single factor, lies the central idea of group-fortification of the Feste type. Two dissimilar elements have to be both protected by the same obstacle and yet spaced some distance apart. But the obstacle (in such conditions mainly wire or grille) itself re- quires local close-defence. This " ultimate unit " has thus not yet been arrived at. Nevertheless, this ultimate unit, in group- fortification, has only to give short-range protection to the obstacle, and in practice it is an infantry-manned stronghold, designed to give fighting protection to its garrison, 1 sometimes provided for, its own local safety with a deep ditch and sunken flanking defences, sometimes organized with a fighting parapet frontally commanding an artificial foreground which is wired, but always having as its real function the protection of an ob- stacle external to itself.
In the case of concentrated main armament, therefore, it would seem that fighting protection for the counter-battery guns, for the traditore batteries, and for sunk ditch defences is required to be designed on such a scale as will enable these elements to defy, actively or passively, the attack guns of the day and the morrow. The same applies to the shelters in which in the case of group fortification the garrison of the infantry work is placed in readi- ness to man the parapets, but not necessarily to these parapets themselves. Further, in proportion as wire replaces the deep ditch, as an obstacle, heavy and expensive work in peace-time is dispensed with.
In the system of deployed main armament, on the other hand, the proportion of permanent work, it would seem, can safely be much less. With modern artillery means, the sites for counter- battery armament are rarely obligatory; observation must be provided for; but the actual position of the guns, and therefore the line of liaison between observation post and guns, are to a great extent at least free from limitations of ground. This being so, the close-defence element of the fortifications may be disposed to the best advantage for carrying out its task that of protecting a system of obstacles suitably placed between the battery zone and the enemy.
In point of permanent work, then, although parts of the bat- tery positions themselves may occasionally require concrete or even armour, concealment of virgin earth, and alternative posi- tions in the great majority of cases afford all necessary protection. For the close-defence guns, on the other hand the element which must be able to endure at all costs the chosen positions are often (if not in most cases) obligatory, and full-scale fighting protection must be given. Even so, there being by hypothesis no necessity to develop frontal fire, and the volume of the re- quired lateral protective fire being relatively little, a permanent work which is essentially a traditore battery and nothing else can be both small and well-covered against frontal fire at an expense much less than that of a great self-contained fort. Its own local protection may be either a ditch with sunk defences or an in- fantry system surrounded by wire, but these auxiliaries, too, would be withdrawn from the crest facing the enemy to positions on the reverse slope. The only case in which it would be necessary for any part of the system to go forward to the crest and front slope would be that in which the artillery observation and com- mand post is combined with the traditore in one work or one enclosed group. In such a case the post in question would un- doubtedly require special treatment as regards its own close- defence. But all that in principle is necessary is that the post and its liaisons should be immune.
On the other hand, the security of the main armament against a rush of hostile infantry was far greater when an obstacle defended by fire completely surrounded it, and military engineers were very loth to impair this security. No doubt, when the obstacle cover- ing the front of the batteries in the deployed order was fully organized, the latter might be considered safe enough for practical purposes so long as the interval-defence remained effectively in action to protect it. But a danger period was foreseen in which the obstacle was not yet fit to perform its function with cer- tainty. The " brusque " or (more accurately) the " abbreviated
1 The term " storm-proof," frequently applied to such infantry works, hardly seems to connote their real function.
attack," proposed by the Bavarian General von Sauer, had many supporters; and as the tendency already mentioned, of modern warfare between " armed nations " is to push the line of resistance as nearly up to the frontier as possible, the fortifica- tions of that line were in fact exposed to instant attack. 2 Those of Verdun and Toul were little more than 20 m., the easternmost fort of Liege only 13 m., from the German frontier, while the western Metz forts could be bombarded from French soil. In former days, this would have mattered less, but the growing mobility of heavy artillery from about 1890 for the first time made it possible to employ true siege artillery within a few hours of the opening of hostilities. The attacker, on the other hand, naturally had to forego some of the powerof hisattacking means in attempting a coup. His truly mobile siege artillery was limited, or supposed to be limited, to the calibre of 21 cm. Heavier pieces though they no longer took weeks or months to arrive in their emplacements, at any rate took days to do so, and by a sort of general agreement (to which however there were exceptions) the situation was met by placing a part of the main armament of the defence called the safety armament inside a closed ob- stacle. Usually it was an existing fort that was adapted to house the safety armament, but sometimes it was included in the design of a new work. The fort thus in practice reverted partially to its old duty of serving as a battery position, while in theory its func- tion had become entirely that of locally protecting a tradilore or other interval defence. The distinction between property and accident was no doubt clear to specialists, but the result was that the generality of armies and peoples continued to look upon a fort as their fathers had looked upon it, till the astonishing events of Aug., Sept. and Oct. 1914 so thoroughly undeceived them too thoroughly, indeed, for in the revulsion, not merely safety-armament guns but even interval-flanking guns were removed from closed works.
In the system of group-fortification, it was naturally much easier to house a safety armament. No element within the ring of wire need cramp any other, or be drawn into the fighting activity of another, or suffer from the shells intended for another. 3 Full fighting protection will be necessary, as is always the case with safety armaments, but, as has been noted above, with more room the same safety can be given with less expense.
In sum, therefore, the necessity of compromise on this question of safety armament has caused the dispersed-elements and the concentrated-elements schools to agree upon: (a) the group or Feste principle for interval-flanking elements, obstacles and de- fence of the same, and safety portion of main armament; (b) the order principle of deployed artillery, with an obstacle covered by flanking fire, for the remainder of the main armament. This, it will be noted, leaves a real liberty for the treatment of particu- lar cases. The proportion of total armament installed as "safety " is whatever the designer chooses to make it in each instance, the Feste being adaptable to any proportioning within reasonable limits fixed by the contour of the ground. A practical check on enclosing an unnecessarily high proportion will always be the expense of giving full fighting protection.
Examples of Croup- Fortification. Types of forts, both main armament forts and others, being described and illustrated in 10.696,
2 To wire a perimeter or frontage of 30 km. to a depth of eight yd. only requires three eight-hour shifts of (in round numbers) 6,000 workers each, as well as mechanical, animal or human transport for about 4,000,000 yd. of barbed wire, weighing 300 tons or so, and 100,000-130,000 stout posts. Other work to be done includes the clearance of the field of fire, the digging of trenches, the con- struction of shelters (if not in existence already), opening of com- munications and liaisons, etc. Land which is occupied by a fort- ress garrison in war rarely belongs to the Government in peace.
'This can be demonstrated by the "theory of probabilities." Assume a main-armament cupola 16 ft. in diameter, under accurate attack by a gun having a probable error of 60 ft. in range and 3 ft. in line. Calculation shows that this will probably be hit by 7 % of the shots fired. Now assume a traditore element having a vulnerable surface on top of 20 ft. from front to rear and 25 ft. laterally. Placed with its front edge 120 ft. behind the centre of the cupola, this will receive 3-62% of the shots aimed at the latter. Placed with the front edge 240 ft. behind, it will be hit by 0-2 % of the shots. In other words, at twice the distance it is eighteen times as safe.
