Firstly, I accept Beeston's proposal to refer ‘to linguistic data as Minaic, Sabaic, Qatabanic and Hadramitic, leaving Minaean etc. for non-linguistic contexts’ (Beeston, 1987a: 13). His argument (at least with respect to the distinction between the ‘Sabaean’ and the ‘Sabaic’) looks entirely convincing:

The Sabaeans were the inhabitants of the area around Ma'rib. But they shared a common language(no doubt with dialectal varieties) with the peoples of the desert fringes between Ma'rib and Wadial-Jawf, of a considerable part of Wadi al-Jawf itself, and of the high plateaux lying in the angle of those two areas—Arḥab, Qā‘ al-Bawn, Qā‘ Ṣany‘ā’ etc. It is dubious whether these other peoples would have called themselves Sabaeans,1 and it might be preferable therefore to term the language Sabaic rather than Sabaean’ (Beeston, 1979: 115).

Given the dependence on aural transmission in the pre-modern musical traditions of the Islamic Middle East it is not at all surprising to find that compositions were hardly ever notated. If there is a typical method of recording the repertoire it is to be identified, rather, in the song-text anthology, which may provide quite detailed indications of formal structure but contains no melodic specification beyond identifying the principal mode of the piece and serves, therefore, as a textual prompt. Indeed, if we disregard as wholly exceptional the rich documentation of the Ottoman repertoire provided by the extensive collections of notated pieces made by ‘Alī Ufḳī and Cantemir, we find for the seventeenth and earlier centuries just one or two not very informative craps recorded by European observers and the occasional samples provided in theoretical treatises.

Tughluqabad is the first of many sultanate and Mughul towns which were purposely planned and constructed on previously uninhabited sites. Built early in the fourteenth century, Tughluqabad was to serve as the capital of the newly established Tughluq dynasty. There were, of course, three earlier Muslim capitals in the vicinity, the first the Delhi of Rāi Pithūrā, converted to an Islamic town after the Ghurid conquest in 588/1192–3; the second Jalāl al-dīn Khahīs Shahr-i naw, which was founded by Muՙiẓẓ al-dīn Kai Qubād (685–8/1286–9) at Kīlukharī (or Kīlugharī) but left incomplete at the time of his death, and the third Sīrī, built by Alՙ al-dīn Khaljī between 698/1298–9 and 700/1300–1 in the fields outside the walls of the older Delhi, but nothing has remained from these towns except parts of the fortification walls and some isolated monuments. The ruins of Tughluqabad, on the other hand, are enshrined in a time capsule. Built between 1320 and 1325 by Ghiyāth al-dīn Tughluq, the town had a brief life, and within a generation was abandoned and its population reduced to the size of a small village. As a result, most of its remains are datable to the short period of its duration in the first half of the fourteenth century. The only exception, as we shall see, are the remains of a small settlement which continued to exist around the old town centre, and in the late Mughal period also occupied the citadel.

The main subject of this paper is the influence of Western science on the philosophy of the young scholar Tan Sitong (1865–98) who was active in the reform movement of the late Qing dynasty. It seeks to show that his understanding of Western science was not as confused and superficial as some commentators have claimed, and that in fact his conception of yitai drew on views of the ether and electricity which were held by some at least of the most eminent Western scientists of his time. It also demonstrates that yitai is the direct descendant of the qi of the Neo-Confucian philosopher Zhang Zai (1020–77), and shows how Tan developed the views of Kang Youwei (1858–1927) on the relationship of qi to consciousness.

Among the Tibetan texts in the Sven Hedin Collection now preserved in the Ethnographical Museum (Folkens Museum), Stockholm, there are three documents which are considered to belong to the Old Tibetan period. Their photos were first published in Bailey (1973), together with comments on a few Khotanese-related words, the texts being numbered Hedin 1, 2 and 3. But since then, they have not been paid due attention. In May 1990, I was able to examine these three manuscripts by courtesy of the Folkens Museum and Professor Staffan Rosén, secretary of the Hedin Collection.

On examination, I realized that all three texts are contracts, but of different kinds: that is, Hedin 1 is a sale contract, Hedin 2 a loan contract, and Hedin 3 a contract of hiring. Each exemplifies one of the three major types of Old Tibetan contract. Although different in kind, they show a close resemblance in both form and content: they are written on paper of similar size and quality. The left edge of each document is torn off in a similar way, but my attempts to join them together showed that they were not likely to have formed one piece. Paleographically, the three texts are written in similar styles, though apparently by different hands. They also have in common somenames of the persons concerned (cf. §IV). All these shared features clearly suggest that these three texts, even though they were not originally one, belong to the same period and location.