Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum . xiv, 529 pp., map [on endpapers]. Wiesbaden: Dr. Ludwig Reichert Verlag, 1989. DM 280.

section on Azeri Turkish, rather oddly excludes Soviet Azerbaijan, and Kleiss's section on archaeology seems to justify that even more oddly by his qualification, 'which since 1829 has belonged to the Russian empire'. Though one cannot expect comprehensive illustration it is noteworthy that the contributors, with the exception of A. S. Melikian-Chirvani's 'Banners' and 'Bargostvan' (horse armour), generally ignore the pictorial aspects of their subjects. Thus the very comprehensive article on dragons, ' Azdaha ', has practically nothing on their astrological, astronomical or iconographical importance; while J. Calmard's discussion of mourning,' 'Azadari', says nothing of the paradoxical canonization of photographs in cemeteries at the graves of those buried there. On the other hand, music, probably for the first time in a reference work on the East, gets the importance it deserves. The criteria for the inclusion of articles are not always easy to fathom. St. Augustine is included on account of his Manichaeism, though it is conceded that that was not Mani's original exposition but a form reconciled, _ in certain respects, to Christianity. ' Aykanom ' is a useful account of the Franco-Soviet excavations but has little apparent of relevance to Iran. Nor does ' Balawaste' in the Khotan oasis, with paintings of c. A.D. 600 of great importance for Khotanese art, seem to be of central relevance to Iranian studies. Absolute consistency in an enterprise of this scale is scarcely to be expected, however, and it is more helpful to suggest its strengths, which combine to make El an essential work of reference to those even tangentially concerned with Iran. It is particularly strong on the nineteenth century, not merely on Bahaism, which mostly falls within this volume, but on Qajar princes, officials and personalities. Of these the most amusing and scandalous was 'Aziz al-Sultan (s.v.) (1879-1940), known as Mahjak TanI, who became the favourite of the besotted Nasir al-DIn Shah and who was apparently introduced to his court as a substitute for the Shah's favourite cat, Babri Khan, which had fallen a victim to jealousy of its honorary Keeper. Other subjects thoroughly treated are the press, banking, planning (' Barnama-rlzr), electricity (' Barq'), bathhouses, sanctuary (' Bast'), and climate (' Bad ', 'Baran' , 'Barf '). In this last case the author critically remarks, ' Iranian material culture shows many signs of failure to adapt to snow '. The articles on tribes in Iran are also impressive, though' Baktiari' is probably typical in treating their history in the form of biographies of influential Khans and then concentrating on ethnography, dialect and carpet-weaving while ignoring e.g. religion. The flora and fauna include an exhaustive zoogeographic and taxonomic survey of bats and oaks (' Balut') and notices of crops or plants of economic, culinary or medicinal importance—bamya (' Bamla'), violets (' Banafsa '), barley, aubergines (' Badenjan '), almonds (' Badam'), etc. Most of these survey lexicography, distribution and their modes of cultivation, their uses in folk-medicine, pharmacology and cookery, sometimes giving characteristic recipes: the most fascinating article is that on hashish (' Bang'), which in view of the strict anti-hashish legislation in Iran must have put its author at some risk. The most powerful form is apparently hashish oil, generally taken with yoghurt and known as dug-e wahdat (' dug of annihilation') for its often fatal effects which can only be counteracted by excessive quantities of animal fat. Annihilation sounds a better prospect. ' Bang' is typical of El in that the volume is rather richer in curious information than EP. There are articles on Bakhshiev Mishi (d. 1972), the author of the first heroic drama in Judaeo-Tat, the language of the mountain Jews of the Caucasus; ' Bax faldisin ' (horse dedication, funeral rites of Iranian origin practised by the Ossetes till recent times); the incidence of favism in Mazanderan and Gilan (' Baqela', broad beans); and the final answer to the puzzling problem of Barthol'd's Christian names (baptismal ' Wilhelm', nom de plume ' Vassily Vladimirovich') in an excellent biographical article by Yu. Bregel'. A rapid survey establishes that to gain the most from El the reader will often require a knowledge of modern Persian. Partly the authors show a preference for American plantnames over English or Linnean terms (' yokeelm ' for hornbeam, Carpinus betulus; ' jimsonweed ', for thornapple, Datura stramonium, etc). Partly the choice of English or Persian for titles is not always consistent. The preface indicates that at some point when El was already under way the decision was made that the entries should be in alphabetical order, ' Bread ' before ' Nan' , ' Berenj' before ' Rice', etc. The problem comes when the Persian term has more common Arabic equivalents. Thus ' Barda and bardadarl' is evidently intended to be the principal discussion of slavery, eunuchs and the ghulam system corresponding to 'A bd, Ghulam and Khadim in EP. Without copious crossreferences such important articles will be difficult to find and, as is generally the case in reference works, the references forward are rather less detailed (because more difficult to foresee) than the references back. This means that articles like' Barg-e Bu' (bay tree) will only be discovered by the browser, or the present reviewer. It is difficult to suggest how to improve the cross-referencing, but it is to be hoped that the problem may be more efficiently solved in subsequent volumes. The list of addenda and corrigenda for Vol. m of El, is unexpectedly large and much larger than in the first two volumes. Practically all are insignificant but many other misprints (also insignificant) remain uncorrected in the text. The only inconsistency of substance I have detected is the translation of Turkish baraq in ' Barakzl' a s ' hairy dog ' and in ' Baraq Baba' a s ' hairless dog' . Both could conceivably be the case, but Clauson citing Khaqani gives only ' hairy dog ' so that I assume the latter is a slip.