.”

(c) De Courteille (i, 6, translating from Ilminsky's imprint), p. 6, “Le Seiḥoun coule au pied de la fortresse qui se dresse sur le sommet d'un ravin, dont les profondeurs lui tiennent lieu de fossé. 'U. S. M. à l'époque où il en avait fait son capitale, avait augmenté à une ou deux reprises, les escarpements qui la ceignent naturellement.”

2.—The key to Bābar's meaning is provided by the word jar, taken in the sense, common in Turkistān, of a ravine cut by water or by man, in the loess of oases, below the general level of the land. Writing of Tashkent, Kostenko (i, 321) says of one subdivision (in which is Jar Kūcha, Ravine Lane) that it is on level ground and is divided by a deep ravine. Of another he says that it is cut by deep ravines (Babar's 'umiq jarlār). These statements, together with the information given by Kostenko and von Schwarz, about the plan of towns, the creation of oases and the characteristics of loess, allow Babar to be understood as saying of Akhsi in the fifteenth century what Kostenko says of Tāshkent in the nineteenth, namely, that its qūrghān stood above the ravines, natural or artificial, of the Kāsan Water and not on a precipice washed by the Saihūn.

3.—Wanting this modern light on the word jar, Bābar's meaning has not been clearly understood; of this there is sign in Erskine's location of Akhsi on a precipice with its walls washed by the river, and in his and de C.'s uncertainty as to the nature of the work done by 'Umar Shaikh. It is now clear that what the Mīrzā did was not escarpment but the excavation of water-channels, whether for the completion of a pseudo-moat or to meet the needs of a population augmented by his residence.

4.—Wanting modern information, again, it has been thought that the walled town abutted on the river, and it has been inferred that Bābar's father, 'Umar Shaikh, met his death by falling into the Saihun (cf. fol. 6b). Bābar's words, however, when taken with other available information, do not demand to be understood as locating the walls on the river's bank. If Akhsi, i.e. the qūrghān, stood back (as it seems to have done) up the riverain slope, the Saihūn might be said to flow beneath it as the Thames flows below Richmond.

Circumstantial testimony is merely accessory to Bābar's plain statement that Akhsī stood above ravines; the Saiḥūn did not flow in a cleft near Akhsī; it could have been no part of the pseudo-moat. Circumstantial only, but weighty, since the permanent influence of the Kasan Water fixes the site of Akhsī both in the twelfth and the fifteenth centuries, is Yāqūt's statement that Akhsī had gardens through “a whole parasang” and entered from every gate. So too is Bābar's that the Akhsī suburbs stretched about 2 miles beyond the town (see infra, n. 61).

5.—It can be only in the passage under discussion that General Nalivkine found testimony by Babar to”what he sets forth in the following extract (Histoire du Khanat de Khokend, p. 53):—“L'emplacement que cette ville occupait alors était un lieu escarpé, assez élevé au-dessus du fleuve, par les eaux duquel il était constamment miné. Aussi la ville, au témoignage du sultān Bābar, reoula-t-elle successivement vers le nord, ce qui obligea d'en reporter dans la même direction et a plusieurs reprises, les murs et les fortifications. II est tres possible que cette destruction progressive du rivage par les eaux ait été l'une des causes qui firent abandonner l'antique capitale du Farghanah, réduite aujourd'hui à l'état de kichlah (qīshlāq, winter quarters) insignifiant. Le site de celui-ci est à quelque distance de la bergé, qui a cesse d'etre affouillee par le fleuve, depuis qu'il s'est forme la un grand bane de sable.”

An obvious objection to the theory that erosion has led to the retreat and dwindling of Akhī, lies in the fact that the Kasan Water does not yet fall into the Saihun. If in the fifteenth century the Saihun was undermining the very walls of Akhsī, a town which in the twelfth century was, Yāqūt says, one parasang from the mouth of the Kāsān Valley, how is it that land on which it stood remains?

Against this objection it might be urged that the water issuing from the valley may have become less and less in volume, whether by general desiccation or because of increased cultivation on the higher reaches of the stream. These points raise problems requiring scientific adjustment between (supposed) erosion, lessened rainfall and increase of cultivation.

6.—Mr. Pumpelly has posited the search for the site of old Akhsi as an archseological task of the future. Approximately, that site is fixed by the Kasan stream and its offtakes. Perhaps the importance of Akhsi bulks too large in literature through the haze of imperfect information; the town was on and of loess; the valuables of past, as of present Turkistan, were movables; treasures of art or architecture are not to be looked for. Akhsī town in the fifteenth century was a small place; the measure of its gardens is the measure, not of its wailed town, but of the oasis lands redeemed from the waste by the help of the Kāsān Water. It became a “capital” by the caprice of one man; it ceased to be one because the boy Bābar's advisers stayed in Andijan.

7.—Cf. p. 114 for distances which would be useful in locating old Akhsiif Bābar's yīghāch were not variable. Ritter, vii,3,733ff.; Réclus, vi, index s.n. Farghāna; Ujfalvy, ii, 168 ff., and his references to Yāqūt; Nalivkine, pp. 14 ff. and 53; Schuyler, i, 324; Kostenko, Tables of Contents, for cognate general information, and i, 320, for Tashkent; von Schwarz, index under related names and especially p. 345 and plates; Pumpelly, pp. 18 and 115.

The maximum time during which Akhsī could have been his capital is twenty-eight years, i.e. from his appointment to the Farghāna Government, as a child, to his death (870 a.h. to 899 a.h. ).

⁶⁰ Maḥallātī qūrghān dīn bīr shar'ī yardq raq tūshūbtūr. Pers. trans. (i.o. 217, fol. 3b), mahallāt o az qila' yak shar'ī dūrtar uftāda. From these passages E. and de C. have understood that the suburbs of Akhsī were a shar'ī (circa2 miles) from the walled town. The Turkī wording is against this, however, (1) in its comparative dīn. .. yārāq rāq, i.e. further than; (2) in its verb, tūshūbtūr, denoting extension; (3) in its use of maḥallāt, suburbs. It is far to go to Yāqūt for support of what Bābar says of Akhsī in the fifteenth century, but as in his century also the gardens depended on the Kāsān Water, it is useful to know that Yāqūt describes all the gates of Akhsī as opening on gardens and waters which stretched a whole parasang (Ujfalvy, ii, 180, who refers to Yāqīt, i, 162). For its maḥallāt not to adjoin a town would be not only a misnomer, but against the uniform plan of the oasis towns of Turkistān (cf. von Schwarz, pp. 133 ff.).

⁶¹ I do not see the point of the Persian proverb Bābar quotes. As suits with his reading that the suburbs of Akhsl were 2 miles from their town, Brskine takes the questions as asked by a person coming out of town and looking for the suburbs. De Courteille (i, 8) translates by, “Ne me parlez plus de village! Ne me parlez plus d'arbres !” If with Erskine, he had not understood the suburbs to be 2 miles from Akhsi, he might be thought to express the fatigue of one making for the walled town and wearying of the long suburban road. As he has not translated accurately, his varied wording suggests that he knew the proverb elsewhere. His rendering supports my location of the suburbs rather than his own.

(N.B. The lacuna in the Elph. MS. ends before the rāq tūshūbtūr of the passage under discussion.)

⁶² Andāq qawūn ma'lūm imās kim 'ālamdā būlghdi, a characteristic idiom.

⁶³ Pers. trans, gawazn. So too Ḥ. MS. beneath the word būghū. Cf. fol. 3b and note, fol. 4 and note.

⁶⁴ Sū, here and in some earlier instances seeming to be a commonnoun. It is used in Turkistān as we use “water” in “Allan Water” and “Water of Leith”.

⁶⁵ Sā'i. Leyden (B.M. MS. trans.) and Erskine have read this as Pers. sāya, and have translated by “entirely in the shade” and “are sheltered along the banks of the stream”. I.O. 217, fol. 4, 1. 4, has sā'i.

⁶⁶ This Persian phrase has been found difficult of interpretation. It has been taken as follows:—

(a) Pers. trans. (I.O. 217, fol. 4), postīn pesh b( )rah.

(b) Pers. MS. quoted by E. (p. 6 n.), postīn-i mīsh burra.

(c)Leyden's MS. translation, “a sheepskin mantle for five lambs.”

(d) Erskine (p. 6), “a mantle of five lambskins.”

(e) Klaproth (p. 109), “pustini pisch breh, d.h. gieb den vorderen Pelz.”

(f) Kehr (p. 12), postin bīsh b( )rah.

(g) De Courteille (i, 9), “fourrure d'agneau de la premiere qualité.”

(g) Pers. annotator of Elph. MS. under the pesh or bīsh, panj.

(i) Ilminsky (p. 6), postīn bīsh b( )rah.

Erskine's five lambskins carry on the notion of comfort started by his previous sāyah. De Courteille also lays stress on fur and warmth, but flowery gardens bordering a torrent seem less likely to prompt a phrase emphasizing warmth and textile softness than one bespeaking ornament and beauty. If the phrase might be read as postīn pesh-perā, what adorns the front of the coat, or as postīn-pesh-i burāh, the fine front of the coat, the gardens would be allowed to recall the gay, embroidered border of a leathern postīn. Cf. von Schwarz's plate, p. 9.

⁶⁷ Shaikh Sulaimān (Konos) explains this as the tamarisk; if this it be, it seems likely to be the Tamarix gallica (Brandis, , Indian Trees, p. 45Google Scholar, and Balfour's Cyclopœdia). Shaw (Vocabulary), “a mountain bush;” Redhouse, (a) a tree of the buckthorn tribe, (b) the red willow, Salix purpurea or Salix rubra, (c) sappan-wood, the wood of the Cœssalpinia sappan. A rod-like plant such as the red willow would suit the several uses of it mentioned by Babar. “Tabalghū has the same meaning as tabarkhūn or ṭabarkhūn. See Vullers, i, 420b, and Meninski, i, 1030, and ii, 3084, s.n., who quotes the Lughat Ḥalīmī and the Lughat Ni'matu'l-lāh. See, too, Rieu, Turkī Cat., pp. 137, 142. It is the Hyrcanian willow” (H. Beveridge).

⁶⁸ Erskine (p. 6), “They also cut it into forked tops of arrows;” de Courteille (i, 9), “On la taille aussi en fleches.” Steingass, s.n. gīz., “a sort of arrow or dart without wing or point, the two ends being small, the middle thick,” a description allowing the scraping (tarāsh) of the Turkī text. Bābar distinguishes the tīr-giz from the auq.

⁶⁹ Tabarruklūq bīla yarāq yīrlār kā īlītlār. Erskine (p. 6), “It is carried to a great distance as a rarity much in request;” de Courteille (i, 19), “On le transporte au loin, où il trouve un débit avantageux.” The text allows the statement that the trees (yīghāch) are carried afar, and this would allow the word yīghāch to be translated all through the passage by “tree” instead of both by “tree”and “wood”. But if the tabalghū were rod-like, a statement about its wood would slip easily into the plural form. The Burhān-i qāṭti' includes the tabarkhūn, the uses of which suit the tabalghū.

⁷⁰ Yabrūja'ṣ-ṣannam, “the mallow consecrated to idols” (Leyden). “The plant called mandragora or mandrake. See the Ulfaz Udwîyeh or Materia Medica of Noureddin Muh. Abdalla Shirazy, published with a translation by Gladwin, Calcutta, 1793. The name aikoti is derived from the Turkī “(qy. Arabic)” word ayek, vivacity, and (Turkī) ot, grass. Mehergîah seems to be merely the Persian translation of the name, from meher, affection, and gîah, grass. It is, however, called atikoti or doggrass, a name which comes from the way in which it is said to be gathered. They have a fancy that any person who plucks up this grass dies; on which account they are said to dig round its roots, and when these are sufficiently loosened, tie it to the neck of a dog, who, by his endeavours to get away, pulls it out of the earth. See D'Herbelot, art. Abrousanam and Astefrenk. The same story is still told.” The mihr-gīyāh (Mandragora officinarum, love-apple) is mentioned in the Ḥadlqatu'l-aqālīm of Murtaẓa Ḥusain Bilgrāmī (Pers. lit. ed., p. 426). Cf. Asiatic Quarterly Review, January and April, 1900, art. Garden of Climes, H. Beveridge. Worldwide superstitions have prevailed and still prevail about the mandrake; some are preserved in English villages. Cf. Genesis xxx, 14, and Song of Solomon vii, 13. De Courteille translates iq-oti by “l'herbe aux ours” and mihr-gīyāh by “l'herbe d'amour” (i, 9).

⁷¹ Seven Villages. Mr. Ney Elias has discussed the location of this place (T.R., p. 180 n.). He mentions that it is placed in Arrowsmith's map of 1878 as a district of Kuratna, in the elbow of the Sir. The Bābar-nāma narrative where Yītī Kint is mentioned allows of Arrowsmith's location. Other names of similar form suggest, like this one, that the numeral in them denotes so many villages served by the same water. Biskent which is in the neighbourhood assigned to Yītī Kīnt, may mean Five Villages.