After disposing, somewhat too neatly, of most of the historically popular ""explanations"" of what makes war break out, Blainey (Univ. of Melbourne, Australia) launches his own theories based on an empirical study of every international war fought since 1700 when Sweden took on Denmark. His thesis rests on the premise that ""war itself is a dispute about measurements"" of international power; fighting erupts when a nation's ""diplomatic currency is overvalued""; peace on the other hand ""breaks out"" when the combatants have reached rough agreement about their relative strength. This may not sound very startling, but Blainey presses onward for more ""clues"" and comes up with some disconcerting evidence. The first dogma to be jettisoned is the notion that ""a balance of power"" is the best guarantee for international amity. On the contrary, says Blainey, the empirical record shows that ""a clear preponderance of power"" is more likely to inhibit conflagrations. Similarly, the theory that a ""harsh"" peace treaty sows the seeds for future wars doesn't hold water. Versailles notwithstanding, decisive wars and tough postbellum settlements hold up longer -- though probably no longer than one generation. Furthermore (and here Blainey scores methodologically), analysts of international relations have mistakenly focused their attention on the ""origins"" and the commencement of conflict; they ought to pay more attention, he says, to the end of wars -- ""the only points in time when the distribution of power can be measured with some objectivity."" Blainey's methodology, which rests on examples and counter-examples drawn impartially from mini and macro conflicts, will draw fire; so will some of his conclusions which tend toward the gloomy war-has-always-been-with-us-and-always-will-be view. But you have to admire the man for even attempting a project so ambitious and iconoclastic.