as they had been understood in the past, no longer responded to the needs of warfare, that produced the multiplicity of designs and proposals for artillery works and infantry works, armoured and unarmoured works, self-contained and mutually inter- dependent works, and so on, characteristic of the period of un- settlement. If most of this ingenuity remained, as it did, un- convincing, this was due to the fact that there were great general causes at work, of which, in default of war experience, only the existence and not the effects could be seen.
The size of armies steadily increased, all the European conti- nental Powers being drawn into a competition based on the numbers of citizen soldiers who could be conscripted and finan- cially maintained. The power of armament increased also, and with it the possibility of holding a wider front per unit of armed force. The special results of this, from -the point of view of fortifi- cation and siegecraft, were the extension of perimeters and the thinness with which a circle or arc of investment could now be maintained. But the more general results were the more im- portant. An army developed along a front of some hundreds of miles could no longer be worked by radial lines of communication centring upon one or two ring fortresses. In Napoleonic practice, a stronghold of some sort was always the centre of operations on which the army's movements pivoted; and only that portion of the theoretical base-line, which was in relation with the strong- hold in use at the moment, formed what has been called the " effective " base. With the modern extended fronts, on the contrary, the effective base has-widened more and more, until it practically coincides with the theoretical base. In other words, each part of a great army has its own lines of communication and its own sources, the connexion between the army line and the base line, or front edge of the home supply area, being a sheaf of more or less parallel routes. Whatever local variations may appear when portions of the system are isolated and examined by themselves, in the ensemble the strategic structure of warfare in civilized countries had become linear.
But the more the front extended, the more difficult it became to collect any considerable force at one part of it for offensive effort. The " parallel battle in all its horror " unit facing unit all along the line was admitted to be the negation of general- ship. But whether the dispositions were made a priori or by manoeuvre within the battle, whether envelopment or break- through were the method chosen, the parallel battle could only be avoided by reducing the living forces on certain passive or semi-passive parts of the line to a minimum. The small econo- mies that could be effected on these parts by judicious tactical arrangements in the open though certainly not to be despised would no longer suffice to give the really considerable superiority of force necessary for decisive victory on the active front. The expedient of economizing force by sacrificing territory had be- come, under modern conditions of social and economic life, more dangerous than it had ever been before. Recourse was had, therefore, to fortification. It became one of the roles, if not indeed the principal role, of permanent fortification to economize active living forces on the passive fronts a principle already applied to field fortification within the tactical sphere. As soon as the competition in numbers had set in all over military Europe, we find permanent fortification developing a new tend- ency to be linear instead of circular in type. The ring fortress becomes a sort of end-redoubt to a long line of forts, usually drawn along some natural barrier. This tendency is shown in the creation of the Meuse line (Verdun-Toul), and the Moselle line (fipinal-Bclfort) in France, the line Namur-Huy-Liege in Belgium, the Sercth line in Rumania, the Bobr-Narew-Bug line in northern Poland, and lastly the Dicdenhofen-Mctz and Molsheim-Strassburg-Istein lines in Germany. In the same way, most of the new ring fortresses that were not so connected by permanent works were so placed as to be keystones of a linear battle-system; conspicuous instances are the systems Lille-La Fere-Laon-Reims-Verdun and Dijon-Langres-Epinal. Of the rest setting aside the fortresses of eastern Europe where poverty of communications enabled permanent works, as of old, to dominate great areas by dominating a few nerve-centres we
find Antwerp, Copenhagen, Bucharest, Paris, designed as self- sufficing ring-fortresses, but not so much for playing a part in a battle-system as for serving as refuge for an army and a govern- ment that for the time being could not maintain its line of battle in the open field. Three of these four were called upon in the World War to play the assigned r61e and it is significant that not one of them did so successfully. Bucharest was evacuated as the result of an unsuccessful linear battle in the foreground; Antwerp was given up by the Belgian field army and Government as soon as the choice had to be made between standing a siege and con- tinuing the war in the open; while in. the crisis before the Marne the evacuation even of Paris was seriously considered. 1 Two ring-fortresses, Maubeuge and Bcsancon, were constructed in France in advance of the battle-system, in order apparently to draw upon themselves a part of the invader's effort condensed on the wings, and to control certain nuclei of communications which might otherwise be useful to him as he pressed forward. But the latter task is really that of a barrier-fort, and indeed Maubeuge was in process of being converted into a pure barrier position when war broke out, the old intention of an isolated defence d, entrance having been abandoned. The linear systems with end-redoubts, on the other hand, performed in the war all the services for which they had been designed, with the exception of Liege-Namur; and even in this case failure so far as there was failure was due not to any vice of principle but to other factors.
The tendency to force a speedy decision in battle at all costs, specially characteristic of citizen armies, could not but rein- force the effects that the size of armies in itself produced upon fortification. With such a tendency on both sides, the initial deployment on each side would inevitably be carried out, wholly or largely, in accordance with an a priori plan of battle. Stra- tegic considerations for the side which had chosen the pure offensive and moral and political considerations for the side which had chosen the defensive-offensive, imposed concentra- tion close up to the frontier, in the first case so as to seize the initiative, and in the second so as to surrender as little of the national territory and resources as possible. Frontier fortification therefore had as its first duty protection of a line or zone of railheads close behind it; and since railway communica- tion is in principle highly sensitive, a system of ring-fortresses at intervals could not give the same protection against sudden raids as linear defences of equal trench-length. But there was a further consideration. An a priori scheme of battle, with frontage and not depth as its main characteristic, is liable to require con- siderable modification when contact has been made and the first serious combats have produced their varied results, and thus a regrouping process begins in the course of the operation itself. In this regrouping, fortification is called on not only to protect the lateral shifting of masses by rail (as for instance the moves of the French IX. and XVIII. Corps from eastern Lorraine to the Ardennes in the middle of Aug. 1914), but also to send away its own local reserves to the area of decision (as in the case of the three French divisions transferred from the Meuse-line front to the Somme at the end of Aug. 1914).
The more penetrable the country, the more pronounced the linear character of the fortifications that must cover it. Not knowing the direction of attack, the defender must cither prepare for it at all points of his allotted frontage of influence, or else resign himself to giving up country that ex hypothesi is economi- cally valuable, and manoeuvre in retreat to gain time. The old policy of devastating a deep zone to cover manoeuvre, occasion- ally practicable when the organization of the state was simple and predominantly agricultural, is almost or wholly inapplicable in an industrial country. In Oct.-Nov. 1914 Hindenburg devastated part of W. Poland as cover for a lateral regrouping. But when it came to including Upper Silesia in the devastation programme, industrial influences promptly intervened to mitigate it. The
1 In the event Paris played the part, originally assigned to Laon- La Fere, of end-redoubt in the battle-system. Antwerp, after cap- ture, was organized by the Germans for the same purpose; viz. to serve as end-redoubt to the Antwerp-Meuse line.
