Amathematician is a machine for turning coffee into theorems,' said Paul Erdos, one of the strangest, yet acknowledged as greatest, mathematicians of the century - if not the most prolific. He lived only for mathematics, having few possessions or even his own home, preferring to wander as a guest from one mathematician's home to another.
Uninterested in the normal trappings of life - money from publications or speaking engagements was simply given away - he eschewed even sex, preferring to spend 19 hours a day solving problems. He lived on amphetamines and coffee as he did so. By the time he died, aged 83 in 1996, he had published more than 1,500 papers and his enormous contribution to a whole range of mathematical problems was undisputed.
Other mathematicians pride themselves on their Erdos number. Erdos number 1 is the coveted code for having worked with Erdos himself. An Erdos number of 2 is earned by mathematicians who have published with someone who has published with Erdos and so on, to infinity - which is the Erdos number for non-mathematicians. Einstein had an Erdos number 2. There is even an Internet site run by two mathematicians at Oakland University, Michigan, which tracks the coveted numbers.
In sum, so to speak, it seems almost every mathematician in America and many in Europe have been influenced by Erdos.
The tongue-in-cheek game of Erdos numbers gives some clues to the way this book is written - from the stories of those with high Erdos quotients rather than family members or those who 'merely' remembered him.
Rather than a linear biography from birth to death, the book careers off on a series of tangents. There is very little about Erdos' parents or childhood in Hungary, or even his death. Instead Paul Hoffman builds a convincing picture of Erdos by drawing amusing portraits of an array of other mathematicians and their work - and not all of them had an Erdos number of 1. Hoffman has said elsewhere: 'I use Erdos' life story to introduce non-mathematicians to a cast of remarkable numerati from Archimedes and Pythagoras to Stanislaw Ulam, one of the chief minds behind the Los Alamos nuclear projects.' It is almost as if Erdos is too complex and too eccentric to be tackled as a subject. It is far easier to collate the impressions of those who knew him, although Hoffman did interview Erdos over the last 10 years of his life, even following him on mathematical sojourns.