This article examines the mythical significance of the famous Afghan Kafir ‘Temple of Imra’ described in Robertson's Káfirs of the Hindu-Kush (1896: 389–92) within the cosmology of the Kalasha (‘Kalash Kafirs’) of Chitral in northern Pakistan. It is known as the ‘Temple of Mahandeu’ in Kalasha tradition, and stories about this sanctuary play an important role in the exegesis of all Kalasha rites. It is, indeed, a focal symbol of Kalasha cosmology: the site of an axis mundi linking heaven and earth with the underworld of the deceased, and the primordial domain of major deities. After examining narratives about this temple, I shall discuss several problems in the comparative religions of the Hindu Kush that such traditions help to elucidate. In recognition of the pioneering scholarship on this subject by Wolfgang Lentz (1974) and Lennart Edelberg (et al., 1959), I present here some Kalasha perspectives on an extraordinary Kafir sanctuary (cf. Jettmar, 1986: 50–51). But in discussing its significance in Kalasha cosmology, I also address broader questions about our present conception of religious knowledge in the Hindu Kush, particularly on the comparative ‘mythology’ of the Afghan Kafirs and of their Dardicspeaking neighbours in northern Pakistan

This article examines the meanings and etymologies of several problematic Sogdian lexemes.I am very much obliged to Dr. J. Cheung (Leiden), Professor V. Livšic (St. Petersburg), P. Lurje (St. Petersburg), Professor M. Schwartz (Berkeley) and Professor N. Sims-Williams (London) who took pains to read the first drafts of this article and made valuable remarks, to Dr. B. Williams (Berkeley) who was very helpful in the matter of everything Chinese, to M. Ellsworth (Berkeley), A. Kassian (Moscow) and A. Kozák (Berkeley) who corrected my style. Final decisions are, of course, my own and none of those listed above are responsible for my errors. I am also grateful to the Houtan foundation (www.houtan.org) for its generous financial support, without which the completion of this article would hardly have been possible.

Burmese music is a virgin field for research. Nobody can bewail this virginity more than myself. To help me there does not seem to exist a single western work on Burmese music though there are such classics as Fox Strangways' on Indian and J. H. Levis' on Chinese music. Many writers on Burma have passed remarks, kind or humorous, on Burmese music. But few have attempted a technical study of it. Nothing Burmese escaped the notice of Sir James George Scott, of course. But I know of only two who left some technical remarks on Burmese music: Paul Edmonds, the author of Peacocks and Pagodas; and P. A. Mariano, in an appendix on Burmese music to that most lovingly produced book Burma by Max and Bertha Ferrars.

In a recent lecture to the Philological Society (now in the press, and to appear in the Transactions, 1947) I had occasion to quote a Central Asian ethnic name Ttaudägara–. The one passage where this name has been found is in the Khotanese manuscript in the Bibliothèque Nationale numbered P 2741. I had prepared this text for publication in 1942, but it remains still unprinted. I therefore print it here since a controversy may be expected over this name as a possible equivalent of Ooyápa, and it is desirable the basic evidence should be in the hands of scholars as soon as possible. A complete translation is hardly possible yet and only the relevant passage is rendered here.

Having been in the first instance a classical scholar, Wilfred Whiteley well knew how important vernacular texts are for an understanding of a people's thought and values. He was also that rare combination of anthropologist and linguist and saw therefore in texts much more than phonetic, grammatical, lexicographical, and other purely linguistic problems, but also the semantic, which in effect means ethnographical, questions raised by them. He saw clearly that a people's whole way of life, and the way they see it, may be revealed in a single text. Consequently, and apart from their significance for the ethnographer who recorded vernacular texts among (generally) a pre-literate people, they can be, especially if presented in both vernacular and with a competent translation, of considerable pedagogic value, the student asking himself such questions as arise from their content.

The following paper does not set out to be a comprehensive comparison of tonal systems—it is rather an account of my own experiences with various methods, with special reference to tone-marking in dictionaries and vocabularies. For this reason a certain amount of biographical detail is inevitable.

Emphasizing the importance of balanced geographical and historical knowledge in the late sixteenth century, Richard Hakluyt remarked: ‘Geographie and Chronologie are the sune and moone, the right eye and left eye of all history.’ In current studies of archaeology and history this emphasis remains apt, for to write a proper account of both artifacts and kings, they must be set geographi- cally in space and chronologically in time. The regions south of the River Yamunā, anciently known as Gopakṣetra, Daśārṇa, and Jejākadeśa (Maps 1 and 2), are rich in antiquarian remains, and have played an important role in the history of India, especially from the Gupta period to the time of the Muslim invasion. Their historical geography, however, has been generally ignored, and scholars have been content to describe the area simply as ‘ Central India’. The purpose of the present essay is to give an introduction to the historical geography of these provinces, and as such it can be taken as a foundation for further studies in local history and archaeology.