165,000 entries, with 30 Oxfords

The use of the term gazetteer to describe a geographical index or dictionary goes back at least 300 years, though the word has also been used to describe journalists, writers of news and even dirty newspapers.

An early gazetteer was produced at the end of the 17th century by L. Echard and bore the title The Gazetteer's: or Newsman's Interpreter: Being a Geographical Index . However, the golden age of the geographical gazetteer was the Victorian era. The British produced innumerable gazetteers of their Indian empire to provide fundamental data for military men, missionaries, administrators and traders. They are still an indispensable source of statistical and other information and are a mine that could be exploited to good effect by geographers and historians far more than they have been.

British commercial publishers also produced some notable gazetteers. One major example of the genre was compiled by the great Edinburgh cartographer, John Bartholemew, while George Goudie Chisholm produced another for Messrs Longman. Both these examples were huge and detailed. In an age of revolution, exploration and colonial expansion they reflected the need for basic data on a changing world.

Some American publishers were also active in this field, and the most notable example was Thomas Baldwin's Pronouncing Gazetteer (1845) which in 1855 became the Lippincott Pronouncing Gazetteer, and in 1952 was renamed the Columbia-Lippincott Gazetteer of the World .

Forty-six years on from that edition, Saul B. Cohen has edited a radically updated and expanded version and the name Lippincott has been dropped. The work comprises three fat volumes, 3,616 pages, boasts a supervisory board of 150 geographers, and contains information on 165,000 places (an increase of 30,000 over the previous 1952 edition).

The range of places covered includes entries on the political world (countries, provinces, states, counties, regions, cities, towns, villages, etc), the physical world (regions, oceans, seas, lakes, lagoons, rivers, mountains, archipelagos, etc) and "special places" (national parks, shopping malls, nuclear facilities, and many more). Even quite small places are included, and there are no fewer than 21 separate entries for Andorra (which compares with 6,068 for Russia). Some entries are just a few lines, while others, where appropriate, are more substantial.

Guides to correct pronunciation are given, old and new names are often provided, and the latest statistical data and names are incorporated. This has not always been an easy task in that, for example, Lebanon has not had a census in over 50 years, and in 1997 during the production of the gazetteer President Kabila inconsiderately changed the name of Zaire to the sadly misnamed Democratic Republic of the Congo while Kazakhstan changed its spelling by omitting an "h".

As with all reviewers of reference books, I turned to obscure places of which I have recent knowledge - Swaziland, the Middle East, Central Europe and the UK - with the intention of checking the quality of information, its coverage, and its modernity. This experience left me deeply impressed, for every place I looked up was included, was accurate and was up to date.

Even though the Gazetteer is not in electronic form, one can pass the time of day happily in various quests. Among those I did was to select some British towns and to see how many settlements in other parts of the world had the same name. London only has nine entries while Bristol has 19 and Manchester has 23. Oxford has 30 while Cambridge (as in so many areas) has a lesser impact with only 25. The small and blasted Isle of Portland has 26, while big Birmingham has five and the great port of Liverpool ten. Plainly size is not everything. I have not, of course, been through all 165,000 entries, but so far of British towns Oxford appears to have the most desirable name for transplanting. The possible reasons for that need not detain us.

In sum, this is an extraordinarily full and detailed work of reference that has been compiled with skill and care. I believe it will be a fundamental source of reliable information for all reputable libraries until it is superseded or complemented by an electronic version. It is just the sort of work whose utility and interest could be markedly heightened by availability in electronic form. In its present manifestation it is magnificently large and far from cheap, and so may be the last of a line of such printed books that has spanned three centuries.

Andrew Goudie is professor of geography, University of Oxford.