Here is a real knowledge of ground and, what is more important, _a careful estimate of time and movement_. I know nothing better in the reconstruction of a mediæval battle than this first-rate piecing together of evidence through common sense upon the flanking surprise movement executed by Simon de Montfort against Foix’s division of the enemy at Muret. It is an unbreakable chain of calculation, and at the same time a full explanation of what happened. This piece of work, in the fifth chapter of the volume here presented to the reader, is as good as anything can be of its kind, and an excellent representative of that new, modern, accurate work now ridding us of the loose stuff which encumbered history through the past two generations. That is the way to reconstruct a mediæval battle in the absence of detailed evidence, to see the movements as they actually took place. I have laid emphasis on this particular section of the book by way of contrast to the insufficiency of so typical a name as Freeman’s. I ought rather, perhaps, to turn to the book as a whole and then again to certain other specific points of excellence which have struck me. Mr. Nickerson’s study is mainly concerned with explaining the nature of the early Inquisition; incidentally he gives us a very clear view of the Albigensian War, and what is especially remarkable in the clarity of his view is the arrangement of the episodes. I note that the author has done what is of first importance in all military chronicling, and that is, the division of episodes _not_ in equal measures of time but by their separate military characteristics. It is a principle too often forgotten even by professional military historians. A war may take twenty years, or fifty, or one. It may, by accident, divide itself naturally into two or three episodes of fairly equal length in time or it may by coincidence fall into episodes corresponding more or less with a successive series of years (e.g.,) Marlborough’s Campaigns in Flanders in the early eighteenth century. But much the greater part of military history is concerned with episodes which have no relation to such more or less equal time-chapters. The general rule is that three or four successive phases of a campaign (or battle) occupy the most disparate lengths of time. The proper way to treat _military_ history is to give to the capital episodes their relative _military_ importance; not, as in the case of a civilian chronicle, to weigh that importance by the time involved. For instance, no one can read a clear account, however short, of the great European War without seeing it as a siege; it is therefore, like every siege (not raised, nor degenerated into a blockade) essentially divided into three episodes:— (_a_) The preliminaries of containment, that is the war of movement prior to the establishment of siege conditions. (_b_) The siege itself. (_c_) The storming of the siege line and the collapse of the besieged.