Mahāyāna Buddhism: the doctrinal foundations . By Paul Williams. (The Library of Religious Beliefs and Practices.) pp. xii, 371. London and New York, Routledge, 1989. £30.00 (cloth), £9.95 (paperback).

that are the best represented in Varanasi. Their performance practice includes, unusually, all the traditional vikrtis (verbal permutations) plus some other rarer ones peculiar to that school. Here Howard shows that there is no significant relation between accent and recitation tone; he suggests (p. 84) that the Madhyandina chanters have cultivated the particularly large number of vikrtis precisely because accent is not followed in the chant, and so another form of mnemonic is required for the exact preservation of the tradition. In the Samavedic chant, most practitioners of which in VaranasT are of the Kauthuma branch, Howard sets out to interpret from first principles the numerical notation of the texts. His achievement here is considerable, especially if we consider the difficulty he experienced in gaining access to the practitioners' own explanations. Howard is sufficiently secure in his reading of the notation to be able to give a transcription, without reference to an actual performance, of the Gayatri, of which he was dissuaded from even requesting a recorded version, so reluctant are the northern Kauthumas to chant this saman for an outsider. All musicologists of India will be indebted to Howard for his meticulous gathering of material and painstaking transcriptions. Some parts of his musicological analysis will be found tantalizing. His brief speculations on the compositional process of saman chant rest on a notional repertoire of melodic elements and phrases or "motives", which he extracts and categorizes from the recordings. The method by which these motives are combined and structured is likened to that of the melodic components in the Hindustani aldp and tan, and as an example Howard cites the svaravistara given by V. N. Bhatkhande for a rdga in his pedagogical work, Hindustanisahglta paddhati. A great deal more needs to be discovered about the melodic form of saman in order to make this comparison intelligible. Are we really dealing here with improvised or formulaic structures, later fixed, or with melodic compositions which predated the texts to which they were fitted? Our thoughts about such questions may be assisted by examining the occasional comments on chant and melody in the ancillary saman and other Vedic texts, but we shall have to be careful to distinguish the quasi-analytical statements of early theory from likely historical reality. Above all, what is the nature of the relationship between saman and the other early musical traditions of India, and are there any detectable melody types analogous to the jdtis described in the early treatises? Howard's comments in this area of musicology are really only peripheral in a work which provides much valuable source material and which should inspire further investigation.