As always, he presents his case with dazzling erudition. "It is difficult to think of any surviving evidence from the period", he cheerfully acknowledges, "which has no relevance to my theme." There cannot have been a more widely read historian or one with his capacity to impose order on so formidable a range of material. His technique, which he inherits from his tutor Christopher Hill, is to pile quotation upon quotation - printbites, we might call them - from books and pamphlets where beliefs or values or prejudices are wittingly or unwittingly disclosed. The method, he admits, can be "bad for the reader's digestion". Hill sustained the reader's attention by Marxist polemical eloquence. Thomas, a more dispassionate writer, holds it by urbanity and wit and by suppleness of language.

Even so, his approach has its difficulties. In The Ends of Life he portrays himself as at least as much an anthologist as a historian, as much a "collector" of material as an "author". Perhaps the pressure of self-effacement helps to explain a certain tentativeness, even at times (it seems to me) a want of clarity, in the argument. Or perhaps he senses the interpretative limitations of his method. There is a familiar objection to the mass democracy of Thomas's quotations. He separates the statements of past writers from their contexts and overlooks differences of literary genre. He has two reasonable answers. The first is that if he were to reconstruct the setting of every remark he quotes we would be here for ever. The second is that, if we find enough people saying the same thing over a period of time, the varieties of context in which they said them are of secondary importance. But when, seeking evidence of a modern tendency to equate our sense of self with our material possessions, he recalls a speech that Henry James has "one of his characters" deliver in The Portrait of a Lady, he does not have space to remind us that the speaker is the odious Madame Merle, whose opinion James's heroine rejects.