In one respect it shouldn't be too hard, since texts describing food are detailed, from the "Life of Luxury", a poem by the Greek Sicilian Archestratus 350BC, to the late Roman Empire recipe collection under the name Apicius.The main ingredients are not mystifying (fish, birds, meats, olives, cabbages, grapes, barley for bread, garlic, cheese). We are familiar with the techniques (roasting, stewing, grilling, and much use of pestle and mortar to make sauces). The one snag is that proportions of ingredients are not specified. The modern cook can only guess."What is very clear to me is that Roman food was strong-flavoured and very exotic," says Ms Grainger.

It is a collaboration with Cambridge classics scholar Andrew Dalby, and its eight chapters encompass a span of 1,200 years, from 750 BC (the time of Homer's Odyssey) to AD450 (a supper at the Roman baths).It is a daring project. Now she has published the results of her efforts, a book of 50 recipes adapted to the modern table, The Classical Cookbook (British Museum, pounds 14.99). The history of technology shows that until the late 18th century science had no impact at all on technology. The industrial revolution with the invention of the steam engine owed nothing to science.

The bicycle, too, has not been designed using basic science - its success depends on the insights of clever mechanics. Today it is all quite different. With so much technology dependent on science, the list is long, from the transistor to the very successful pharmaceutical industry. Kealey's case rests upon his assertion that there is no simple linear relation between the funding of basic science and wealth creation. Research does not necessarily lead to new and moneymaking inventions Here, I regret, I have to agree with him. There are certain topics I will not discuss either in public or in private. Some ideas are so silly that to discuss them implies that they are to be taken seriously.