ÁñÁ«ÊÓƵ

When The Whole Earth Catalog, the classic alternative guide to all things American, gave it a mention, the review said simply "Good old Scientific American". The formula may be a familiar one - Scientific American has not had a full-scale redesign in the almost 30 years I have been looking at it - but it seems to work. One hundred and fifty years old in 1995, and claiming to be the oldest continually published magazine in the United States, it sells well and always has at least one or two must-read pieces per month. Its September special issues are definitive, and this year are joined by a look at key technologies for the 21st century; while the Scientific American books add up to a rich - if sometimes excessively orthodox - library.

Naturally this big birthday book, Triumph of Discovery, is an essential part of the sesquicentenary. So it is a pity that the chance has not been taken as enthusiastically as it might have been.

The strange thing about Triumph of Discovery is that its faults are the very ones that Scientific American makes it its business to avoid. For one thing, it is a mess. Somebody had the obvious and valid idea of putting into the book a chronology of scientific achievement over 150 years: and then made the mistake of scattering it through the book in five-year bites instead of concentrating it at one point in the book to make a single fascinating read. The reader wants to see the sweep of discovery from the era of Mr Otis, the elevator inventor, to that of the Challenger space shuttle and its disaster. To offer it in teaspoonfuls is to diminish the main point of the book. (The chronology's insistence on the Manhattan Project being an all-American affair is also inaccurate and chauvinistic.) Something that can always be said in favour of Scientific American is that anything it publishes will have been edited well and its significance made clear to the reader. But this does not apply to Triumph of Discovery. The principal editing device used to order the book's material is the alphabet; so the articles begin with Addiction and Aging and end with Viruses and Vision. It is hard to imagine anyone sitting down to read this jumble of matter for pleasure.

And that is a shame, because despite the cunning camouflage there are many good things hiding here. Most of the essays cover only three pages, with large type and a fair few illustrations, making the book as rotten value in money terms as Scientific American itself is excellent. But this length is just right for a reader wanting a quick look at nanotechnology, the origin of life, or sleep, three of the more diverting entries.

Less useful are the chapters of which the editors are probably most proud -reviews of big subjects by mega-authors, including Carl Sagan on the solar system, Anne and Paul Ehrlich on population, Lester Thurow on industrial policy and Steven Weinberg on unified theories of physics. Any assiduous reader will have had plenty of chances to catch up with what these people think: a few more idiosyncratic or heretical pieces would have been more fun.

Alas, it is almost impossible to count the missed opportunities in Triumph of Discovery. For example, there is almost nothing about the history of the magazine itself apart from a note on its launch, on its relaunch in 1947 after it got into financial problems, and a mention of its 1986 acquisition by Holtzbrinck, the German publishing group that also owns Nature, sandwiched tastefully in the chronology between Chernobyl and Challenger. Most newspapers and magazines are prone to excessive anniversary celebrations, but this is taking things too far the other way.

Worse perhaps is the negligible use the book makes of the bottomless archives of the Scientific American itself. Apart from one or two pages and a smallish number of pictures, no attempt has been made to give the flavour of the Scientific American heritage to the book. I would far rather have read more of the way the magazine has written about, say, viruses over the years than a review article on the subject, even though the authors in this case, Robert Gallo and Howard Streicher, take a much more historical approach to the issue than many of the other contributors. (There seems to have been no systematic briefing of the authors on the right app- roach to this crucial matter of style.) Unlike some organisations, Scientific American is likely to be around to correct these failings in its 200th anniversary volume. In the meantime, your money would be better spent on the magazine than on the book.

Martin Ince is deputy editor of The THES.