Sir Banister Fletcher's A History of Architecture, Twentieth and Centennial Edition.

The American Institute of Architects was quite right recently when it made Banister Fletcher its Book of the Century. The tome has been bought for 100 years by every architectural student who could posssibly scrape together the shillings. It has no rivals in its marvellous analysis of the great works of the European past. If you want to compare St Peter's to the Pantheon or Hagia Sophia, you turn to Banister Fletcher for the measured drawings and a succinct description.

The problem for its successive editors has been to keep up Fletcher's scholarly rigour and extend it. The first (and many subsequent) editions were what we would now consider almost racist. The history of architecture ran as a tree from its earliest roots in Egypt through the Greek, Roman, and Romanesque to be crowned absurdly with the American. My edition (the 16th), purchased at the same time as my drawing board with what seemed to be a huge amount of money on my first day at architecture school, has all the great descriptions but falls about a bit in the twentieth century, when Sir Banister must have had a glass of port too many as he finished his final revision on 'Coronation Day 1953' and invented an architect called Frank Gropius (p869), whom the author could not find space to show (with Le Corbusier and many others). Partly as a result, all we young up-and-coming Modern architects thought of Fletcher with some disrespect (but we never dared to ignore him, and every trip to Italy or France was organised after extensive study of the great book).

Dan Cruickshank was an obvious editor for the centenary edition for his lucid prose, respect for tradition and knowledge of the contemporary scholarly scene. He may have made mistakes but I am unable (as yet) to find them. (Like the proper gent that he is, he takes all blame for errors, and yet carefully explains the contributions of his numerous collaborators). The new edition builds on the 19th one (edited by John Musgrove) in which Fletcher's original structure was somewhat changed to make the book less Eurocentric. But it still keeps a reasoned description of the relationship of architectures to the societies that created them, and the resources of material and technology that made them possible. Fletcher's great work can still be seen as a product of William Morris's teaching that architecture and society are indivisible and always interactive.

Cruickshank's edition is much the biggest yet (the work is almost at a stage when it should go into two or even three volumes and have a concise Sir Banister for the pocket). Much of the new work is about the twentieth century and explores areas that have not been covered properly elsewhere (for instance China from 1900 to 1950 and Latin America in all of this century).

All this is good, but you long for magisterial drawings to back up photographs and written descriptions. Why for instance waste half a page on a rather faded plan of Foster's Hongkong Bank without scale or north point? Fletcher's own drawings (most of which are used in this edition) are exemplary and were made for what he called 'A history of architecture on the Comparative Method'. In the next edition (which must surely be edited by the excellent and explorative Cruickshank), the publishers should find resources to show twentieth-century work in comparable ways to the buildings of previous ones.

It has faults, but no architect or student can afford to be without it. In our hundredth year, we hail an old friend and exact contemporary that has always informed this magazine - and sometimes inspired it too.