,\. v •/■> ,^ oV y '% ■\ A e>- "■ " AV % * CY 1 .* - A ft//?? * S . ^\ ~- ~ ;'!^k '^ ^ .V ^ v tf * AV O V \ V O >, a OA" / O >, a OA" 'III \V s , o.V •S » -^ *>' A BATTLEFIELDS OF THE WORLD WAR H ^ \ AMERICAN GEOGRAPHICAL SOCIETY RESEARCH SERIES NO. 3 W. L. G. Joerg, Editor BATTLEFIELDS OF THE WORLD WAR Western and Southern Fronts A Study in Military Geography DOUGLAS WILSON JOHNSON Professor of Physiography in Columbia University Formerly Major, Division of Military Intelligence, U. S. A. WITH A FOREWORD BY GENERAL TASKER H. BLISS Member of the Inter-Allied Supreme War Council NEW YORK OXFORD UNIVERSITY PRESS AMERICAN BRANCH: 35 West 32ND Street LONDON, TORONTO, MELBOURNE, AND BOMBAY I 92 I .3* COPYRIGHT, I 92 I BY THE AMERICAN GEOGRAPHICAL SOCIETY OF NEW YORK COPYRIGHT IN GREAT BRITAIN MAI d.\} \oC\ THE CONDE NAST PRESS GREENWICH, CONN. ©CI.A614473 p CONTENTS CHAPTER PAGE List of Illustrations v Foreword, by General Tasker H. Bliss .... xiii Introduction xv I The Battlefield of Flanders: The Wet Clay Plain Battlefield i II Military Operations on the Battlefield of Flanders 53 III The Battlefield of the Somme : The Dry Chalk Plain Battlefield 84 IV Military Operations on the Battlefield of the Somme 134 V The Battlefield of the Marne: The Plateau- AND-LOWLAND BATTLEFIELD 215 VI Military Operations on the Battlefield of the Marne 269 VII The Battlefield of Verdun: The Cuesta-and- Lowland Battlefield 316 VIII Military Operations on the Battlefield of Verdun 377 IX The Battlefield of Lorraine : The Cuesta-and- Mountain Battlefield 415 X Military Operations on the Battlefield of Lorraine : . . . . 473 BATTLEFIELDS OF THE WORLD WAR XI The Battlefield of the Trentino : The Alpine- AND-PlEDMONT BATTLEFIELD 488 XII Military Operations on the Battlefield of the Trentino 523 XIII The Battlefield of the Isonzo: The Karst Plateau Battlefield 541 XIV Military Operations on the Battlefield of the Isonzo 563 XV The Battlefield of the Balkans: The Range- and-Basin Battlefield 572 XVI Military Operations on the Battlefield of the Balkans 605 Index 635 LIST OF ILLUSTRATIONS (m=map; bd =block diagram; d=diagram; p = photograph) » Battlefield of Flanders Fig. Page 1 Index to the battlefield maps (m) 2 2 Generalized sketch of Battlefield of Flanders (m) ... 3 3 Mont Kemmel (p) 4 4 Mont des Cats (p) 5 5 Geological column of formations of Flanders battlefield (d) 10 6 Cross-section of Battlefield of Flanders (d) 12 7 Topographic belts of Flanders battlefield (m) 13 8 Gohelle portion of southern Flanders plain (p) . . . . 16 9 Marshes of the Scarpe (p) 19 10 Canal used as defense line (p) 29 1 1 Scattered population of Flanders plain (m) 30 12 Compact villages in Somme region (m) 31 13 Flanders plain southeast from Mont des Cats (p) . . . 34 14 Flanders and Somme battlefields (bd) 39 15 Drainage ditches and canals in maritime belt of Flanders battlefield (m) 43 16 Passer elle across flooded area along the Yser (p) .... 45 17 Sandbag defenses commanding flooded valley of the Yser (p) 45 18 Morass produced by artillery fire on tree-covered part of water-soaked plain of Flanders (p) 46 19 Defensive position along canal draining Flanders plain (p) 49 20 Barbed-wire defenses along coastal dunes (p) 50 21 Extension of defense lines across sandy beach into sea (p) 51 22 Yser barrier, showing three main lines of resistance (m) 56 23 Flanders plain south from Mont des Cats (p) 76 Cf. also PL I (see p. xi) Battlefield of the Somme 24 Generalized sketch of Battlefield of the Somme (m) . . 85 25 Chalk area of northern France (m) 86 26 Parallel valleys of the Somme and neighboring rivers (m) 88 vi BATTLEFIELDS OF THE WORLD WAR Fig. Page 27 Location of Somme battlefield at natural gateway con- necting Flanders with the Paris region (m) 92 28 Influence of anticlines and synclines on Somme battle- field (m) 93 29 Somme plain looking west across La Boisselle ravine (p) 94 30 Great Plains near Bismarck, North Dakota (p) . . . . 95 31 German camouflage of road across Somme battlefield (p) 97 32 Fault cliff forming northeast face of Vimy Ridge (d) . . 101 33 View down backslope of Vimy Ridge (p) 102 34 Looking up backslope of Vimy Ridge (p) 103 35 Flanders plain northeast from Vimy Ridge (p) .... 104 36 Mont Kemmel-Vimy Ridge region (bd) 106-107 37 Typical hill and valley cross-section in Somme battlefield (d) '. 108 38 Geology of portion of Somme battlefield (m) 109 39 Geological cross-section of typical hill of Somme battle- field showing possible locations of trenches, dugouts, etc. (d) 118 40 Marshy valley of Somme River near Amiens (p) ... 122 41 Typical portion of marshy valley of the Somme (m) . . 123 42 Quadrilateral of Caesar's Camp (m) 126 43 Typical dry valley in chalk of Somme battlefield (p) . . 129 44 Oise Canal at Chauny (p) 131 45 Three sectors of operations in 191 6 Battle of the Somme (m) 146 46 Field of operations of First Battle of the Somme (m) . 147 47 Main wire defenses of Hindenburg Line (p) 167 48 Mont Renaud, commanding Oise valley route to Paris (p) 174 49 Fortified summit of Mont Renaud (p) 175 50 Battlefield of Villers-Bretonneux (m) 181 51 Open valley of Somme River northeast of Villers- Bretonneux (p) 182 52 Some of the major topographic elements utilized in con- structing defenses of Arras bastion (m) 190 53 Somme valley at Brie crossing above Peronne (p) . . . 196 54 Concrete machine-gun shelter on Hindenburg Line (p) . 201 55 Main German defensive positions in northern France (m) 208-209 Cf. also PL II (see p. xi) LIST OF ILLUSTRATIONS vii Battlefield of the Marne Fig. Page 56 Generalized sketch of Battlefield of the Marne (m) . . 216 57 North-south cross-section of Marne plateau (d) . . . . 217 58 East-west cross-section of Marne plateau (d) 217 59 View northeast across the Champagne showing vineyard- covered lower slopes of plateau escarpment in fore- ground (p) .' 218 60 Infertile chalk areas of the Champagne with vineyard- covered slope near base of Marne plateau escarpment in foreground (p) 219 61 The western theater of war (bd) 220 62 Marne plateau and Champagne lowland (bd) . . facing 222 63 Contrast between east-west topographic belts of central plateau and north-south concentric belts of Paris Basin (m) 226 64 Chemin des Dames region (bd) facing 234 65 Typical cliff slope in northern portion of Marne plateau (d) 232 66 Parallelism of topographic elements in Marne plateau (m) 239 67 Flooded floor of Petit Morin valley (p) 241 68 Flat-floored valley of Marne River in vicinity of Meaux (p) 242 69 Military value of two types of valley profiles (d) ... 243 70 Valley of the Marne near Dormans (p) 245 71 Destroyed bridge over the Marne at Chateau-Thierry (p) 246 72 Inundated valley of the Vesle near Fismes (p) 247 73 Escarpment of Marne plateau south of Epernay (p) . . 253 74 Looking northeast over Basin of Rheims from tree-top observatory (p) 256 75 Moronvilliers massif as seen from tree-top observatory (p) 257 76 Supposed former courses of the Soude, Somme, Vaure, and Maurienne (m) 265 77 St. Gond marshes (p) : 267 Cf. also Pis. Ill and IX, A {see p. xi) Battlefield of Verdun 78 East -west section across Battlefield of Verdun (d) . . . 318 79 Generalized sketch of Battlefield of Verdun (m) .... 319 viii BATTLEFIELDS OF THE WORLD WAR Fig. Page 80 Woevre lowland as seen from crest of Meuse plateau scarp (p) 322 81 Woevre lowland looking northeast from crest of Meuse plateau scarp at Gironville (p) 323 82 View northeast from Mont d'Amance across Seille low- land (p) 324 8^ Meuse and Moselle fortified defensive lines (m) .... 329 84 Northern Argonne and Meuse plateaus (bd) . . . 338-339 85 Butte of Vauquois (p) 349 86 Meuse plateau west of Verdun looking northwest from near Fort Marre (p) 350 87 Looking north from spur bearing Fort Marre, showing even sky line of dissected Meuse plateau (p) .... 351 88 Irregularly eroded scarp of Meuse plateau facing Woevre lowland (p) 352 89 East-facing escarpment of Meuse plateau and Woevre lowland (p) 354 90 Crest of Meuse plateau with ruins of Combres (p) . . . 357 91 Portions of the Barrois, Meuse, Moselle, and Saffais plateaus (bd) 358 92 Abandoned valley of Moselle River in Meuse plateau west of Toul (p) •. 360 93 Mont Sec, an outlying erosion remnant of Meuse plateau (p) ' 362 94 Verdun region (bd) 364 95 Cross ridges and connecting "bridges" upon which the eastern defenses of Verdun were based (m) 366 96 Meuse plateau east of Verdun (p) 368 97 Flooded valley of Meuse River in Meuse plateau (p) . 371 Cf. also Pis. IV and JX, B {see p. xi) Battlefield of Lorraine 98 Generalized sketch of Battlefield of Lorraine (m) ... 417 99 Northwest-southeast section across Battlefield of Lorraine (d) 418-419 100 Forest of Champenoux in lowland at base of Mont d'Amance plateau (p) 420 101 Saffais plateau escarpment and lowland east of it (p) . 422 LIST OF ILLUSTRATIONS ix Fig. Page 102 Looking southeast over lowland from crest of Saffais plateau (p) . 423 103 Valley of the upper Moselle north of the Remiremont (p) 426 104 Upper valley of Moselle River as seen from base of Ballon d' Alsace (p) 427 105 Vosges Mountains looking north from Ballon d'Alsace (p) 428 106 Nancy region (bd) 432-433 107 Valley of the Natagne traversing Mont d'Amance plateau bastion (p) 434 108 Ste. Genevieve plateau forming northern face of Mont d'Amance natural bastion (p) 435 109 Butte of Mousson (p) 436 no Eastern portion of Mont d'Amance natural bastion (p) 438 in Diagram to illustrate protective value of forest cover on steep upper slope of Moselle plateau scarp (d) ... 441 112 Marginal tree fringe of Meuse plateau scarp, also charac- teristic of Moselle and Saffais plateaus (p) 442 113 Cote deDelme seen from crest of Moselle plateau scarp (p) 444 114 Looking southward from "Amance bastion" of the Grand Couronne across "Nancy curtain" to "Luneville bastion" (p) 446 115 Lowland east of Saffais plateau (p) 450 1 16 Looking eastward from crest of Vosges down valley cut in steep eastern face of range (p) 458 117 Valley of the Meurthe showing edge of scarp of "sandstone Vosges" (p) 462 118 Natural bastions and curtains of Nancy region (m) . . 483 Cf. also Pis. V and VI {see p. xi) Battlefield of the Trentino 119 Italian theater of war (m) 490 120 Generalized sketch of Battlefield of the Trentino (m) . 493 121 New Italian military road ascending Monte Grappa massif (p) 494 122 Trench of the Adige with high ridge of Con! Zugna in distance (p) 495 123 New Italian military road zigzagging up slope of Monte Baldo (p) 496 x BATTLEFIELDS OF THE WORLD WAR Fig. Page 124 Teleferica on slope of Adamello group (p) 497 125 Comparison of Italy and California (m) ....... 498 126 Crossing glaciers of high Adamello group by dog sleds (p) 499 127 Special costumes worn by "Alpini" in mountain warfare on snow-covered terrain (p) 500 128 Monte Pasubio massif as seen from airplane (p) . . . . 501 129 Coastal marshes bordering Adriatic at foot of Piedmont plain (p) 503 130 Multiple channels and sand bars of Piave River (p) . . 505 131 Italian machine-gun positions along dike bordering Piave River (p) 505 132 Piave River crossing Piedmont plain as seen from Monte Grappa (p) 508 133 Italian first-line trenches behind defensive barrier of the Piave (p) 511 134 Inn corridor in vicinity of Innsbruck (p) 512 135 Flat-floored trench of the Vintschgau (p) 513 136 Wild mountain gorge forming Finstermunz Pass (p) . . 515 137 Konigspitze, part of the glacier-clad Ortler group (p) . . 517 138 Upper bench of Asiago plateau (p) 518 139 Gun position at exit of tunnel excavated in solid rock of an Alpine mountain (p) 521 Cf. also Pis. VII, X, B, and XI, B (see p. xi) Battlefield of the Isonzo 140 Generalized sketch of Battlefield of the Isonzo (m) . . 543 141 Barren upland of the Carso (p) 545 142 Isonzo front showing the Carso and Bainsizza plateaus (m) 547 143 Ridges of the Alps in northern part of Isonzo battlefield, and Isonzo River at Caporetto (p) ........ 548 144 Matajur-Cucco ridge viewed from the plain (p) . . . . 550 145 Italian front across karst plateau (p) 552 146 Sink hole in limestone of karst country (p) 553 147 Defensive position on western margin of the Carso (p) . 555 148 Hermada hills on southern border of Carso (p) .... 556 149 Isonzo River flowing through Plezzo basin (p) .... 558 150 Isonzo River at base of Carso plateau (p) 561 Cf. also Pis. VIII and X, A (see p. xi) LIST OF ILLUSTRATIONS xi Battlefield of the Balkans Fig. Page 151 Barriers of the Balkan Peninsula (m) 574 152 Generalized sketch of Battlefield of the Balkans (m) . . 577 153 Rugged limestone mountains of southern Albania (p) , 578 154 Flat floor of Struma basin (p) 580 155 Argyrocastro basin (p) 582 156 Wagon train passing through flooded Morava valley (p) 586 157 Austrian forces assembled along Save River barrier (p) . 587 158 Flat desert floor of Monastir basin (p) 594 159 Natural defenses of the armed camp of Saloniki (m) . . 598 160 Northwestern Serbia (m) 607 161 Italian military road ascending to pass giving access to Chimara basin (p) 621 162 Exposed floor of broad valley of Voyussa River (p) . . 622 163 Voyussa River north of Valona (p) 625 164 Mountainous terrain in great bend of the Cherna (p) . . 626 165 Moglenitsa basin and mountainous triangle between the Cherna and Vardar Rivers (m) 629 Cf. also PL XI, A (see below) PLATES IN SEPARATE POCKET I Topographic map of the Battlefield of Flanders, 1 :300,ooo II Topographic map of the Battlefield of the Somme, 1:300,000 III Topographic map of the Battlefield of the Marne, 1:300,000 IV Topographic map of the Battlefield of Verdun, 1 :300,ooo S V Topographic map of the Battlefield of Lorraine, 1 :300,ooo VI Block diagram of the Battlefield of Lorraine VII Block diagram of the Battlefield of the Trentino * VIII Block diagram of the Battlefield of the Isonzo TY J A Panorama of the Meuse plateau and Woevre lowland ' \ B Panorama of the Douaumont sector of the Verdun • Battlefield A Panorama of the Isonzo-Carso region ^ B Panorama of the Asiago plateau \ A Panorama of the Belashitsa Range and Struma basin B Panorama of the Adige trench and bordering mountains X • Figs. 13 and 23 are reproduced with the permission of the Director of Military Intelligence of the War Office, London; Figs. 10, 16, 17, 18, iq, 20, 21, of the Direction des Informations Militaires, Ministere de xii BATTLEFIELDS OF THE WORLD WAR la Defense Nationale, Brussels; Figs. 31, 44, 48, 49, 70, 71, 72, 85, 90, °3> 97j I0 9> IT 3 an d Pl s - IX,A and IX,B, of the Chef de la Section Photographique de I'Armee, Paris; Figs. 127, 128, 120, 130, 131, 132, 133, 138, 139, 141, 143, 144, 145, 146, 147, 148, 149, 150, 154, 155, 138, 164 and Pis. X,A, X,B, XI, A, XI, B, of the Direzione del Servizio Fotografico, Comando Supremo, Regio Esercito Italiano, Rome. Pis. I to V inclusive are reproduced from the five-color Carte de France, 1:200.000, published by the Service Geographique de I'Armee, Paris. Each plate is made up of the parts of the sheets of this map necessary to cover a given battlefield. The sheets used are: 2 {Dunkerque-Bruges) , 4 {Lille), Q {Amiens), 10 {Mezieres), 11 {Longwy), 16 {Paris), 17 {Chalons), 18 {Metz), 19 {Karlsruhe), 25 {Melun), 26 {Troyes), 27 {Nancy), 28 {Strasbourg), 35 {Vesoul), 36 {Zurich). FOREWORD Every informed General Staff officer understands, of course, the broad relations between topography and the strategic plans for a campaign or the tactical dispositions for a battle. These were known and applied as well by Alexander, Hannibal, Caesar as by any modern commander. Many of these relations are so simple and self-evident that, in the rudest ages of war, they have been instinctively noted and taken advantage of; and, where not, the resulting disaster may generally be traced to self-confi- dence or indifference to the plain teachings of Nature. Many a commander has unconsciously taken into his calculations cer- tain basic conditions of geology, just as Moliere's grocer spoke prose without knowing it. But, in war, as between two equally skillful generals, at some all-important time in the game one of them has to play to the other's lead. It may be that then only one of them is favored by all the physical and seasonal conditions of the terrain. The other, most likely, can not defer his action to a more convenient and favorable time or place. His adversary's card is on the table; he must play his own then and there or forfeit the game. All Nature seems to be in alliance with his opponent. It is just then that his success — which may consist in preventing defeat quite as much as in gaining a patent victory — may depend upon a deeper knowledge of Nature and her processes. If the sun, the very stars in their courses, the tides of the sea, the running streams — everything on the surface of the earth and above it has marshaled its influence in favor of his enemy, then, though he may have to surrender topography to that enemy, he himself may have to burrow beneath it all and exert his strength amid the unseen, tumbled, and contorted ruins of old Chaos. Not only there, but on the surface too, he must know how to exert his propaganda on that Nature which seems to have entered into an unholy alliance xiv BATTLEFIELDS OF THE WORLD WAR against him, to play one of her laws or processes against another, to slay her ally with venom-tipped arrows skillfully stolen from her own quiver. It is difficult to say whether this work of Professor Johnson has been written more for the benefit of the geographer and geologist, or of the military student, especially the student of the operations in the Great War, or of those who like to read charming descrip- tions in sweetly flowing English, of the physical landscapes in both their gay and gloomy moods — the plains of Flanders, the chalk uplands of Picardy and Artois, the pastures and vineyards of Champagne, the glens and forests cf the Vosges, or the moun- tain valleys and tarns and peaks and precipices of the Alps and the Balkan ranges — which are so clearly portrayed that neither the scientist nor the military student nor the lover of Nature in literature need visit them in order to see and understand them. To all of them it will be a classic, and to none more than to another. But it is to the military student that I especially com- mend it in order that he may see, among other things, how the art of war has gathered to itself and absorbed not only all trades and arts, all the elder sciences, but has now laid hands on this youngest one of all, this youthful David of human knowledge, to help us to do the one thing that can ennoble our own art and science — to slay the evil giants of wrong and oppression. Tasker H. Bliss INTRODUCTION The Problem Stated "Do the mountains defend the army, or does the army defend the mountains?" The problem is an old one, and has claimed the attention of military authorities in all countries and in all times. Expressed in broader terms, it is the oft-debated question as to the relative influence of topography upon strategy and tactics under changing conditions of warfare. It is an ever-recurring question, for each "revolution" in methods of combat brings in its train a body of opinion intent on demonstrating that, under the new conditions of fighting, topographic obstacles have lost their significance, strategic gateways no longer exist, and commanding positions no longer "command." Then, as opposing forces share in the new discoveries, or profit in equal measure by new systems, the fundamental importance of topography reasserts itself, and each side maneuvers for an advantageous position on the terrain as one of the prerequisites to victory in battle. The question is still a live one. The warfare of today employs a variety of inventions and technical devices, each of which may appear to reduce, if not to destroy, the influence of topography upon military operations. What protection is a river channel, when the modern military engineer can throw bridges across it in a few hours, defended by artillery which can reach the enemy many miles beyond the farther bank? What need has the artil- lery for hill positions, when guns are now commonly concealed in valleys and ravines, firing with marvelous accuracy upon objec- tives the gunners never see? With sound-ranging and flash- ranging devices to spot enemy batteries, with airplanes and aerial photography to locate these and other objectives and to exercise surveillance over enemy movements, of what significance is a paltry elevation of some few tens or hundreds of feet, dignified in earlier wars as "dominating heights?" So might one multiply, indefinitely, queries the common answer to which would seem to xvi BATTLEFIELDS OF THE WORLD WAR be that in the warfare of the present time the inventions of Man have reduced to insignificance the role of Nature. Strategic Frontiers If this be true, the matter is one of far-reaching importance, not only to the military leader but also to the statesman. The problem of strategic frontiers and the question of the influence of topography on warfare are inevitably linked the one with the other. If the character of the terrain is of no consequence in modern fighting, much of the argument for strategic frontiers falls to the ground. On the other hand, if a frontier may be strong or weak, impregnable or vulnerable, according to the nature of the terrain which it traverses, then the statesman must weigh this aspect of the territorial problems which come before him for adjustment. Whether he believes that strategic claims are but cloaks covering a multitude of imperialistic sins, or holds the view that strong frontiers may be locks discouraging inter- national burglary and hence aids to maintaining world peace, he must take cognizance of strategic frontiers if they really exist. Others will demand them, and the wise statesman will bring to the consideration of such demands a full knowledge of the strength and weakness of the supporting arguments. The Value of Terrain One object of the present volume is to demonstrate the fallacy of the contention that modern methods of warfare have reduced to insignificance the r61e of the terrain as a factor in strategy and tactics. This demonstration might be made by abstract argu- ment: by showing that, despite the enormous improvement in the artillery and other arms of the service, it is still the infantry which must drive back the enemy and conquer the ground on which he stands, and that whatever affects the movement of infantry remains a vital element in the fighting; by pointing out that one of the most effective agents in breaking through wire- defended trench systems, the tank, finds in rivers and marshes more serious obstacles than such topographic barriers ever consti- INTRODUCTION xvii tuted for lighter and more mobile weapons; by proving that direct and uninterrupted observation of enemy back areas from concealed positions on topographic eminences held constantly in easy communication with the artillery is, for the speedy and accurate control of gun fire, greatly superior to observation from airplanes, because such observation is often interrupted by weather conditions, is limited in some measure as to its efficiency by difficulties of communication, and is reduced in value by the necessity of flying high to avoid enemy bombardments or by the exigencies of aerial combats with enemy planes. Or the demonstration might rest more heavily upon the expert testimony of those responsible for the gigantic military operations of the World War: upon the opinion of De Castelnau, proved by events to be correct, that the peculiar terrain of the Nancy region would enable limited forces to defeat a numerically superior enemy in a modern battle; 1 upon Haig's dictum that the whole war might be considered a series of struggles for topographic position; 2 upon Ludendorff's statement that by 1916, after the Battles of Verdun and the Somme, "the decisive value of the artillery observation and the consequent necessity of paying great attention to the selection of positions had also become apparent ;" upon his testimony, a year later, that "another tactical detail which was emphasized everywhere was the value of ground observation for artillery. Only by that means could the attacking hostile infantry be annihilated, particularly after penetrating our front, or fire be concentrated on decisive points of the battle- field;" and, finally, upon his assertion in 191 8, despite the fact that loss of strong defensive positions had by that time weakened his faith in the value of commanding ground in defense, that "in the attack in the war of movement the capture of some high ground brought about the tactical decision. Its possession must therefore be striven for as a matter of principle." 3 1 See Chapter X. 2 Personal communication. 3 Erich von Ludendorff: Ludendorff's Own Story, August 1914-November, 191 8: The Great War from the Siege of Liege to the Signing of the Armistice As Viewed from the Grand Headquarters of the German Army, 2 vols., New York, 1919; references in Vol. 1, p. 324; Vol. 2, pp. 103, 200. xviii BATTLEFIELDS OF THE WORLD WAR The writer has preferred, however, to let the demonstration rest upon a simple presentation of the record. No one who knows the terrain of the principal battlefields of the World War, and who follows carefully the operations of the contending armies, can doubt that under modern conditions of warfare the terrain is, more than ever before in military history, an important factor in strategy and tactics. This is not to imply that armies should put their whole trust in topographic barriers or commit the error of seeking victory by standing permanently on the defensive upon some strong natural position. The folly of such tactics was repeatedly demonstrated by the Germanic generals of Napoleon's time, and this extreme view of the value of "commanding ground" has been justly condemned by Clausewitz and other writers on the art of war. Nature offers no absolutely impregnable positions. In the words of Foch, "All terrains may be crossed by the enemy unless defended by rifle shots, that is to say, by active and valiant men." 4 It is rather to support the view of Napoleon that strong natural positions are one of the factors which enable an army to maneuver advantageously in the face of a superior or equal enemy; the view of Clausewitz that even the larger army must not despise a source of strength which enables it to hold the enemy at bay with limited forces along one part of its front while it concentrates its major effort elsewhere; and the view of the modern general who finds in the better observation and artillery control offered by a favorable terrain elements which often prove decisive in both defensive and offensive battles. To the fullest appreciation of the important influence of topography on modern warfare two things are fundamentally essential, two things which unfortunately have not been easily obtainable. The first is a reasonably accurate mental image of each battlefield, a picture of those salient features of the terrain which might be expected to influence the dispositions and move- ments of armies. The second is a record of the military operations on each battlefield presented, not in the usual terms of advances or retreats to this or that line of villages and towns, but in terms 4 Ferdinand Foch: Des principes de la guerre, Paris, 1917. p. 29. INTRODUCTION xix of the plateau scarps, ridge crests, valley trenches, marsh barriers, and other topographic features which obviously affected the fighting and its results. Scope of the Present Work In the present volume the writer has endeavored to supply these two prerequisites to an understanding of the influence of topography upon the World War. For each important battle- field on the western and southern fronts there is presented, with the aid of maps, perspective diagrams, and photographs taken for the special purpose of showing essential elements of the terrain, such a picture of the stage upon which the drama of war was enacted as is hoped will prove both entertaining and helpful not only to the specialist, but equally to the reader who is neither military expert nor geographer. Throughout the descriptions an effort has been made to unite scientific accuracy of treatment with a phraseology as far as possible devoid of technicalities. If these descriptive chapters prove of service to the future historian of the war who would understand fully the events he chronicles, if they aid the military student in analyzing the operations of each battle and campaign, if they "provide the geographer with new illustrations of the influence of topography on one phase of human activities, if they furnish the intelligent traveler with a new form of guide to European battlefields, and if they give to other readers a few hours of pleasure and of profit, the author's labors will be most abundantly rewarded. For each battlefield described there is a companion chapter recording in proper sequence the major military operations which were carried out upon it. Here for the first time, so far as the author is aware, the reader will find the operations of each battlefield presented as a unit and in terms of the terrain by which the plans of campaign and the movements of armies were most profoundly affected. In these chapters the author has not attempted to write history but to offer a partial interpretation of the significance of certain events in history. Time has not permitted that investigation of original sources essential to an xx BATTLEFIELDS OF THE WORLD WAR accurate and well-balanced chronicle of the events of the war, and the reader will doubtless discover errors both of omission and commission which would mar the value of the record, did it make any pretense at giving an adequate history of even the major events of the great conflict. It is believed, however, that no such errors will be found to diminish the usefulness of these chapters as a general resume of the principal battles and campaigns on each sector of the western and southern fronts, as affected by the terrain upon which they occurred. No treatment of the Rumanian and Russian fronts is included in the present text. The operations on those fronts, in so far as they materially influenced the main struggle, came to an end about the time the author published his "Topography and Strategy in the War." The reader who would complete his survey of the European battlefields with some account of the eastern terrains and the principal operations upon them, may consult that volume. Acknowledgments are due Messrs. Henry Holt and Company, the publishers of "Topography and Strategy in the War," for permission to incorporate in the present work the descriptive matter relating to the northern and central Balkans, as well as certain paragraphs concerning the Italian front and several text figures which appeared in the earlier work. Basis of the Present Work From the beginning of the war the author followed closely the military operations on large-scale topographic maps, aided by such special knowledge of the terrain as his previous travels and geographic investigations had given him. The results of these studies appeared from time to time in the publications of the American Geographical Society, and later in the book mentioned above. The cordial reception given these very general essays encouraged the belief, particularly after America entered the struggle, that a more critical and detailed study of the principal battlefields of the World War would give results of value, an idea warmly supported by the American Geographical Society. An exceptional opportunity was presented when the author, INTRODUCTION xxi having previously been ordered by the Secretary of War to report to the Secretary of State for special service abroad, was directed by the latter to proceed to Europe "for the purpose of making special studies in military geography for the use of this Depart- ment [of State] in connection with the work being done at the direction of the President by Colonel E. M. House." Prepara- tions were already being made for the Peace Conference which must follow the conclusion of hostilities, and it was considered essential that the American representatives should bring to the discussion of the strategic claims of different countries as full a knowledge as possible of the strategic and tactical value of land- forms under modern conditions of warfare. A part of this knowledge could best be gained by studies on the ground while operations were still in progress, and while the men directing them had all phases of their problems freshly in mind. The American Geographical Society was then housing the staff of the "Inquiry" assembled under Colonel House's direction, and co- operating in its investigations; and it made from its own funds a generous appropriation to assure the author every facility in prosecuting his European studies. It was under these favorable conditions that the author, accom- panied by one of his former students, Lieutenant S. H. Knight, professor of geology in the University of Wyoming, visited the Belgian, British, French, American, Italian, and Balkan fronts. The War Offices at London, Paris, and Rome generously placed at our command large-scale maps and relief models of the battlefields, reports on operations, and other material of value in the prosecution of the work; and also facilitated in every way our studies along the various fronts. As a rule the generals com- manding on these fronts not only welcomed us at their messes, where opportunity was afforded to discuss problems of military geography with the most competent leaders in each given sector, but manifested a personal interest in the investigation, often tracing in detail upon maps or models the course of operations, and contributing from their wide experience suggestions of the highest value. A member of the staff familiar with the terrain, xxii BATTLEFIELDS OF THE WORLD WAR its defensive organization and the operations upon it, was usually assigned to accompany us and to assist in our studies. Upon the conclusion of hostilities the author was called upon to undertake geographical investigations in preparation for the Peace Conference which were only indirectly related to the study of European battlefields. But at the Conference he had an oppor- tunity, first as Chief of the Division of Boundary Geography on the American delegation and as technical adviser meeting with various commissions, and later as a member of several of the international Territorial Commissions, to participate in extended discussions of the strategic and tactical value of the terrain along certain proposed frontiers. If the author has thus enjoyed some unusual advantages in the prosecution of the studies upon which the present volume is based, it is but fair to add that other circumstances limited the extent to which these advantages could be utilized. The enemy was still in possession of large areas which are of necessity in- cluded in any adequate survey of the principal battlefields, and it has been possible to visit but a limited portion of his former holdings since the armistice. The time available for the battle- field studies was seriously curtailed by the necessity of carrying on other investigations for the "Inquiry," the results of which were needed in preparation for the Peace Conference. Before a begin- ning on this volume could be made, the assignment to service at the Peace Conference postponed the work for many months, a postponement which was prolonged by further service as an adviser on geographical questions to the Department of State after the author's return from Paris. Not until February, 1920, could the considerable task of preparing this volume be taken up, and then only in connection with regular university duties and under special conditions which made absolutely necessary the completion of the manuscript by early autumn. Thus have circumstances beyond the author's control limited unduly the time at his disposal, while lack of access to a large library during half the working period proved a further handicap. If the reader feels that inadequate consideration is given to certain works on INTRODUCTION xxiii the great war which have recently appeared, and discovers other deficiencies, of which the author is only too conscious, it is hoped he will agree that the omissions do not materially affect the main purpose of the volume, and that he will show such indulgence as the circumstances cited may seem to deserve. Acknowledgments It is evident that the conditions under which the author visited the battle fronts and prosecuted his studies must place him in an embarrassing position in acknowledging his indebtedness to the many who have assisted in this work. The number of those who rendered valuable aid is so large that merely to name them would be impracticable. It must suffice to express here my pro- found gratitude to all of them, and to record my sense of obliga- tion to each and every one who in government bureaus or on the battlefields placed his materials, his expert knowledge, and his invaluable assistance at my service. A special measure of appre- ciation is due to Marshal Joffre, General de Castelnau, General Gouraud, and General Bourgeois of the French Army; Field Marshal Sir Douglas Haig, General Rawlinson, and General Biddulph of the British Army ; General Diaz, General Badoglio, and General Ferrero of the Italian Army; and General Henrys and General Milne of the French and British forces in the Army of the Orient, for critical analyses of certain military operations on different terrains, for valuable suggestions and criticisms of the problem under investigation, for providing unusual facilities for prosecuting the work, or for personal discussions and explana- tions in the field. To all of these the author desires to acknowl- edge his sense of obligation, without in the slightest degree mak- ing them responsible for any statements of fact or any inter- pretations which appear in the following pages. To Colonel Delfino De Ambrosis, military geographer on the Italian General Staff, for invaluable assistance and advice at the War Office at Rome, at the Military Geographical Institute at Florence, and along the Italian front; to the Italian geographer, Colonel Filippo De Filippi, and his gallant brother, Captain De Filippi of the xxiv BATTLEFIELDS OF THE WORLD WAR Royal Italian Navy, who gave his life to save others when his destroyer was sunk in the Adriatic, for manifold services at London, Paris, Rome, and in Albania; to Lieutenant Colonel T. Edgeworth David, Chief Geologist of the British Expedi- tionary Force, for important data on the geological conditions encountered on the Somme front and in Flanders; to Captain Alan G. Ogilvie, geographer with the British forces in the Balkans, for valuable criticisms of the text describing the Balkan front; and to my assistant and companion in the field, Lieutenant S. H.. Knight, for efficient aid at all times, my cordial thanks are gladly given. It is a special pleasure to express my gratitude to my former chiefs, General Marlborough Churchill and Colonel R. H. Van Deman, for their unfailing and generous assistance through- out my service under them, given both directly and through their representatives at London, Paris, and Rome. In the descriptions of the terrain the author has profited much from the published works of French students, especially the series of admirable geographic monographs which includes Raoul Blanchard's "La Flandre," Albert Demangeon's "La Picardie," Emile Chantriot's "La Champagne," and the smaller work of Bertrand Auerbach entitled "Le plateau lorrain." Frequent citations of these works will be found in the text, but it is only fair to state that in endeavoring to paint for the reader a picture of each battlefield the author has, consciously and doubtless also unconsciously, drawn upon the works in question to an extent which cannot fully be indicated by specific citations. His own mental image of the battlefields is due to the excellent descrip- tions of his French colleagues, as well as to his personal studies on the ground; and if the attempt to transmit this image to the reader has been successful, no small share of the credit must go to the distinguished Frenchmen who are well-known authorities on the geography of their country. For detailed illustrations of the influence of different topographic elements upon certain phases of the fighting, Arthur Conan Doyle's volumes on the British campaigns in France and Flanders have repeatedly been laid under tribute, because they present in INTRODUCTION xxv a telling manner the reaction of individual units to the difficul- ties which opposed them. For those photographic illustrations which are not from the author's own negatives, acknowledgments are due to the Photo- graphic Services of the French, British, Italian, and Belgian armies, as indicated under each reproduction in the text. The block diagrams, with one exception, were prepared under the author's general direction by three of his former students: Pro- fessor F. K. Morris, of Pci Yang University, Tientsin, China; Professor A. K. Lobeck, of the University of Wisconsin; and Professor S. H. Knight of the University of Wyoming. As the author has modified these drawings to meet special needs dis- covered during the preparation of the text, it is but just for him to credit their excellence to his three colleagues, without making them responsible for the precise form in which the drawings now appear. In drawing the generalized sketch maps of the battle- fields the author has adopted the simple and effective method employed by Professor W. M. Davis in his "Handbook of Northern France," and in some cases has utilized portions of his maps, modified to meet the needs of the present volume. To Miss Ellen Churchill Sem pie thanks ?re due for permission to use certain data assembled by her relating to the passes and routes of the Alps. The Library of Columbia University has extended exceptional courtesies both in the matter of placing large numbers of books, maps, and other documents at the author's disposal while he was absent from the city, and in pro- viding expert assistance in tracing records and verifying data essential to the work. Special acknowledgments are made to Miss I. G. Mudge, Reference Librarian, for invaluable assistance throughout the preparation of the manuscript. A very heavy debt of gratitude is owed my colleagues in the Department of Geology at Columbia University, whose generous sympathy and fraternal assistance have alone made it possible to complete the work in the allotted time. To the American Geographical Society, and to its Director, Dr. Isaiah Bowman, it is a pleasure to record the author's hearty xxvi BATTLEFIELDS OF THE WORLD WAR appreciation of the warm support they have given his under- taking from its very inception. Not only has the financial sup- port been prompt and generous, but in such matters as the pro- viding of needed secretarial assistance, the preparation of maps and diagrams, and the meeting of other needs which arose in the course of the work, the author's every request has met an imme- diate and cordial response from the Society and its officers. Many improvements in the text have resulted from the valuable suggestions of the Society's Research Editor, Mr. W. L. G. Joerg, to whose untiring co-operation the author is indebted for much aid which lay far beyond the usual limits of editorial assistance. Finally, it is the author's privilege to express his thanks to General Tasker H. Bliss, who generously undertook to read the proof sheets of this volume, and who brought to the task an un- equaled breadth of knowledge based on his long and dis- tinguished military, career, his membership on the Inter-Allied Supreme War Council and on the Armistice Commission, and his service as one of the five American Commissioners to Negotiate Peace. A similar friendly service was performed by General Francis Vinton Greene, whose intimate knowledge of military history and well-known ability as an author and critic made his suggestions and criticisms extremely helpful. To both General Bliss and General Greene the author would acknowledge a very great debt of gratitude. Many improvements in the text have resulted from their generous assistance ; but on the author alone rests the responsibility for statements of fact and expressions of views which the text sets forth. Douglas Wilson Johnson Columbia University December 31, 1920 CHAPTER I THE BATTLEFIELD OF FLANDERS: THE WET CLAY PLAIN BATTLEFIELD One afternoon in September, 191 8, a British staff automobile left corps headquarters "somewhere in Flanders," sped eastward over good roads for a few miles, then plowed into the ruts and mud holes of a newly recaptured portion of His Majesty King Albert's dominions. It was typical Flanders weather, for a driz- zling rain was falling and low clouds or banks of fog drifted over the plain. The machine skidded into a slimy shell hole and half overturned. Three officers stepped out, belted their raincoats tighter about them, and slopped on through the sea of mud, into the wall of mist. A quarter of an hour later you might have seen them toiling up the slope of a low hill along a sandy path winding pictur- esquely through a wood of conifers, each heavy step weighted down with a mass of clay which clung tenaciously to boot and spur and picked up sand, leaves, and twigs from the narrow trail. Muddy and bedraggled, the three crept in between the sand bag walls of a shelter near the crest and waited. The chill wind drove the lowering clouds about them, and through the gray curtain there came from the north, east, and south the rumble and crash of a great battle. Then, as if raised by magic hands, the fog curtain slowly lifted and parted. A flood of golden sunshine burst through, lighting up a vast green-carpeted plain on which rivers and ponds glittered like silver spangles. Stretching in a vast crescent across the stage thus revealed to the waiting observers was a line of flashing tongues of flame, a semicircle of steel and fire which, from the sea on the northwest to the uplands of. Artois far away to the southward, was slowly blasting the Note. For Chapters I and II the reader should constantly consult the detailed map of the battlefield in the pocket (PI. I) and the block diagrams (Figs. 14 and 36). 2 BATTLEFIELD OF FLANDERS Kaiser's hordes back into the Fatherland. It was the last great Battle of Flanders in full swing, and from the summit of Mont Kemmel our officers were with their glasses sweeping the vast field Fig. i — Index map to the maps of the battlefields of the western front. For each rectangular area shown here there will be found in the text a generalized sketch map (Figs. 2, 24, 56, 79, 98), and in the pocket at the back of the volume a detailed large-scale map (Pis. I-V), both of which should be consulted while reading the two chapters devoted to each battlefield. from end to end. For the low sandy hill, scarcely to be discovered when one stands among the trees and orchards of the plain, is none other than the famous "Mount" (Fig. 3), stained red with the blood of many thousands and standing in the very center of a GENERAL ASPECT battlefield which numbers its victims by the hundreds of thou- sands. Its summit rises barely 500 feet above the sea and only 350 feet over the adjacent plain; yet it is one of the dominating points of Flanders. In all the vast plain bounded on the north- west by the sea, on the south by the chalk uplands of the Somme battlefield and the foothills of the Ardennes, and extending east- ward beyond Brussels and Antwerp, there are only two other points higher; and both of these are in the same range of hills as Mont Kemmel. Look westward and one may see them, if the trees are not too much in the way: Mont des Cats (Fig. 4), some 550 feet high, with the lesser heights of Mont Rouge, Mont Noir, and others lying between; and far- ther on, at the western end of the ridge, the king of the Flemish moun- tains, Mont Cassel, boasting all of 560 feet. But one does ill to scoff at these tiny mounds, which over-enthusiastic patriots have compared with the Alps. Each summit commands a view of which a real mountain would have no cause to be ashamed: in the foreground a charming landscape of green meadows bordered by rows of trees and dotted with red- tiled farmhouses; here and there a shady grove or picturesque Flemish village; far away to the northwest the green of trees and meadows, bounded by a thin band of white, the sand dunes along the coast. The Fig. 2 — Generalized sketch map of the Battle- field of Flanders. Ruled areas are uplands; the dotted area is a coal field. For the topographic details and place names referred to in the text see PI. I and Figs. 14 and 36. 4 BATTLEFIELD OF FLANDERS eye sweeps the vast plain (Fig. 23) to the north, east, and south and is arrested only when it discovers to the southwest, barely emerging from the blue mist on the horizon, a low elevation which can be nothing else than the historic Vimy Ridge. The view is broad, not because the hill is high but because the plain is so extremely flat. One is tempted to say "absolutely flat," and the exaggeration would be pardonable. Very few places Fig. 3 — Mont Kemmel, a strong point near the apex of the natural "Ypres bas- tion." Although of moderate elevation only, it offered commanding observation over a vast expanse of the plain; and hence its possession was bitterly contested in important battles. When the photograph was taken the Germans held the eastern (right-hand) portion of the hill, the British the main crest. in Europe can show a surface so nearly level. Roads, railways, and canals cross the country in every direction, and usually in straight lines, for seldom is there even a modest elevation to be avoided. The whole extent of Flanders provides but a single railway tunnel, and a canal may run for many miles without a lock. The rivers seem to flow on the plain instead of cutting valleys in it, and hills are so rare that even the most insignificant attracts more attention than many alpine crags. Houses and trees are usually the most imposing features in the landscape. Surely if ever a region was fully entitled to the term plain, Flanders is that region. GENERAL ASPECT 5 Of the battlefields of the World War the rolling surface of the Somme region just to the south can best be compared with the Flanders plain. Yet how few the elements of correspondence, how many the points of contrast! Two different worlds seem to come together where the chalk of the Somme region meets the clay of Flanders. One is a rolling upland, the other a level low- land. The even-flowing, clear streams of the chalk country do : : - 1 Fig. 4 — Mont des Cats, in the middle distance, rising slightly above the general level of the plain of Flanders and serving as a natural observatory of the highest importance. not in the least suggest the unruly rivers whose turbid waters repeatedly flood the fields and homes of the Flemish peasant. The boundless horizon of the treeless Somme plain is poorly counterfeited in the usual view on the plain of Flanders, where apparent vastness is nullified by the trees which limit one's vision. Nothing could be farther apart than the aridity of the chalk surface and the humidity of the water-soaked clay plain. The dreary stretches of open fields on the chalk, devoid of human habitation except where thickly clustered dwellings form a village, find no counterpart in the green, tree-bordered meadows and gardens among which are scattered the homes of the widely disseminated Flemish population (cf. Figs. 11 and 12). Surely the imagination can picture no two geographic regions belonging in 6 BATTLEFIELD OF FLANDERS the same class which are more dissimilar than the plain of Artois and Picardy and the plain of Flanders. The real dissimilarities are so pronounced that they lead to belief in others which do not exist. Ask the peasant who knows both regions, whether it rains more in Artois or Flanders, and he will tell you the latter receives much more rain. Yet there is only one place in all Flanders where the rainfall is not less than the lowest rainfall on the Artois upland. 1 The smaller rainfall on the flat clay plain cannot quickly escape, and produces a far greater impression on man than does the heavier rainfall, which soon disappears into the fissured chalk. Again, ask the peasant whether the wet and dry seasons are the same in Artois and in Flanders, and he may answer that there is no dry season in Flan- ders; that it rains often and much, the year round. But the Flemish rivers rise in the summer and autumn, and a study of sta- tistics demonstrates that, while the sky of this country is much of the time overcast and some rain falls on an average nearly every other day throughout the year, the greatest precipitation occurs in the summer and autumn months. The peasant remarks the many days of rain but does not measure the quantity; and the soldier is apt to be like him. He is more affected by the length of time he shivers in wet clothing and stands in wet trenches than by the amount of water that falls on him. Therefore he will tell you that the low belt of plain nearest the coast is the rainiest part of Flanders, when in reality it is the zone of least precipitation; and will yearn for the "less humid country" to the south, where nearly twice as much rain falls! The physical conditions of the Flanders battlefield responsible for such effects will soon claim our attention. Strategic Importance of the Flanders Battlefield To one standing on the Mont des Cats and looking out across the low plain of Flanders eastward toward Brussels, Antwerp, and the gateway of Liege, southward toward Arras and the 1 Raoul Blanchard: La Flandre: Etude geographique de la plaine flamande en France, Belgique et Hollande, Paris, 1906, p. 27. STRATEGIC IMPORTANCE 7 gateway to Paris, westward toward the sea and the Channel ports, the strategic significance of the region cannot be wholly lost. It forms part of that great belt of plain across which one may travel by rail from the Pyrenees to northeastern Russia without passing through a single tunnel and without rising 600 feet above the level of the sea. Across it lies the only path by which armies may advance into France without encountering formidable mountain or plateau barriers. But it is more than a pathway; it is one of the important meeting places of northern and southern Europe. English, Dutch, Danes, Scandinavians, and Germans from the north, Venetians, Genoese, Spaniards, and Frenchmen from the south, have throughout the centuries met on the Flanders plain in commercial intercourse or armed conflict. Whether one prefers the view that "Flanders occupies perhaps the most superb international situation which exists in Europe," or the more pessimistic conception that "when God had made this good Flanders he put it between all in order that it might be devoured by one after another," he cannot doubt the very great significance of its location. On its level surface have been fought out some of the greatest struggles of history: a few of them slowly and unimpressively, like the conflict between the Latin and Teutonic tongues; many of them in the rush and furor of battle, when the warriors of many nations clashed in arms and some of the world's famous generals made or lost their reputations in the mud and marshes of the plain. This battle- field of the World War has always been the battleground of Europe. In 1914 the Flanders plain offered the German General Staff something more than the smoothest pathway between moun- tains and sea along which to launch its enveloping movement designed to crush the French armies in the space of a few weeks. The plain was provided with that abundant network of roads, railways, and canals which is the natural product of a dense population inhabiting a region of very little relief. Nowhere else could the Germans find such admirable facilities for transporting and supplying a great army. An intelligent agriculture had 8 BATTLEFIELD OF FLANDERS made the loamy parts of the plain highly productive, while the region was also noted for its rich pastures and its cattle. Here, then, was the food supply necessary for the support of unex- ampled concentrations of men. Once in undisputed possession of the plain, the Teuton armies could continue southward through the low gateway (p. 92) between the Ardennes Mountains and the upland of Artois, to outflank the French armies and capture Paris; or could strike farther west to gain the Channel ports, threaten British communications with the Continent, and bring to the very doors of England the menace of imminent invasion. The complete extinction of a country which had dared to antago- nize theGermancolossuswould be accomplished when all the Flan- ders plain was occupied, and a valuable lesson thereby given to other countries still neutral; while the political advantages to be derived from possession of Flanders at the end of the war would be enormous. Along the southern part of the Flanders plain lay the rich coal fields of northern France. About them had grown up densely populated industrial centers whose products were vital to the economic and military power of the Republic. To seize this region would strengthen Germany's fighting machine and weaken the opposition it would have to overcome. In German opinion the blow might well realize the aspiration of the Teutonic militarists, as expressed by Bernhardi, "to crush France so completely that she can never again come across our path." But if the Flanders plain offered Germany high inducements as a route for the invasion of France, it imposed obstacles pecu- liarly formidable. Far more important than any topographic bar- rier was the neutrality of Belgium, the violation of which by Ger- many would call into action moral and material forces capable of exerting a decisive influence upon the issue of the conflict. Even the topographic obstacles were far more serious than they may have appeared; and as soon as Allied man power should be as- sembled for their defense they would prove impassable. Notwith- standing the fact that both the Central Powers and the Allies repeatedly launched major offensives on the Flanders plain, not SURFACE FEATURES 9 one of them was successful after the first German onrush in 1914 before the natural barriers were adequately defended, until the Allied advance late in 191 8 after the German armies had . already been defeated elsewhere and compelled to begin their withdrawal to shorter lines. On the Somme plain, along the Chemin des Dames, farther south on the Marne plateau, and elsewhere, great drives succeeded; but on the Flanders plain all failed because of physical obstacles which deserve our careful consideration. It is to the detailed topography of the Flanders plain that we must now turn our attention. Beginning thus with the lowest, flattest, and most monotonous of all the battlefields of the World War, we shall pass next to the slightly higher and more dissected plain of the Somme, then in orderly sequence to the trenched plateau of the Marne, the still higher plateaus of the Verdun country with their intervening lowland belts, the loftier domes and ridges of the Vosges, and finally to the towering peaks and precipices of the Alps and the complex mazes of the Balkan ranges. Surface Features of the Flanders Battlefield Look over the Flanders plain from the summit of Mont des Cats and you will see no evidence of system in the topography. Rivers and canals seem to run at will in every direction across the level surface, while the few low hills are scattered irregularly here and there. Now take a detailed topographic map of the region and search there for any sign of symmetry in the surface of the country which might escape the more limited vision of the unaided eye. Again all seems hopeless confusion. Rivers change to canals and back again to rivers, and wander aimlessly toward the ocean or directly away from it. Roads and railways make a confused network without suggesting any definite pattern imposed by the form of the land. Here and there an isolated hill or low ridge rises suddenly out of the vast expanse of plain. The ridge may trend east-west or north-south. And who shall say whether the Mont Cassel-Mont Kemmel ridge 10 BATTLEFIELD OF FLANDERS Upper sand formatio Cloy formation should be continued eastward to connect with the scattered hills beyond the valley of the Lys, or whether it bends sharply northward to form an integral part of the Messines-Passchen- daele swelling? As for the forest patches, what could be more "hit-or-miss" than their distribution? Nevertheless there is order in the apparent chaos. To see and appreciate it we need only know the very simple geological struc- ture of the region. The rocks are the skeleton upon which Nature fashions the varied forms which we behold upon the earth's sur- face; and, just as the artist must begin by studying the human skeleton if he would end by painting aright the beauty of the human form, so must a knowledge of rock structure pre- cede any understanding of the beauties of natural landscapes. Fortunately in this case the task is an easy one. If we omit certain beds, of im- portance only beyond the limits of the region which particularly interests us, we need consider but four formations as responsible for the form of the Battlefield of Flanders (Fig. 5). At the base is the chalk, really characteristic of the Somme battlefield to the south and only appearing in the southern portion of the Flanders field to give a transition zone possessing some features common to both regions. Patches of clay of variable thickness lie in irregular depressions in its sur- face. Next above comes a series of argillaceous sands, sometimes partially consolidated to form sandstones, from 130 to 170 feet in total thickness. For sake of convenience we will call this "the Lower Sand formation Chalk formation (with clay in surface depressions) Fig. S — Geological column show- ing the succession of formations responsible for the topography of the Flanders battlefield. SURFACE FEATURES n lower sand formation." Overlying it is a great deposit of clay, in places nearly 600 feet thick, containing lenses of sand scattered throughout its mass. It is the formation which more than all others is responsible for the typical Flemish landscape. Let us call it simply "the clay formation." In its upper portion the clay formation gets more and more sandy until it merges with the overlying series of beds, "the upper sand formation." This con- sists for the most part of loose sands with layers of clay frequently interspersed, but toward the top the sands are partially consoli- dated into more resistant sandstone, while one layer is of con- glomerate. The total thickness of the upper sand formation is not great in the central part of the region under discussion but reaches several hundred feet farther north. Geologists have sub- divided these several formations into a much larger number of beds and given each a special name; but for our purposes the simpler division is sufficient. Well records show that all four formations dip gradually down- ward toward the north or northeast (Fig. 6). Now it is the rule that, when a series of inclined beds is beveled across by erosion to form a plain, that plain is "belted," i. e. has parallel zones differing in surface form or character of soil, or both, according as different formations are exposed. Thus parts of our own Atlantic coastal plain have a cotton belt, a pine belt, and a rice belt, named from the products of the soil formed on three different rock types suc- cessively appearing at the surface. The Flanders plain is no exception to the rule, as will appear from Figures 6 and 7. Each of the four formations described above produces a characteristic surface zone, except that the chalk, disappearing and reappearing as a result of changes in the inclination of the beds, gives a mixed or transition zone toward the southeast, but does not show at all in the plain southwest of Hazebrouck (Fig. 7). The lower sand formation is represented by a narrow southern sand belt, the thick clay formation by a broad clay belt, and the upper sand forma- tion by a northern sand belt. Turn again to the map (PI. I) and, remembering that sandy soils are apt to be poor and hence not so completely stripped of 12 BATTLEFIELD OF FLANDERS their forests, note the distribution of woodlands. It will now appear that there is a southern zone where woods are more numerous than usual, extending from near St. Omer to the vicinity of Orchies, southeast of Lille. These woods are either on the southern sand belt or near the southern edge of the clay belt where the clay is thinnest; except that about Orchies itself they are in the transition belt. A geological map of the Orchies region would show that there is here an oval basin in the chalk, with the clay formation in its center and the lower sand formation outcrop- ping all around the sides, between the chalk and clay, as shown Fig. 6 — Generalized north-south cross-section across the Battlefield of Flanders, showing the "belted" plain resulting from the northward inclination of the geologi- cal formations. near the right side of Figure 7. The forests really describe an oval surrounding Orchies, following the pattern of the lower sand formation; but the eastern end of the oval lies beyond the limits of our map. North of Roulers there is a second zone of more abundant forested areas, corresponding to the northern sand belt. Between the two zones stretches a less forested region, in the midst of which the sand-capped ridge of Mont Cassel-Mont Kemmel is partly wooded. Assuredly there are other patches of forest the presence of which is to be explained on various grounds; but already there begins to appear some semblance of system in the surface features of the Flanders plain. If the different formations of the plain have varying resistances to erosion, the dipping beds ought to produce, in addition to soil belts, corresponding belts of varying relief. Thus the dipping beds of the coastal plain of England give the parallel sloping up- lands, or "cuestas", of the Chiltern Hills and the Cotswold Hills, separated by the Oxford lowland ; and the inclined layers of the SURFACE FEATURES 13 Paris Basin form a remarkable series of parallel cuestas which exercised an important influence on the grand strategy of the war. 2 In both of these cases the rocks making the cuesta uplands were very resistant as compared with the weaker beds eroded into lowlands. When, as in the Flanders plain, there are no really re- sistant beds, but all are weak and worn down to a low surface, we can expect to find only a faint de- velopmentof uplands, corresponding to such faint differences of resistance to erosion as may exist between un- consolidated clays and unconsoli- dated sands. When we study the map with this moderate expectation in mind, we are not disappointed. The Mont Cassel-Mont Kemmel ridge, capped by the upper sand formation in which there are some layers partially consolidated, ap- pears to be the remnant of an east- west trending upland or cuesta which had its continuation in hills extending on eastward beyond the limits of the battlefield (Fig. 2). From near Dixmude a second line of low hills — much lower, it is true, and more irregular in form, but having generally a steeper slope toward the south and a gently sloping upland toward the north — suggests by its typical asymmetrical form and by its position the remnants of a cuesta trending parallel to the Mont Cassel-Mont Kemmel ridge. Just north of it is a third cuesta-like elevation, also very low and inconspicuous but not wholly without significance. Both of these northern cuestas (Fig. 2) are located in the upper sand formation (Figs. 5 and 6). On the basis of this interpretation the Messines-Passchendaele ridge is part of the gentle backslope of the Mont Cassel-Mont 2 D. W. Johnson: Topography and Strategy in the War, New York, 1017, pp. 1-49. See also Chapters V-X, below. Fig. 7 — Diagrammatic sketch map of the topographic belts of the Flanders battlefield. i 4 BATTLEFIELD OF FLANDERS Kemmel cuesta, left uneroded between the headwaters of certain branches of the Lys River on the east and of the Yser River on the west (Fig. 2). Where sloping uplands of the cuesta type are developed, the river system ordinarily shows a certain systematic relation to the cuestas and lowlands. Some rivers, often the principal ones, flow down the initial slope of the plain, which is down the dip of the beds, crossing cuestas and lowlands more or less nearly at right angles; others flow parallel to each other through the lowlands which they have eroded on the weaker beds, to join the streams of the first type or to enter the sea directly. Thus the Thames flows southeast down the dip of the English coastal plain beds, joined by smaller streams coming into it from the Oxford and other parallel lowlands, while the Severn and the Cam flow through the lowlands directly to the sea. We may now discover that the rivers of the Flemish plain have also a trace of this same systematic arrangement. The upper Yser, the Lys, the Deule past Lille, and, beyond the limits of our map, the Scheldt, Dendre, and others flow northeastward down the dip of the for- mations. The branch of the Yser following the marshy lowland from near Lichtervelde westward past Handzaeme to Dixmude lies in the lowland between the first and second cuestas; the stream which now forms the Bruges-Ghent canal follows the low- land between the second and third cuestas described above, and other short streams or canals show a similar orientation; while the lower Aa flows from St. Omer northwestward directly to the sea. If one shades heavilyon a map those parts of the rivers and canals trending from southwest to northeast down the dip of the rocks, or northwest-southeast (often nearly east-west) at right angles to the direction of dip, he will be surprised to note what a large proportion of the drainage system falls into one or the other of these two classes. The fact that the principal streams trend strongly toward the northeast, while the streams in the low- lands are usually not at right angles to them but more nearly east- west in their trend, has been explained by the fact that the uppermost beds, on which the principal rivers first took their SURFACE FEATURES 15 course to the sea, had more of a northeastward slant than the lower beds, in which the branch rivers later cut their lowlands. However that may be, we need only remember that there is a system of northeast-flowing rivers and canals in Flanders, crossed by another system trending more or less nearly east-west, how- ever much they may be obscured by other watercourses travers- ing the level plain in all possible directions. After the four principal formations of the region had been beveled by erosion to form the four topographic belts already described, the northwestern part of the plain was depressed below the sea and covered with 60 to 70 feet of marine sands. Then the sea withdrew, and the submerged area changed to a fresh-water morass in which peat was formed in thicknesses varying from a few inches to as much as ten feet or more. During the fourth and fifth centuries of the Christian era the sea again covered this maritime zone, burying the peat under five or ten feet of sand and clay. As these deposits gradually built up to the surface of ordinary high tide, and as the formation of the dune ridge along the coast and the building of artificial dikes helped to keep out the sea, we now have a fifth belt some ten miles broad, cutting across the other belts at an oblique angle and often called the "maritime plain" (Fig. 7). Over the surface of Flanders there is spread a thin coating of loam which usually conceals from view the older formations pre- viously described. This loam is sandy toward the bottom, more clayey in its upper part. Where it mantles the clay formation of the extensive clay belt it is as a whole much more argillaceous than where it covers the sand belts. Apparently the underlying formation influences the character of the thin overlying loam. Certainly it is the fundamental beds which determine the topographic belts of Flanders, for the superficial mantle of loam affects the topography but little. Only on the chalk of the tran- sition belt do we find the loam fundamentally different from the basal formation. There it plays an independent role of some importance, as we shall see when we study its typical develop- ment in the Battlefield of the Somme. In Flanders the loam 16 BATTLEFIELD OF FLANDERS CO i> CD <« to d TRANSITION BELT 17 reflects the form and character of the underlying beds so well that one is apt to lose sight of it in describing the features of the country. THE TRANSITION BELT Let us, having in mind the general picture of the Flanders battlefield which we have just gained, consider next those fea- tures of the different topographic belts most likely to affect mili- tary operations. The transition belt lies lower than the Somme battlefield to the south but higher than the typical Flanders plain farther north. From the base of Vimy Ridge the rolling chalk surface, partly coated with loam, slopes gradually downward toward the northeast, forming the faintly inclined plane which the natives call the Gohelle (Fig. 6). It is in the deeper rocks, under this part of the chalk, that are found the coal deposits mined at Lens and neighboring localities. A typical view of the Gohelle may be had from the summit of Vimy Ridge (Fig. 8). The nearly treeless undulations of the chalk appear barren and dreary in contrast with the tree-bordered gardens and meadows of the clay plain beyond. An occa- sional woodland patch and lines of trees along the roads relieve the monotony of open fields. Villages are numerous, but they are usually the ugly corons, long rows of red-roofed miners' houses, built together and all of the same plain style of architecture. Uglier still are the huge black dumps of waste from the mines, often wrongly called slag piles, which may rise a hundred feet or more above the surface of the plain and form the most promi- nent feature in the landscape. Their "command" of the adja- cent country gave them a high military value. For, while less flat than the clay plain farther north, the Gohelle is nevertheless a surface of such faint relief that from an elevation of a hundred feet the observation is really commanding. From our vantage point on Vimy Ridge not only are Givenchy-en-Gohelle, Vimy, Oppy, and other near-by villages in full view, but the great mining center of Lens, Drocourt where began the Drocourt-Queant switch of the Hindenburg Line, Loos of bloody memory, and even the more distant Vermelles and the towers of Vv ingles are easily 1 8 BATTLEFIELD OF FLANDERS distinguished, together with a score of other villages and towns which help to make up this important mining district. Topo- graphic barriers are few; but the underground workings of the mines provide secure shelter for vast quantities of military stores; the corons and other villages and towns are well adapted to serv- ing as fortified strong points in a network of trenches; cellars and caverns in the chalk offer concealment and protection from bom- bardment to large bodies of troops; the mine dumps give valuable observation ; while the network of roads, tramways, and railways necessary to serve a mining region furnish abundant means of transport for military purposes. Except for its lower relief and the presence of the coal mines and their accessory features, the military aspects of the Gohelle resemble those of the Somme battlefield farther south. We will reserve fuller discussion of the effect of the chalk upon military operations until we consider its typical development in that field. Among the few topographic barriers in the chalk of the transi- tion belt the marshy valleys of the Scarpe and Sensee Rivers deserve special attention. The Sensee might be taken as the boundary between the transition belt of the Flanders plain and the typical chalk plain of the Somme; and both it and the Scarpe have their sources in the latter region. But their military history has perhaps been more particularly bound up with that of the Battlefield of Flanders. Both rivers have long been regarded by French military authorities as important links in the northern defensive system of their country. The broad marshy valley of the Scarpe (Fig. 9) prolonged eastward, from near Arras, the natural barriers of Vimy Ridge and the rectilinear valleys of the Canche and Authie. Today the stream is canalized for much of its course, and the marshes partly drained; but inundations can still turn the valley into a serious military obstacle. At its junc- tion with the Scheldt the river helps to make a natural quadri- lateral, formerly of much military importance, enclosing the great Forest of Raismes, bounded on three sides by marshy river valleys capable of being inundated at will and further strength- ened by military engineering works at three of the corners. The TRANSITION BELT 19 marshes of the Sensee are broader than those of the Scarpe, and much of the area "is at all times of the year a large bog or quag- mire and, following heavy rains, a veritable lake almost a kilometer in breadth, at some points filling the river basin. On each bank peat bogs prevent access to the stream. The Sensee Fig. 9 — The marshes of the Scarpe, from earliest days one of the important natural defensive barriers of northern France. Compare this humid, verdure-clad valley with the dry chalk upland shown in Figs. 29 and 43. (From a photograph by A. Demangeon.) Canal and the stream itself make the valley still more impassa- ble. ... In case of war, locks permit raising the water surface 2 meters higher." The junction of the marshy Sensee with the marshy Scheldt at Bouchain (Fig. 42) has always been a point of much strategic value. The northern side of the famous quadri- lateral known as Caesar's Camp (p. 125) was formed by the Sensee marshes, and Bouchain defended one of its corners. When the French under Marshal Villars were striving in 1710- 171 1 to prolong the struggle against the Allies, led by the brilliant Marlborough, they took refuge behind the Scarpe and Sensee 20 BATTLEFIELD OF FLANDERS barriers. Villars inundated the Scarpe to Biache eight miles east of Arras- by damming it, whence the line crossed to the Sensee marshes by a canalized depression which was likewise flooded. Thus were these natural obstacles organized into a part of the famous ne plus ultra line which taxed the military genius of Marlborough to the utmost. The floods of the Sensee could be crossed only on two causeways, one at Arleux* and the other at Aubigny, both of which were protected by fortifications. So formidable was the barrier that it could not be taken by direct assault, and Marlborough had recourse to a most remarkable and complicated stratagem which resulted in giving him posses- sion of the crossings. 3 Marlborough's despatches are filled with references to the "morasses" of the Sensee, Scarpe, and Scheldt which now prevented him from attacking the French, now pro- tected him from enemy assaults. In 1794 the French and their Allied enemies were again facing each other across the barrier of the Sensee marshes. These are but scattered instances in a long history in which the Scarpe and Sensee played an important role, the later chapters of which saw the French Army of the North retreating behind the line of the Scarpe after being defeated by the Prussians near Amiens in 1870, and the Germans shielding themselves from Allied attacks by flooding the Sensee marshes during the war of 1914-1918. The chalk of the Gohelle dips under the oval basin containing the lower sand and the clay formations (Fig. 7). Within this oval, rudely outlined by forest patches following the lower sand out- crops (p. 12), one finds the typical landscapes of the low clay plain characteristic of most of the Flanders battlefield. It is a bit of real Flanders in the midst of the chalk, and has been given a spe- cial name, the Pevele. The regional name is sometimes attached to a village name, after the French custom, as in the case of Mons- en-Pevele. Beyond the basin the chalk again bulges up to the sur- face, forming the region called Melantois just south of Lille and * Unless otherwise stated, places mentioned in Chapters I and II may readily be located on PL I in the pocket, on or near the river or other topographic feature with which the names are associated in the text. 3 J. W. Fortescue: A History of the British Army, 7 vols, to date, London, 1899- 191 2; reference in Vol. 1, pp. 540-547. SOUTHERN SAND BELT 21 Carembault farther southwest (Fig. 6). Both these regions repeat the topographic features characteristic of the chalk plain of Gohelle, except that the elements contributed by the coal mines are largely lacking. THE SOUTHERN SAND BELT As the lower sand formation outcrops only as a narrow band and is there characterized by no particularly resistant or non- resistant layers, it does not exercise a striking effect upon the landscape. Forest patches are, however, a little more frequent on or near it, and the bare hills of the chalk begin to give place to a more pleasing landscape of meadows and gardens interspersed with trees. The sand formation has, furthermore, real importance as a source of water supply. Rainfall penetrating the porous sand descends to great depths, held in by the clay covering of the chalk below and by the great clay formation of Flanders above. Wells driven through the clay formation to pierce the lower sands give some of the most abundant flows of water in all Flanders. This is an item of considerable military importance, for we shall discover that, curiously enough, in the water-soaked plain of Flanders it is a difficult matter to secure sufficient uncontaminated water to supply the large demands of a great army. THE CLAY BELT It is on the low, flat clay plain that the topography of Flanders finds its typical expression. The fact that the clay is fine-grained and not yet consolidated into rock makes it a more ready prey to erosion than the chalk. Hence it has been worn down more rapidly and to a much lower level than the upland of Artois, even lower than the chalk plain of Gohelle in the transition belt. As a whole the clay plain rises but slightly above the level of the sea. The rivers are very little lower than the interstream areas, and pronounced valleys like those in the chalk plain farther south are quite unknown. It has well been said that the whole of the plain is one great valley. If a river happens to cut against a low ele- vation in the clay so as to give a wall a few yards high, it is a topographic feature worthy of special remark. 22 BATTLEFIELD OF FLANDERS On so flat a surface the rainfall finds it difficult to flow away. Nor can it escape readily underground, for the clay is one of the most impervious formations imaginable. Unable to sink down- ward or to flow laterally, the water remains stagnant over large areas, forming ponds and marshes, or rises until it slowly creeps, halting and hesitating as to what course to take, toward one of the sluggish rivers which wander with apparent aimlessness over the level land. The ground is saturated with moisture, whether it be the clay itself outcropping at the surface or the thin deposit of clayey loam resting as a mantle upon it. Either one gives a sticky, slippery mud which is the abomination of Flanders. Even when it dries, as it does when some days of sunshine interrupt for longer than usual the succession of cloud, mist, and rain, the clay is an enemy of man; for the peasant cultivating the hardened soil must pour water on the plow and get men to help force the share into the ground as it moves painfully forward. The hardened mass cracks open in all directions, letting the next rain descend into the fissures to make the sides of the clay blocks so slippery that the divided mass glides on itself, thus producing a landslide where there is slope enough to permit its descent to obstruct some railway, road, or canal. Flanders Mud Flanders mud demands more than mere passing notice. In the early days, before metaled roads and railways had made their appearance and before the system of canals was so extensive as now, the mud made large parts of the country difficult of access — some parts absolutely inaccessible. This condition endured through the seventeenth and eighteenth centuries, for in 1635 Bailleul could barely procure the necessaries of life during winter, and that thanks to navigation upon the becque, the roads being too bad. The intendant Barentin states in 1699 that the rich castellany of Cassel could not, on account of the lack of passable roads, distribute the provisions it produced nor assist neighboring castellanies suffering want. A project for the defense of the country after 1735 declares that between Armentieres and Cassel the main highways are impracticable almost the entire year. . . . The prefect Dieudonne in 1804 refers CLAY BELT 23 sorrowfully to the bad condition of the roads "which have, in the nature of the soil, a perpetual cause of deterioration, the low, wet, and often peaty soil having no solidity." The state of the roads was still worse in the plain of the Lys. The character of the ground, even more clayey and wet than around Cassel and Ypres, rendered them impracticable in all seasons. In 1761 the town of St. Venant complained of being "inaccessible through the bad condition of the roads;" and in 1766 the sheriffs explain that in wet periods it is the mud which prevents approach to the town, while in dry periods it is the ruts which are so frightful "that it is impossible to pass either on horseback or in carriage without risk of smashing everything." For several years the old Roman road from Cassel to Arras remained impassable between Estaires and La Bassee at a spot called 'le Trou Gallot', "where opens an abyss which must cost the lives of all who pass that way." The only means for those on foot to traverse the country in winter (for travel by carriage was scarcely to be considered) was to jump step by step, on the blocks of Bethune sandstone [from the southern sand belt] which the foresight of the authorities had caused to be placed along the side of each road and which were called pierres de marchepied, or stepping stones; shod with shoes heavily metaled to prevent slipping upon the stones and disap- pearing in the mire up to the waist and armed with long poles to aid in jumping from one block to another, the people of the country became so used to this mode of traveling that accidents were rare. 4 Today the main highways are paved and many of the lesser roads improved. But an army cannot restrict its movements to a limited number of roads; and, when it must fight in any given region, modern methods of warfare compel operations on almost every square foot of the terrain. Thus the mud of Flanders remains an element of the first importance with which military leaders perforce must reckon. When Philip Augustus let his army become trapped in the morass southwest of Ypres in 1197, he but shared the experience of many before and after him, from the days before the Roman conquest to the days when the Ger- man General Staff and the Allied commanders saw one great offensive after another stop short in the mud. It would be diffi- cult to enumerate all the ways in which the seemingly unfathom- able mud of Flanders affected the fighting powers of the opposing armies in the fateful years 1914-1918. The damage done by 4 Blanchard, La Flandre, pp. 446-447. 24 BATTLEFIELD OF FLANDERS artillery fire was greatly reduced when the shells exploded in a sticky clay. Shell holes filled with water which could not drain away, turning the battlefield into an almost impassable morass which blocked the advance for which the bombardment was supposed to be a preparation. Munitions and other supplies could not be hurried forward in time to support an advance properly, and reinforcements of troops floundered in the mire behind the lines while their hard-pressed comrades were forced to relinquish captured positions for want of prompt assistance. Heavy guns could not leave the metaled roads, were delayed by the congested traffic confined to these narrow paths through the wet plain, and could not be distributed at will when they reached the firing line. Even the light field artillery used in supporting an attack had to be equipped with material for making the shell-torn surface passable before it could advance, and one artillery officer is reported to have said: "I am carrying forward my guns and ammunition, the material for making my road as I go along, and the material for fortifying my new position. . . I am half expecting orders to bring along an acre of ground with me, too." 5 Assaulting columns found it difficult to scramble out of the slippery trenches and were mowed down by enemy fire as they advanced slowly through a tenacious clay into which they sank more than ankle deep. Rifles became so clogged that they could not be fired; and, when they were wrapped in cloth to keep the mechanism clean, were not ready for instant use. The wounded lay half buried in the mud, and many were suffocated. Even the well and strong were caught in. fatal mud traps, for detailed offi- cial reports on the operations in Flanders contain not infrequently such statements as "part of Company bogged in communication trench south of St. Eloi; two men smothered" and "three men suffocated in mud near Voormezeele." In a British assault on the low clay mound near St. Eloi in April, 1916, the attackers had to lie flat and distribute their weight evenly in order to prevent sinking into the mire. As it was, a number of the men were engulfed and suffocated. In many parts of the plain trenches 6 Frank Fox: The Battles of the Ridges, London, 1918, p. 95. CLAY BELT 25 slumped in so fast that new positions could not be consolidated, and important points captured at a heavy cost in lives had to be relinquished because the survivors could not protect themselves in the soft clay before heavy counterattacks were delivered. Even well-established trenches required constant repair. The effect of ever-present, everlasting mud on the morale of an army is a factor difficult to evaluate but certainly not to be ignored. The mere labor of keeping rifles, guns, shells, and other equipment dry and clean is alone a heavy task. The cleaning of gun carriages, automobiles, trucks, horses, and other means of transport adds to the burden. Superhuman efforts would be necessary to keep men and equipment up to that high standard of cleanliness which has earned for the British troops the enviable if somewhat unpoetic name of "the spit-and-polish army." In Flanders the soldier's best efforts left him discouraged. Forever busy cleaning the sticky deposit from himself and everything about him, he forever found himself and his equipment caked with the mire. Cold, wet, tired, and disgusted, the unhappy fighter in Flanders would crawl into his straw-floored dugout, leaving his clay-coated shoes at the entrance, and lie shivering as he cursed the eternal mud which was by far his worst enemy. The author has visited the fighting fronts from the sand dunes of the Belgian coast to the entrenched camp of Saloniki and observed during the conflict the conditions under which men fought from the polders below sea level to the glacier-clad heights of the Alps. He has no hesitation in saying that, of all the com- batants, those who fought on the plain of Flanders endured the most terrible physical conditions. The Rivers of the Clay Plain Mud is not the only enemy of an army condemned to operate on the clay plain. Although the rainfall is moderate and notably less than that on the arid chalk uplands of Artois, we have seen that it is held at the surface by the impervious clay and cannot escape rapidly from so flat a region. This excess of water makes trouble in numberless ways. Where it flows into the sluggish 26 BATTLEFIELD OF FLANDERS rivers they are flooded, for with their faint gradients they cannot rapidly discharge the large total volumes thus received. When they rise, their flood waters spread far and wide, because the country is so low and flat. Almost all the streams of the plain, from the smallest rivulet to larger rivers like the Lys, have one or more floods every year, especially in the late summer and fall and in the early winter. Half a dozen floods in the same stream in a single year are not unusual. In 1882 the Yser overflowed its banks fourteen times. The floods of the Lys are the most danger- ous of all in the Flemish plain, and the river has not inaptly been called "the scourge of Flanders." The rising waters not only submerge its broad, shallow valley and flood the city of Ghent, but force the Scheldt to back up and overflow its banks, fill the Bruges-Ghent canal to overflowing, and interfere with the regime of other canals connecting with its valley. Elaborate works have been undertaken to control the rivers. Many have been straightened and canalized, their rising waters are held in by dikes, their great breadth has often been reduced to a narrow channel, and the extensive marshes bordering their courses have been partly reclaimed. Yet even today disastrous floods are not unknown, and the marshes are still numerous. Where marsh and forest combine, as at Ploegsteert Wood in wet weather, the obstacle is peculiarly formidable and can only be threaded on wooden footbridges, or passerelles, raised above the morass. The river channels are barriers which armies may pass only with the aid of bridges, while the valley floors can be flooded by opening the dikes and the difficulty of passage thus greatly augmented. Almost every phase of the defensive value of rivers, marshes, and inundations discussed by Von Clause- witz in his classic work "On War" finds exemplification on the Battlefield of Flanders. 6 The tactical and strategic value of the rivers of Flanders cannot be doubted by one who reads even a few pages of the long military history of this region. Froissart's "Chroniques" contain many references to the line of the Lys, among them the detailed 6 Carl von Clausewitz: On War, London, 191 1, Vol. 2, pp. 263-295. CLAY BELT 27 and quaint description of how in the latter part of the fourteenth century Philip van Artevelde commanded the destruction of all bridges over the river on a front of fifty miles or more, in order that it should be "not in the power of the King of France nor of his Frenchmen, that they should pass the river of the Lis." The advisers of the French King held the barrier in profound respect, and when asked, "This river of the Lis, is it so evil to pass that one cannot find passage save at the well-known crossings?" one of them replied, "Sire, yes; there is no ford, and the stream flows upon marshes which one cannot traverse." Some advised ascend- ing the river until the French army could pass around its source, others counseled turning northeast into another part of Flanders where the advance would be blocked by no such obstacle. Bolder spirits urged an attack on the river barrier, assuring the others that God would aid them to make a crossing; and their advice prevailed. But so difficult did the task prove that even those who urged the attack repented that they had not taken the long route round the river sources or marched in some other direction across an easier terrain. When success was finally achieved, it was by means of a surprise crossing effected secretly with small boats at a poorly guarded point. 7 In the succeedingpagesofthe"Chroniques"onereadsthesamecomplaints of marsh and mud, the same references to soldiers standing knee- deep in the mire, which became so familiar in the despatches of 1914-1918. If we turn to the wars of the French Revolution, we find the Allied armies holding the line of the Lys early in 1794, then fleeing in disorder when the French capture a section of the barrier. Next it is the French who make the line of the Lys their position, advancing from and returning to it with the exigencies of the campaign. In the disastrous Allied offensive in May, known as the Battle of Tourcoing, Clerfayt's corps, one-fourth of the whole army, was held up by the French standing behind the Lys at Wervicq and so counted for nothing in the first day's operations. Two other columns were held up at the Marque, "a stream impassable except by bridges, owing to soft bottom and 7 Jean Froissart: Chroniques, edit, de Lettenhove, Brussels, 1870, Vol. 10, pp. 106-126. 28 BATTLEFIELD OF FLANDERS swampy banks." In the preceding year the line of the Yser at Bambecque was held so effectively by the Allies that ten times their number of Frenchmen were unable to dislodge them by direct attacks; while the line of the Yser-Ypres canal from Nieuport to Ypres, which was to figure so largely in the World War of 1914-1918, served as the main defensive position of the Allied armies' right wing in 1794. 8 It is unnecessary to multiply instances in which the rivers of Flanders have in the past served as natural lines of defense. In the recent war the Lys again played an important defensive role, the Yser blocked some of the greatest offensives launched by the Germans, while other streams, including so small an example as Kemmel Brook, contributed in no small degree to the results of operations in the clay plain. Rivers and canals (Fig. 10) com- bined with the mud of the plain to oppose a serious obstacle to one of the most effective of modern offensive weapons. In most parts of the clay plain it was practically impossible to use tanks; in others, especially in more sandy and less humid areas, they were utilized, sometimes by equipping them with fascines of wood which could be dropped in front as the tanks advanced. But, not- withstanding this device to improve the terrain over which they moved, the work of the tanks on the low plain was never so effec- tive as on the chalk uplands farther south. As for cavalry, mili- tary writers have from earliest times pictured the difficulties occa- sioned that arm of the service by the endless network of small drainage ditches, the marshes, the larger canals, and the rivers. Artillery and even infantry find the drainage ditches an obstacle which forces them to keep to the roads or to the larger dikes of rivers and principal canals. Submerged Trenches and Dugouts The presence of an excess of water on or close below the surface proved a serious embarrassment in the operations of trench warfare. Where water stood in ponds or marshes in faint depres- sions of the low plain, trenches were impracticable. Elsewhere even shallow trenches might encounter the zone of permanent 8 Fortescue, Vol. 4, Part I, Chs. 5 and 10. CLAY BELT 29 saturation close below the surface, and so be permanently inun- dated. Even where the land was slightly higher, trenches in the clay caught rainfall which could not drain off through the imper- vious material. Hence the terrible sufferings of men compelled to stand in cold water or liquid mud ankle-deep, knee-deep, sometimes even waist-deep. What this means to an army can Fig. 10 — A canal used as a defense line and crossed by a temporary footbridge. (Belgian official photograph.) only be fully understood by one who has experienced it; but some idea of the truth can be gleaned from reading the reports of officers of all grades, now become part of the British War Office archives. As one report modestly observes, "The difficul- ties of this part of the country are worthy of note. The trenches are very wet, and the water is up to the men's knees in most places." Such phrases as "men knee-deep in water," "trenches full of liquid mud 2 to 3 feet deep," "trenches full of water 3 feet deep in places," "trenches untenable owing to flooding," "ground so wet only able to dig down 2 feet," occur in endless repetition. 30 BATTLEFIELD OF FLANDERS One officer reports his men as "in pitiable condition coming out of trenches; wet through, caked with stinking mud from head to foot, and perishing with cold." The state of the men's feet became unbearable, and much space is devoted to casualties from this cause. Eight hundred men unable to walk were left behind in Fig. ii — Scattered population on the low plain of Flanders, where abundance of water close to the surface enables every farmer to have his well and his home in the midst of his fields. (From Lille sheet of 1:80,000 topographic map of France.) Compare with Fig. 12. a single village. The terse report of another officer speaks vol- umes: "Trenches full of liquid mud. Smelt horribly. Full of dead Frenchmen too bad to touch. Men quite nauseated." Can one marvel that the physical conditions of the Flanders battlefield tried the souls of men as they were tried nowhere else on the long fighting front? If deep trenches were hardly practicable, it goes without saying that still deeper dugouts and all that elaborate system of sub- CLAY BELT 3i terranean fortifications which honeycombed the chalk of Artois and Picardy were impossible in the low clay plain. Small, shal- low dugouts were excavated in places; but shelters built on the surface of bags of earth and other material were essential over broad areas, while railway embankments, dikes, and other Fig. 12 — Compact villages grouped about wells in the Somme region, separated by uninhabited open country. The great depth to water in the chalk formation makes wells few and costly. (From Amiens sheet of 1:80,000 topographic map of France.) Compare with Fig. n. artificial structures were much utilized (Fig. 17). Even the "trenches" were often mere breastworks of sandbags, and attack- ing parties carried with them the quantity of bags necessary to consolidate and hold a new position. Contaminated Water Supplies It might be supposed that with such excesses of water on every hand, the fighting forces would at least be spared the 32 BATTLEFIELD OF FLANDERS trouble of securing proper water supplies. There is, of course, no lack of quantity, and it is easily accessible. The sandy lower portion of the loam mantle covering the plain is full of water which cannot sink deeper because of the impervious clay imme- diately beneath. One has only to sink a shallow well some few feet through the more clayey upper half of the loam to reach this water horizon. Every peasant can afford to have his own well; and, since travel over the muddy plain is difficult, he digs his well and builds his home conveniently in the midst of his fields. This is in striking contrast with the conditions on the chalk plain of the Somme battlefield, where all the peasants of a district will cluster their homes close about a few deep wells dug at the common expense. A map of the Flanders plain (Fig. 11) con- trasts strongly with a corresponding map of the chalk region, with its compact villages and uninhabited open country (Fig. 12). In the defensive systems organized by the opposing armies in Flanders there was little to suggest that network of isolated fortified towns and villages which figured so largely in the Somme campaigns. But if water was close at hand it was also easily contaminated. Surface wash, carrying with it impurities from every possible source of filth, came to rest in the lower part of the loam, where it was held by the clay below. The muddy rivers and canals, dis- colored and poisoned by refuse from the manufacturing cities and towns along their courses, not only were unfit for use them- selves but often contaminated the shallow water zone of the loam into which the peasants sank their wells. The loam hori- zon was, furthermore, too limited to supply the huge quantities of water demanded by countless great manufacturing estab- lishments. Hence the water-soaked Flanders plain became a region of numerous deep and costly artesian wells. When vast armies camped upon the plain in 1914 the water problem became aggravated, and especially so after drainage washed the decaying bodies of the battlefield and passed into the water-bearing horizons. Even the deeper waters were often poisoned by surface supplies which penetrated down the sand CLAY BELT 33 formations or down the sandy lenses in the clay formation. Hence the development of sanitary supplies of water in the enormous quantities required by large armies created a difficult problem. Nothing less than a full knowledge of the underground structure of the battlefield could enable an army to deal success- fully with the vital matter of determining which layers in the different formations would carry uncontaminated water and at what depths they would be encountered in different parts of the field. Vegetation of the Clay Plain Another consequence of the humidity of the clay plain is the vigorous growth of vegetation which covers it with a mantle of green. There is not now, nor does there ever appear to have been, any such vast forest as once covered the rolling chalk plain of the Sorame region. In its original state a succession of grassy meadows and swamps studded with trees and inter- spersed with more continuous woodland patches, the vegetation has since been modified by man so that today rectangular patches of meadow, field, and garden cover most of the land, the scattered trees are less numerous and are largely confined to the borders of roads, canals, and rivers, or to the margins of meadows, fields, and gardens; while the patches of veritable woodland are much more restricted than formerly. Wherever one may be on the clay plain, he has the impression of standing in the midst of a large and rather intensively cultivated clearing in a forest. When he advances to where the forest seemed unbroken, the woodland dissolves into scattered groups and rows of trees bordering culti- vated areas or transport lines, and the scene is the same as before. Distant views are rare, and one has no such sense of the vastness of the plain as he gets on the barren but much more uneven up- land of Artois. It is only when he mounts to the summit of some hill and views the flat country from above (Fig. 13) that he sees over and beyond the trees to more distant parts of the plain. Hence the incalculable military value of even faint elevations on so level a surface. Histories of military campaigns in Flanders 34 BATTLEFIELD OF FLANDERS / CLAY BELT 35 contain frequent reference to the difficulty of observing enemy movements on account of the trees. Aside from the wooded areas on the upper sand formation, found capping the scattered hills and ridges left isolated by erosion here and there over the plain, the principal forest patches are those rooted in the clay formation where it lacks the loam covering. These surface exposures of the clay are called clyttes, and, as they furnish a cold, wet, sticky, inhospitable soil, poorly adapted to cultivation, the peasants often leave them covered with trees. The Houthulst Forest north of Ypres is a good example of a type of woodland in which forest fighting is at its worst, because to the other difficulties usually presented by a fortified wood are added the horrors of a water-soaked, tenacious soil giving a morass in which troops cannot entrench. Hills on the Clay Plain The hills of the clay plain are perhaps its most important topographic feature from the military point of view. A general can well afford to sacrifice thousands of men in order to conquer and hold a small hummock a hundred feet or less in height; for an observer on that hill can save tens of thousands of his comrades by directing artillery fire against enemy positions, batteries, troop concentrations, railways, and roads, and other objectives easily located from his point of vantage but absolutely hidden behind a wall of trees to one who stands on the plain itself. An insignificant mound rising only 30 feet above the plain and called "the Bluff" was the scene of terrific fighting in the early part of 1916 because its value for artillery observation made it the key to a section of the British line. To secure control of this one point the enemy excavated tunnels and exploded five mines under it and then launched a successful infantry attack. From the captured elevation the Germans could enfilade other British positions, and two weeks of British counterattacks failed to regain the lost mound. So valuable was the tiny hillock that the British now made more elaborate preparations for its recap- ture. A frontal attack was necessary, since one side of the hill 36 BATTLEFIELD OF FLANDERS was protected by a canal barrier, the other by a marsh. The story of the struggle, of the final successful British assault, and of the later unsuccessful German counterattacks is clearly pic- tured by Conan Doyle in the third volume of his detailed account of "The British Campaign in France and Flanders," while Sir Douglas Haig states that between January 16, 1916, and June 7, 191 7, a total of twenty-seven mines were exploded at this point alone. Anyone who doubts the significance of a 30-foot hill on a flat plain should read the pages of unofficial and official reports on these operations and try to measure the labor expended and the blood spilt in contests for possession of the Bluff. A still more sanguinary conflict raged for six long weeks in the spring of 1916 for possession of "the Mound" near St. Eloi, a small bump of clay about 25 feet high. The British positions on the Mound were violently attacked in the middle of March, fol- lowing a heavy bombardment and the explosion of a mine under the hillock. The attack was successful, and the defenders had to evacuate a considerable stretch of their line, dominated as it was by the Mound. In unsuccessful counterattacks the British, swept by a hail of projectiles from the higher point, in a short time lost over 900 men, including 40 officers. Thereupon they excavated five mines under the Mound, placing as much as 30,000 pounds of ammonal in a single one. These were exploded the last of March, and a furious combat lasting several days resulted in a British victory. Then began a series of German counterattacks continuing for some weeks, in the course of which the mine cra- ters, filled with pools of blood and mud in which the men fought waist deep with rifle and machine guns clogged with dirt, changed hands repeatedly. Not until the end of April was the position again definitely in British possession. To gain Hill 60, a low ridge fifty feet high in the Zillebeek re- gion from which the Germans dominated a broad stretch of the plain in front while their own back areas were concealed, the British on April 17 discharged six subterranean mines and began a contest which raged with fury for weeks. Within four days the British casualties rose to more than 3,000 men and 100 CLAY BELT 37 officers on a front hardly longer than two or three city blocks, and the blood-drenched hillock had changed hands several times. Still the battle raged. The Germans delivered their first bom- bardment with poison gas shells, and in the first days of May discharged several poison gas waves which alone cost the British another thousand men and drove them from the position. In the fight for Mt. Sorel and Observatory Hill in June of the same year the Canadians sacrificed 7,000 men. 9 The Ypres Bastion If an insignificant hill may be of such vital importance, who shall set a value on such heights as the east-west ridge from Mont Kemmel to Mont Cassel, rising several hundred feet above the plain? Dominating the lowland from a central point, these "Monts de Flandre," as they are called, should form the key position of Flanders. It is perhaps not too much to accord them the dignity of such a title. In Roman times Mont Cassel was crowned with fortifications, from which radiated a system of Roman roads to bind the surrounding plain to the central strong- hold. The eleventh, fourteenth, and seventeenth centuries saw fierce battles rage about it. When Philip of Valois overthrew the Flemish rebels holding Mont Cassel in 1328, all Flanders submitted. While the French held southern Flanders against the combined forces of the Allies in the wars of the French Revolution, they made of Mont Cassel an entrenched camp, from which military operations were carried out in all directions. So throughout history one might trace the important r61e played by this ridge of hills and particularly by its culminating point, Mont Cassel, which was in 19 14 to serve as the observa- tory from which Foch would direct the first battles of Flanders and from which he could, in fact, on a clear day see the flashes of the guns from the dunes at Nieuport to the chalk upland at Vimy Ridge. But the real strength of the hill position is only apparent when taken in connection with that remnant of the backslope of the 'A.ConanDoyle: TheBritishCampaigninFranceandFlanders:i9l6,London,i9i7. 38 BATTLEFIELD OF FLANDERS former cuesta, or asymmetrical upland, which trends northward from near Mont Kenimel to the vicinity of Passchendaele (Fig. 2). This Messines-Passchendaele ridge is lower than the Mont Cassel-Mont Kemmel remnant of the southern crest of the cuesta, and, being part of the gentle back slope of the former upland, it gets progressively lower toward the north, where it is in a sense continued by fragments of the second cuesta near Staden so as to give a line of heights bending northwest nearly to Dix- mude. In front of this northern sector, and strengthening it, lies the marshy Handzaeme lowland, a natural moat be- fore the upland barrier. The Messines-Passchendaele crest is broader than the higher ridge, affording space for large bodies of troops; and it commands a wide stretch of the fiat plain to the east and west. Especially toward the east, the direction of chief danger, the crest dominates the vast plain like a watch- tower on a castle wall. In the angle between the east-west and north-south ridges lay Ypres, defended by these higher lands on the east and south, and on the west by several parallel branches of the Yser River and the Yser Canal. About Ypres there is, furthermore, a girdle of forested areas, partly on the arenaceous soils of the upper sand formation capping the higher portions of the ridges, partly on the compact clay exposures, or clyttes, which here protrude through the loam mantle in unusual num- bers and which, we have seen, are so inhospitable to agriculture that they remain wooded. Altogether the Ypres region is a mili- tary stronghold of the highest importance, and this importance is fully attested by the number of pages in history which record bloody struggles for its defense or conquest. Siege and assault have followed each other in a long succession of which the great Battles of Ypres of the World War were merely the most recent chapters. The Mont Cassel-Mont Kemmel ridge and the Messines- Passchendaele ridge meet in an apex near Mont Kemmel, form- ing a "bastion" pointing southeast, the flanks of which present a most formidable natural obstacle to an enemy advancing from either the south or the east. Neither to the north nor to the CLAY BELT 39 south of it could hostile troops press very far westward without incurring heavy risks, so long as the obstacle remained unshaken to threaten their flank and rear. On the south especially would a Fig. 14 — Block diagram of the Flanders and Somme battlefields, showing salient features of the topography. westward advance of necessity be limited, for the Vimy Ridge side of the Arras bastion (p. 189) and the Mont Cassel-Mont Kemmel wall of the Ypres bastion formed the two jaws of a giant trap which might at any time prove fatal to an army 4 o BATTLEFIELD OF FLANDERS pushing too far over the plain between them (Figs. 14 and 36). The Mont Kemmel-Mont Cassel ridge is continued westward by the hills east and west of Watten until it meets the northwestern prolongation of Vimy Ridge west of St. Omer. Between the two bastions there is thus a triangular re-entrant, or "curtain," of low plain hemmed in by higher land. Into this curtain an enemy could safely venture only if both bastions were conquered and danger from the higher land removed. We shall not be surprised, therefore, to find much of the fighting in the Flanders plain centering on the attempted destruction of the Ypres bastion, "the key of Flanders," nor to find the Germans, after pushing a salient across the plain between the Ypres and Arras bastions, voluntarily withdrawing from the dangerous position when repeated attempts to destroy the two jaws of the trap had failed. The hills are apt to be drier than the adjacent portions of the plain, either because they are capped by the upper sand forma- tion or because they rise higher above the permanent ground- water level and shed rainfall more readily from their sloping sides. Holders of the hills thus possess more comfortable positions, as well as better observation. But they are exposed to a new danger, for in the better drained ground subterranean mine war- fare can be prosecuted on a scale impossible in the water-soaked lowland. It is true that tunneling operations must encounter dangers from underground water in the hills, but where layers of sand alternate with layers of clay, engineers guided by geological advice may escape much of the trouble suffered by those who remain ignorant of the detailed structure of the terrain. Even in rainy Flanders there is a "dry" season, which causes the ground- water level to sink lower than usual; but, owing to a lag in its movement, it continues to sink for some time after the wet season has begun, and the rise does not cease until after the beginning of the dry season. Failure of the Germans to realize this fact re- sulted in the drowning of some of their tunnels driven just above the water level as determined when the heavier rains ceased. The chief geologist of the British forces, familiar with the subterranean movements of the water table, saved his army all trouble from NORTHERN SAND BELT 41 this source. The high importance of mining operations in Flan- ders, and the necessity of a knowledge of underground structure in directing them, were fully demonstrated in the Battle of Messines Ridge, which began with the greatest mining operations in the history of warfare. THE NORTHERN SAND BELT It is only by way of contrast with the clay belt that the north- ern zone may be called sandy. The land is so low and flat that water stands always on or near the surface just as in the clay country. The sands themselves are usually argillaceous, and layers of true clay alternate with them. So also the loam cover- ing contains a large clay content, even though more sandy than farther south. Hence the water-soaked soil is muddy, its humid- ity favors a vigorous growth of vegetation, and one who passes from the clay belt to the sand belt is not impressed with any sud- den or marked change in the character of the country. Seldom is there even a trace of that appearance of aridity which one usually associates with a sandy region. We have already seen (p. 12) that forested areas are more numerous in the sand belts and that there are remnants of two faint cuestas (three if we count a very short intervening one) indicated by the steeper southern and more gentle northern slopes of two slightly elevated zones. In a belt where the "ele- vations" are matters of a few feet only and where the "steeper slopes" are scarcely noticeable as such, the topography would be called "flat" by any one who did not give it a very careful exami- nation. Nothing but the fact that on a flat battlefield elevations of half a dozen feet may be significant makes it worth while to emphasize such faint topographic elements. It is in the northern sand belt that contamination of deeper waters is most likely to occur. This is due to the fact that the clay layers interspersed with the sands are not continuous over broad areas. Impure waters carried downward through the sand to the first clay bed move along its surface until its margin is reached, then descend to the next layer, and so on. The fact 42 BATTLEFIELD OF FLANDERS that a well is deep is therefore no guarantee that its water is fit for use. "At Bruges wells 40 meters deep are as dangerous as those which barely penetrate below the surface." 10 In general, however, there is little in the northern sand belt to differentiate it from that part of the Flanders battlefield just to the south. Certainly the points of resemblance are far more striking than the differences. It is therefore unnecessary to re- peat the previous descriptions of features which find their typical expression in the clay plain and which are only modified in a moderate degree in the northern sand belt. THE MARITIME BELT Low, flat, and wet as is the rest of Flanders, the maritime belt is lower, flatter, and wetter. Only an occasional area rises above the level of high tide, and vast stretches lie well below tide level. The dunes and the dikes alone prevent the belt from being almost completely submerged when the tide is high, at which times some areas would have 10 feet of water over them. The flat plain of the clay belt is undulating, almost hilly, when compared with the remarkable flatness of the low land bordering the sea. Trees are rare, houses few, and villages still less numer- ous; hence wide vistas open to the view. Yet the eye sweeps the broad expanse of fields and pastures in vain for an elevation sufficiently high to be detected by the unaided vision. Only careful measurement would reveal the fact that occasionally the surface swells gently a very few feet above the average level. Here one has an impression of vastness which is lacking in the tree-studded plain of the interior. On so low and so level a surface the problem of excess humidity is necessarily present in its most difficult form. With the whole belt menaced by invasion from the sea on one side, by flooding from the rivers which flow into the lowland on the other side, by deluges of rain from a leaden sky above, and by eruption of brack- ish water from the earth beneath; and with no chance for all these waters to escape with the aid of man's labor, it must be evi- 10 Blanchard, La Flandre. p. 56. MARITIME BELT 43 dent that only a severe and never-ending struggle can keep the land fit for human habitation. The line of dunes must be guarded and strengthened, and breaches in them blocked by arti- ficial dikes. When during the great siege of Ostend in 1601 the Dutch removed the dunes northeast of the town in order to pre- vent the Spaniards from attacking it on that side, the sea poured Fig. is — Drainage ditches and canals in the maritime belt of the Flanders battlefield. through the gap and flooded many square miles of the maritime plain. During the War of the Spanish Succession, when the French under Marshal Saxe were besieging the English base at Ostend, the commander-in-chief of the British forces in vain urged his Dutch and Austrian allies to open the dikes and flood the country as a measure of defense. In this case failure to profit by the natural topographic advantages of the region led to the fall of Ostend. In 1793 the French under Souham, besieged in Dun- kirk by the Allied armies, opened the sluices and deluged all the 44 BATTLEFIELD OF FLANDERS former marsh region east of the fortress, thereby contributing largely to the ultimate success of the defense. These are but isolated instances of the many occasions on which inundation of the low plain has been resorted to as a means of defense in war. The river waters must not be permitted to enter the plain but must be carried across it between dikes raised high enough to confine the waters until they reach the sea. Here gates must be installed and faithful watchmen see that they are closed at high tide to keep the sea from flowing into the land but opened at low tide to permit the accumulated river waters to escape. If the gates are kept closed, or are opened at high tide, the land may be flooded, especially if sluice gates in the river dikes are also opened. On the other hand, evacuation of the canals may be resorted to in order to deprive an enemy of transport at a critical moment. This was demonstrated during the siege of Dunkirk in 1793, when the defeated Allies on retiring had to abandon their thirty-two heavy guns to the French because the latter, "having control of the sluices, had shut off the water from the canal, so that it was no longer of use for transport." 11 Disposal of rain falling directly on the flat surface is a heavy task. It cannot sink into the water-soaked ground; evaporation is slight in the humid region; and the neighboring rivers are at higher levels. An intricate network of drainage ditches must be dug to lead the water to collecting points, whence it may be raised by windmills or steam pumps to the diked rivers. In the lowest part of the plain the complex of ditches and canals (Fig. 15) is particularly striking. Even detailed maps show but a fraction of the total number of ditches, for they are sometimes spaced only 30 or 40 yards apart. They break the country up into tiny morsels of ground bounded by water and oppose to an army a formidable obstacle. The uselessness of cavalry on such a terrain has often been commented upon in past campaigns; the artillery finds itself embarrassed at every turn; while that heavy modern arm of the service, the tanks, can scarcely negotiate the 11 Fortescue, Vol. 4, Part I, p. 132. MARITIME BELT 45 Fig. 16 — Passerelle across the flood waters at the border of the submerged zone along the Yser. Many of the advanced positions could be reached only by means of these narrow footbridges. (Belgian official photograph.) Fig. 17 — Sandbag shelters and defenses commanding the flooded valley of the Yser. Note the wooden footpath over the muddy surface and the fact that the water-soaked soil compels resort to exposed sandbag shelters in place of concealed underground dugouts. (Belgian official photograph.) 4 6 BATTLEFIELD OF FLANDERS larger channels. Even infantry finds such terrain difficult to cross, as will readily appear from perusing accounts of the sieges of Dunkirk. This position is protected on the east by a former marsh, the Moeres, or Great Moor, occupying the lowest part of the maritime plain, which was drained by a system of rectangular ditches in 1624, but partially flooded during the sieges. Fig. 18 — The water-soaked plain of Flanders between Ypres and Dixmude, show- ing the morass of mud, shell craters, and fallen tree branches produced by artillery fire on a tree-covered part of the low surface. (Belgian official photograph.) One naturally expects in so low and humid a region to find un- sanitary conditions and sickness among the difficulties to be combated by large military forces. Such great marshes as the Moeres near Dunkirk did indeed curse all the surrounding plain with malarial fevers, and even today the evil is not completely eliminated. The clairs, or openings, left by the extraction of great quantities of peat when that labor was the chief industry of the plain, contain large bodies of stagnant water, and not all the marshes have been drained. The water near the surface is con- taminated with all sorts of impurities; that deeper down is brackish. "All the wells are bad, or ought to be so considered." 12 12 Blanchard, La Flandre, p. 292. MARITIME BELT 47 The brackish waters, which penetrate through the sands below the maritime plain under pressure from the higher sea, not only destroy the value of wells but may even push to the sur- face, especially near the dunes, and kill trees and gardens. Dur- ing the sieges of Dunkirk the Allied armies found the region "wholly destitute of drinking water, that in the canals being brackish, and that found in the wells unpalatable;" and in 1793 it was the combination of "incessant fighting, a swampy en- campment, bad drinking water, and fever" which brought the besieging armies to the verge of disaster. 13 That man should have so long maintained the struggle against his liquid foes is due in part to the fertility of the loam which covers the maritime plain. The low-lying reclaimed polders have the richest soil in all Flanders, and despite its clayey consistency it rewarded the toiler with abundant harvests. Its natural rich- ness is soon reduced, but wise methods of agriculture have main- tained the productiveness of the plain. Fields and pastures together are able to support a population which, if not so dense, is more prosperous than that of neighboring belts. The scarcity of trees is explained not through any defect of the soil ; nor, as in the case of the Somme plain, by the fact that the fertility of the soil caused all the land to be cleared and cultivated. It seems due rather to the greater exposure of the plain to the winds, which, sweeping in from the open ocean, bend and deform such trees as persist in growing on the level expanse. As in the rest of Flanders, so in the maritime plain the clay is a most important element to be reckoned with. The layer of peat which underlies almost the entire region, but shows at the surface over limited areas, is usually covered with the argillaceous loam. Here also, therefore, a dense, sticky, slippery mud is charac- teristic of the water-soaked land and opposes to military opera- tions all those embarrassments described on previous pages. The awful mud of the Yser is the strongest memory which one carries away from that part of the battle front. Even the passer elles above the flood waters (Fig. 16) were coated with the ls Fortescue, Vol. 4, Part I, pp. 124, 132. 48 BATTLEFIELD OF FLANDERS slimy deposit tracked there by countless feet, and one slipped and slid, continually in imminent peril of gliding into the waters. Un- drained marshes still exist, and mud, water, and marsh, especially when torn into a frightful morass by shell fire, produced a surface horrible to contemplate (Fig. 18). Deep trenches and dugouts were of necessity little known on the lower parts of this front. Those which existed were often filled with water and liquid mud, and life in them was a night- mare. Breastworks and shelters of sandbags were brought into use (Fig. 17), and every faint knoll rising above the general level became a fortified island in the sea of mud. Railway embank- ments and dikes were much-prized strips of higher and drier land, often used as principal lines of defense (Fig. 17). But it was the flood waters, released from canals and diked rivers and permitted to flow in from the ocean, which formed the ultimate line of resistance to an enemy attack. It is true that even an empty canal, with slimy sides and mire in its bottom, was no easy obstacle to cross (Fig. 19.) Yet this was as nothing compared with the vast stretch of waters which filled the so-called valley of the Yser when the gates at Nieuport were opened and the whole country from the coast to Dixmude, formerly an estuary of the sea, reverted to its original state. The line of the Yser, repeatedly utilized as a defensive barrier in the wars of the past in Flanders, was to play a grand role in the World War. THE DUNE BELT Between the low maritime plain and the sea stretches the great barrier of sand, from one to several miles broad, capped by dunes which keep out the ocean waters. The dunes are from 30 to 100 feet high and, while sheltering more trees than the level plain, show large spaces of barren white or yellowish sand, con- spicuous even at great distances (Fig. 20). The sandy soil is much drier than that found on the plain; but the water lies close below the surface, as the Duke of York discovered when he found it impossible, during his siege of Dunkirk, to protect his flanks by trenches because they encountered water at a depth of two DUNE BELT 49 feet. 14 In form the individual dunes are extremely irregular but often arranged in two or more bands parallel to the sea and enclosing broad depressions or basins called pannes, occasionally containing ponds of fresh water. The higher parts of the dunes permit the digging of dry trenches, which must, however, be specially walled to prevent the slumping of the loose sand (Fig. Fig. 19 — Defensive position sited along a canal draining the low Flanders plain. Even when the water in the canal is low, its muddy bottom and marshy borders make it a moat of some tactical value. (Belgian official photograph.) 21). Shells bursting in dune sands do less damage than in almost any other type of terrain. Because the dunes are higher and less humid than the plain, they contain an unusually dense population. They are out of the reach of floods and at all seasons of the year serve as a highway for traffic along the coast in preference to the low and muddy plain. Both the flat strand of the beach between the dunes and the sea, and the depressions between the principal dune ridges, parallel 14 Fortescue, Vol. 4, Part I, p. 128. 50 BATTLEFIELD OF FLANDERS to the sea, have been used as routes of travel and for the advance of armies. The dunes contain the least objectionable waters for drinking purposes, as the rain falling on them and sinking into the loose sand forms a water table highest in the center and declining toward the sea on the outside and toward the polders on the inside. Wells a few feet deep get a limited supply fit for household uses. Fig. 20 — Barbed wire defenses to prevent an enemy advance along the dry high- way formed by the strip of dunes along the coast. Because of the water-soaked character of the maritime plain, the dunes have always been a favorite route for advancing armies. (Belgian official photograph.) Higher, drier, better provided with drinking water, and afford- ing some shelter from an enemy, the dunes are far better adapted to military operations than is the adjacent plain with its ditches and canals, its water and its mud. Before the modern day of military operations on a grand scale one could truly say: "When an army ventures into the maritime plain, it is not on the vast open fields that it delivers battle; it is upon the sands that it encounters the adversary. The rare combats which have taken place in the plain have had for their theater the sands : the Battle DUNE BELT 51 of the Dunes of Dunkirk, the Battle of the Dunes of Nieuport. It is by the dunes that the Duke of Guise lays hand upon Calais; by the dunes that Conde, Turenne take Dunkirk, that the Duke of York approaches it in 1793, that the Archduke Albert persists in the attack on Ostend." 15 In the great Battle of the Dunes of Dunkirk, in 1658, the Spanish took up a strong position in the sand Fig. 21 — Extension of the defense lines from the coastal dunes across the sandy beach into the sea, to block the natural pathway formed by the strand. Like the strip of dunes, the strand has in successive wars served as an easy line of advance for hostile troops. (Belgian official photograph.) hills, with their right flank protected by the sea, their left by a canal in the plain, and the key to their position held by four regiments on a particularly high mass of dunes. So formidable was the position and so well protected its flanks that the com- mander of the English troops for a time despaired of forcing it. Success was achieved only when the dominating point on the highest dune was stormed at a heavy sacrifice. The line of the dunes is occasionally broken through by rivers 15 Blanchard, La Flandre, p. 226. 52 BATTLEFIELD OF FLANDERS or canals, the mouths of which have been transformed into har- bors with the aid of artificial locks, basins, and other engineering works. Such are the harbors of Calais, Dunkirk, Nieuport, Ostend, and Zeebrugge. All are exposed to the danger of silting up; but the outwash from rivers and canals, sometimes aided by tidewater accumulated in special basins for the purpose, usually serves to keep them open. Any obstacle placed in the entrance by checking the outflowing currents will speedily cause the blocking of the port, a fact of which the British took advantage in their brilliant exploit in closing the German submarine base at Zee- brugge. In front of the harbors the bottom of the sea is ridged with submarine banks of sand parallel to the shore which break the force of storm waves, while between the banks and the coast are channels, often 30 feet or more deep, where vessels may lie in comparative security until a favorable moment for entering port. Opposite Nieuport these banks reach their maximum development and so afford protection to naval craft which might desire to protect the flank of an army standing behind the line of the Yser from possible attack by enemy forces advancing along the dunes. We have now sketched in outline the salient features of each topographic belt in the Flanders battlefield. In all of them we have discovered that the points of resemblance are more striking than the points of difference. There is thus a real geographic unity in the Flanders plain, which the several phases of its topography do not materially affect. Everywhere the plain is low, everywhere it is comparatively flat, everywhere except on the limited hill and dune areas it is a sea of sticky mud in wet weather. In all parts are found sluggish rivers and canals which may be used in time of war as defense lines by opening the sluice gates and inundating the broad, shallow valleys. Few parts do not have the terrain cut up by countless ditches filled with water and the ground permanently saturated below a very shallow depth. Everywhere the water-supply problem is a serious one in spite of the extreme humidity of the country. The Battlefield of Flanders has its own consistent and striking character and is like no other battlefield of the World War. CHAPTER II MILITARY OPERATIONS ON THE BATTLEFIELD OF FLANDERS About the middle of October, 1914, the Allied forces, following the Battle of the Marne and the check at the Aisne, were pro- longing their great flanking maneuver, sometimes called "the race for the sea," northward from the Somme region into the low country south of Ypres. At the same time the Belgian army, retreating from Antwerp, was falling back toward the line of the Yser. Thus from opposite directions were in progress the move- ments which would soon crystallize the battle front across the Flanders plain. On October 11 British cavalry entered the southern edge of the Battlefield of Flanders, crossing the southern sand belt and driving the enemy cavalry from the Forest of Nieppe south of Hazebrouck, one of the woodland areas characteristic of the southern margin of the clay belt and the sand belt. Pushing on across the plain to the Mont Cassel-Mont Kemmel ridge, it drove an outpost of the German army from the Mont des Cats (Fig. 4) after a stiff fight and captured the high ground farther east. Thus was the securing of the Ypres bastion one of the first objective of Allied strategy in this field. With a view to a further advance eastward Sir John French "ordered General Allenby, on the 15th, to reconnoitre the line of the River Lys, and endeavour to secure the passages on the opposite bank." But the barrier was so strongly held by the Germans to the northeast of a point near Armentieres that "the Cavalry corps was unable to secure pas- sages or to establish a permanent footing on the eastern bank of the river." 1 The infantry and artillery on entering the plain swung east- ward to support the left wing of the French, which was already x Sir John French: Despatches: Mons; The Marne; The Aisne; Flanders, London, 1914, pp. 119, 128. 54 BATTLEFIELD OF FLANDERS engaged with the Germans southwest of Lille. Immediately Sir John French found himself embarrassed by the nature of the terrain, and in his despatch covering the campaign he early notes that "the ground throughout this country is remarkably fiat, rendering effective artillery support very difficult." 2 Mean- while both British and French made fruitless attempts to dis- lodge the Germans from their strong position at La Bassee, where they had fortified themselves on a slightly higher bit of dry land surrounded by low, flat, marshy ground partly underlain by peat bog. "This position of La Bassee has throughout the battle defied all attempts at capture either by the French or the Brit- ish," reports the Commander-in-Chief of the British forces. Elsewhere the Allies pushed on, much handicapped by the ditches and dikes and unable to reach Lille. Meanwhile General Rawlinson's forces, which had been aiding the Belgians in their retreat from Antwerp, fell back to the Ypres bastion, taking up a position on the Messines-Passchendaele ridge and facing east in the neighborhood of Gheluvelt, while French reserves were massed inside the bastion, in and west of the town of Ypres. Still anxious to obtain a footing on the east bank of the Lys the Commander-in-Chief directed that Rawlin- son's forces should advance eastward from the ridge to the river at Menin and force a passage over the barrier there. But Raw- linson feared to abandon his advantageous position on the ridge, in view of the fact that overwhelming enemy forces were threaten- ing him from the east and northeast. He accordingly made representations to this effect and after a short advance returned to his lines on the eastern side of the Ypres bastion, while French cavalry on his left drove back advanced parties of the enemy beyond the Forest of Houthulst in order to secure the north- western continuation of the ridge toward Dixmude. They were soon pressed back from these northern hills to the line of the Yser River, but most of the Ypres bastion was still in Allied hands, while the enemy held the important line of the Lys just to the southeast, along which he was assembling very heavy forces. 2 Sir John French, Despatches, p. 121. BATTLE OF THE YSER 55 During these movements in the south and center of the Flan- ders plain, the Belgian army, supported by French forces, was falling back from Antwerp. King Albert, "judging that no other line offered as great advantages, decided to establish the army on the Yser and to place this line in a state of defense." 3 On October 15 the Belgians and French took up the new position along the west bank of the Yser from Zuydschoote, five miles north of Ypres, past Dixmude to the sea at Nieuport, leaving out- posts on the eastern bank. The Allied front was now based on the river-canal barrier from the sea to a point south of Dixmude, thence on the ridge positions from near Passchendaele southward beyond Gheluvelt, whence it cut across the plain to follow south- west up the western side of the Lys nearly to Armentieres. From here it ran across the plain past La Bassee to the edge of the chalk upland of Artois. In the north this position was a strong one, but it was held by wholly inadequate forces in the face of an en- emy already greatly superior in numbers and constantly growing stronger. The Battle of the Yser The moment had now arrived when the German armies, con- fronted by a continuous line of enemies from Switzerland to the sea, must give up any hope of outmaneuvering the Allied line by a turning movement around its left end and launch a direct attack against some part of the front. The sector selected was the plain of Flanders, and the two immediate objectives were to force the Yser barrier and smash the Ypres bastion. If these two for- midable natural obstacles could be conquered, the Channel ports might be seized, direct communication between England and the Continent endangered, and the left wing of the Allies rolled up or forced to fall back on the Artois upland. In the latter case the Allied line would form a dangerous strategic salient with its apex near Arras, which, broken at any point by a German offensive, would compel an Allied withdrawal to the line of the Somme. 3 Military Operations of Belgium in Defence of the Country and To Uphold Her Neutrality: Report Compiled by the Commander-in-Chief of the Belgian Army for the Period July 31st to December 31st, 1914, London, 1915, p. 65. 56 BATTLEFIELD OF FLANDERS Thus the German line would be greatly shortened and the fertile fields of Picardy added to the German holdings. The way for an advance on Paris would once more lie open. The first attack was directed against the line of the Yser from Dixmude to the sea at Nieuport, and both Belgians and French Fig. 22 — The Yser barrier, showing the three main lines of resistance: the Yser River, the Noord Vaart-Groote Bever- dyk stream, and the Dixmude-Nieuport railway embank- ment. All the area from the railway embankment to a line east of the Yser River was ultimately flooded. «quickly concentrated the bulk of their available forces behind this part of the barrier to meet the shock. A report by the Com- mander-in-Chief of the Belgian army on the "Military Operations of Belgium" describes the Yser line as "an excellent defensive position" and "tactically a strong one." The left flank rested on the sea, which was in friendly hands and where the submerged banks off Nieuport offered fair shelter to a fleet engaged in sup- porting land operations by its fire. The river itself was about 65 BATTLE OF THE YSER 57 feet wide, diked on both sides, and passable only by bridge or boat. Fortunately the western bank commanded the eastern one, being about 6 feet higher. Roughly parallel to the river and just west of it runs another stream about 30 feet wide, called the Groote Beverdyk in its upper portion, the Noord Vaart farther down near where it joins the Yser at Nieuport (Fig. 22). This stream forms a good second line of resistance. Just west of it is a third line, the embankment of the Dixmude-Nieuport rail- way, raised from 3 to 6 feet above the wet plain. All this region, a gulf only a few centuries ago as we have already seen, was low, marshy, and muddy, intersected with a network of ditches filled with water, largely devoid of cover for attacking forces, and subject to inundation at any time (Fig. 15). The principal point of weakness in the line was the large meander of the river near Tervaete, half way between Dixmude and the sea, called "the Tervaete bend." As this meander is convex toward the north, an enemy on the northern side can concentrate fire from three directions upon the defenders within, render it unten- able, and then effect a crossing. 4 At its two ends the line was buttressed by strong positions. Nieuport is a center upon which six canals or streams converge, all of them passing east or north of the town, while the dunes offer cover for troops holding a bridgehead north of the watercourses. This bridgehead was of vital importance, for at Nieuport were located the gates which would admit the sea to inundate the Yser valley, should this become necessary as a means of defense. The security of this important buttress was further guaranteed by Allied control of the sea. Dixmude, at the other end of the line, was defended by strong positions in its suburbs, protected by the partly submerged marshy bottom of the Handzaeme low- land on the east, by the Handzaeme canal, and by a railway embankment on the east and south. Altogether the line was one of formidable strength. Never before had an enemy attempted an invasion by this route. Ludendorff chides the Allied commanders for their 4 Military Operations of Belgium, p. 72. 58 BATTLEFIELD OF FLANDERS stupidity in seeking to attain great strategic ends which tactical conditions rendered impossible of achievement. The best example which the war affords of such unwisdom is the attempt of the German high command to gain the grandiose strategic ends outlined above by an assault on the impregnable natural barrier of the Yser. Such a blunder can be excused, if at all, only on the ground that the German leaders believed their overwhelming numerical superiority was more than a match for any natural obstacle, however strong, when defended by nothing more than a wretched remnant of the little Belgian army and a few French forces hastily rushed to their assistance. On October 16 the Battle of the Yser began with a recon- naisance in force against Dixmude. By the 18th Nieuport and parts of the river front between the two buttresses were feeling the enemy pressure; a British flotilla, supported by French ships, was shelling the Germans attacking along the historic highway, and "dryway," of the dunes; and the battle was becoming general. One determined assault after another, launched against the Nieuport bridgehead, was beaten off with heavy losses to the enemy. Violent attacks against the Dixmude bridgehead, some of them delivered in the darkness of night, nearly achieved suc- cess at one moment, but the lost positions were recovered. A furious bombardment deluged the whole front and the Belgian back areas with shells, while heavy assaults on the Tervaete bend, fully recognized by the Germans as the weakest point along the barrier, menaced the Belgian center. By October 21 seven German divisions were flinging their whole weight furiously against the line of the Yser. That night, under cover of dark- ness, a crossing was effected on a temporary bridge at the danger- ous Tervaete bend, and, despite the efforts of the Belgians to execute the order to hold the chord drawn across the base of the meander "at all costs by clinging to every inch of the ground," the condition was so grave there on October 23 that the Belgian command appealed to the French for support, at the same time again ordering the Belgian center "to hold out to the last extrem- ity." Fortunately, every attempt made by the Germans to cross BATTLE OF THE YSER 59 the barrier elsewhere had proved a costly failure. On the other hand, an Allied offensive by troops debouching from behind the barrier had to be abandoned in order to help the imperiled Bel- gian center. The German commanders now concentrated all their efforts against the weak center at Tervaete bend and drove the defenders from their position along the chord. The latter thereupon fell back behind the line of the Groote Beverdyk. This exposed the right flank of those defending the river farther west, and they fell back behind the lower course of the same stream, where it is called the Noord Vaart (Fig. 22). Only the Dixmude end of the main river position still held, and the enemy now made a supreme effort to capture that part of the line. Fifteen assaults in close succession were hurled back by the gallant Belgians and the heroic French marine fusiliers, and still the enemy returned to the onslaught with dogged perseverance. 5 The defenders, hopelessly outnumbered and reduced to a state of extreme exhaustion, lacking reserves, and faced with a shortage of munitions, now required all the aid that the natural advan- tages of their position could give them. "The key of the position was not at Dixmude, nor at Pervyse, nor at Ramscappelle, nor at Ypres; but in the pocket of the chief hydrographic officer who guarded the locks at Nieuport." 6 A council of war on October 25 decided to flood all the Yser region east of the railway embankment forming the third line of resistance. Accordingly all aqueducts through the embank- ment were sealed up, and the sluices at Nieuport were opened when the tide was high in order to let in the sea, and closed when the tide was low to prevent the accumulating river, rain, and marine waters from escaping. It was none too early, for on the 26th the Germans in irresistible numbers drove the Allies from behind the Beverdyk, compelling them to seek protection behind the railway embankment. But the terrible strain of continual assaults through mire and water, over rivers, canals, and ditches, 5 Military Operations of Belgium, p. 81. 6 Charles Le Gome: Dixmude, Rev. des Deux Monies, Vol. 26, 1915, pp. 169- 194, 370-403; reference on p. 390. 60 BATTLEFIELD OF FLANDERS was beginning to tell on the invaders. As they paused for breath, the new ally of the Allies made its appearance. About the 28th the Germans discovered that water was rising in the ditches and beginning to flow over the flat land. The peril was imminent, and on the 29th they began a new series of furious attacks, hoping to break through before it should be forever too late. Their only hope lay in capturing the Nieuport buttress, containing the engineering works by which alone the floods could be controlled. But this strong position, supported by the Allied flotilla off the coast, defied every assault. Just to the east the gray- clad invaders, struggling through the mire and the rising waters, "hideous, drenched to the waist, and smeared with mud to the tops of their heads," stormed the railway embankment and entered Ramscappelle. They maintained themselves in the village for a brief space but were soon thrown back. Elsewhere the line held. The waters continued to rise. Soon from the Dixmude-Nieu- port railway embankment the Allied defenders looked out across a flood two or three miles wide, covered with debris and swollen German corpses (Fig. 17), which effectively blocked the Kaiser's frantic drive toward the Channel ports. The Teutonic hordes were faced by a liquid trap which yawned to engulf them. Beneath the muddy waters were hidden ditches and canals into which men would suddenly plunge over their heads, and bottom- less mud which would hold them fast in the flood. The level surface of the liquid could offer no shelter from the fire of sharp- shooter and machine gunner, and those who fell wounded must drown where they dropped. Numerical superiority and Prussian discipline were alike helpless before such a barrier. The Germans retired northward beyond the flood, abandoning quantities of arms and ammunition. And while they might later seize the ruins of Dixmude north of the river, the barrier itself was abso- lutely impregnable. The Battle of the Yser was over. The First Battle of Ypres While the Battle of the Yser was still in progress the Germans were preparing the blow designed to shatter the Ypres bastion. FIRST BATTLE OF YPRES 61 The Allied command, judging that the best defense would be an offensive movement eastward and northeastward from the bastion, and seeing in the situation along the Yser an added reason for attacking the German forces opposed to them, on October 21 ordered an advance northeastward against the northern part of the Messines-Passchendaele ridge, still held by the Germans. After taking the ridge the offensive would be developed as far as possible in the direction of Roulers and Thourout, thereby flanking the Germans attacking the line of the Yser, with Bruges and Ghent as more remote objectives. A little ground was gained, but the stiff German resistance, the growing menace of increasing German numbers, and the necessity of awaiting Allied reinforcements made further attacks against the northern end of the ridge inadvisable. The troops were then ordered to strengthen their positions along the ridge east and southeast of Ypres from Zonnebeke to Messines and to hold fast until a new French army hurrying northward should enter into the line. On October 22 and 23 the Germans attacked with great deter- mination, gaining some ground but suffering heavy losses. The Allies counterattacked with equal energy and tried on the 23rd to continue their offensive, but were held up at several points, their center in particular being unable to cross the small brook flowing from near Passchendaele down the west side of the ridge. For nearly a week the Allied forces endeavored without much success to drive the enemy from the northern part of the ridge about Passchendaele. On the 29th the Germans, exhorted by their leaders to strike "the decisive blow," returned to the assault with redoubled energy. They assembled enormous forces opposite the three- mile sector of the ridge between Gheluvelt and Hollebeke and for two days hurled them against the barrier with almost irresist- ible fury. On either side the main attack was supported by other troops. At a terrible cost they won the crest in the Ghelu- velt region and at several points farther south. The moment was critical in the extreme. But additional French forces were 62 BATTLEFIELD OF FLANDERS thrown into the fight; the anxious leaders were encouraged by the imperturbable optimism of the genius now directing the battle from the highest point in Flanders as later he would direct the whole gigantic struggle from the highest position of military authority; and the men responded magnificently to the demands of their leaders. The enemy was hurled back down the eastern slope, and the Kaiser, now on the front, awaited in vain the moment for entering Ypres. Furious counterattacks again gave the Germans a foothold on the crest, but the bulk of the territory previously held by the Allies remained in Allied possession. On November 5 the Kaiser returned to Germany. The First Battle of Ypres was in reality a bataille de rencontre. Two offensives, hurled against each other, both came to grief. If the Germans were unable to secure full possession of that part of the Messines-Passchendaele ridge southeast of Ypres, the Allies failed equally to gain the portion northeast of the town. The Ypres bastion was not securely in the hands of either com- batant, and a resumption of the struggle for mastery of this critically important terrain could not long be delayed. For the Allies the tactical situation was anything but satisfactory, be- cause their line described an awkward salient about Ypres, per- mitting a concentration of enemy fire upon that junction point of all important roads, railways, and canals in the vicinity. German possession of part of the ridge seriously complicated the situation and called for an opening out of the salient by pushing the enemy northeastward and southeastward beyond the heights and into the plain below. For the Germans there was no hope of victory in Flanders, now that passage of the Yser barrier was definitely excluded, until their hold on the Ypres bastion had been extended into conquest of the entire obstacle. Both combatants prepared an immediate renewal of the struggle. A continuous and violent bombardment by the Ger- man artillery and the massing of new forces behind their front made their intentions clear. On the Allied side orders were given November 6 for an offensive which should clear the enemy from his footholds on the bastion northeast and southeast of Ypres FIRST BATTLE OF YPRES 63 and open out the salient by advancing its two sides. But every attempt at progress encountered violent resistance. On Novem- ber 9 the attacks and counterattacks became more violent along the whole line from Dixmude to beyond Messines, and on the 10th the Germans launched their supreme assault. A division of the Guard, secretly brought to Flanders and advised by the Kaiser that he counted on them to succeed where their comrades had failed, was hurled against the Allied front. Along the Yser Canal from Dixmude southward, now flooded by the waters backing up from the locks at Nieuport, a subsidi- ary attack gave to the enemy that part of Dixmude east of the barrier and secured him a precarious foothold on the west bank at the Maison du Passeur, halfway between Dixmude and Ypres. Elsewhere the barrier held firm. But the grand assault was delivered along the Ypres bastion farther southeast. It made progress, but only at terrible cost. "The regiments of the Guard . . . had been, north of Gheluvelt and between Zonnebeke and Passchendaele, so badly used up that they were gasping. Elsewhere certain regiments . . . had been, in the region of Poelcappelle, almost completely annihilated. 'On the 10th,' wrote one of the soldiers, 'we launched an assault in which almost the entire batallion was wiped out. In my company, in one hour, all fell except one officer and fifty men.'" 7 Even at such a price the gains were very moderate. The Messines end of the Messines- Passchendaele ridge was made securely German for the time, while the northern, or Passchendaele, end they still firmly held. In the center it had proved impossible to dislodge the Allies from their hold on the crest about Gheluvelt. Capture of a small part of the difficult terrain had exhausted the enemy's offensive power. This second phase of the battle, sometimes called the "Second Battle of Ypres," left the fate of the bastion still undecided. The Allies held the southern wall from Mont Kemmel westward and a central section of the eastern wall; while the Germans were 7 Louis Madelin: La Bataille des Flandres, Rev. des Deux Monies, Vol. 40, 1917, pp. 241-276, 506-539; reference on p. 530. 64 BATTLEFIELD OF FLANDERS firmly established along the two ends of the eastern wall. Neither offensive had succeeded. Ypres was exposed to direct enemy observation from the Messines-Passchendaele ridge and suffered from accurate artillery fire. The salient, instead of being widened, had been rendered more cramped and dangerous than ever. On the other hand, the German effort to crush the obstacle had failed, the left wing of the Allies was still secure, and the Channel ports seemed farther away than ever. Neither side could succeed in any large operations in Flanders until the Ypres bastion was wholly conquered. Certainly the struggle for mastery was merely ad- journed. The Second Battle of Ypres During the winter of 1914-1915 there were many local com- bats around the Ypres bastion, in the course of which the Allies gained a little terrain in the direction of Passchendaele and else- where but lost more of the ridge crest near Gheluvelt and east of Mont Kemmel. It was on the morning of April 23, 1915, that the French and Canadian troops holding the naturally weak seg- ment of the front connecting the Yser River-Canal barrier near Lizerne* with the Messines-Passchendaele ridge barrier near Zonnebeke, saw a greenish-yellow cloud rolling toward them across the flat plain. In order to flank the Allies from their last hold on the eastern wall of the Ypres bastion, the Germans were resorting to a new and barbarous offensive weapon, poison gas. Completely taken by surprise, strangled by an enemy they could not combat, the Allies retreated in disorder. On the northwest the Germans succeeded in crossing the canal in the confusion and established bridgeheads on the west bank at Lizerne and neigh- boring points. The northern side of the salient was pushed a couple of miles nearer to Ypres, and the crest of the ridge cleared of Allied troops. Nothing but the most heroic action on the part of the defenders of the bastion prevented a complete break in their front and the loss of the entire stronghold. Fortunately the rush was stopped in front of Ypres; long and bitter fighting ejected the Germans from the west bank of the canal, and the * Five miles north of Ypres. Not shown on PI. I. LOCAL COMBATS 65 Yser barrier remained intact. But the east wall of the bastion was in German hands, and the defenders of Ypres were subjected to all the tortures of an inferno as accurately controlled artillery fire poured upon them from every part of the salient. Local Combats In the southern part of the Flanders plain La Bassee had con- tinued to prove a strong point which held firm despite repeated local fluctuations of some magnitude north and south of it. To the north the British in March introduced, at Neuve Chapelle, the system of massed artillery fire on a limited front; but, despite its success in destroying the Germans' front trenches, only a small advance could be driven across the level plain where the attackers were fully exposed to a deadly fire. The British lost over 10,000 men on a narrow front in a few hours. During the first battle of Vimy Ridge in May the British again lost 8,000 men on the same terrain. South of La Bassee the British co- operated in the second battle of Vimy Ridge (September) by undertaking an advance on Lens across the more rolling plain of the Gohelle, in the transition belt. Loos was captured, and the low but important Hill 70, dominating Lens on the north, was seized but could not be held. The German positions in the chalk were strong, the British organization for supporting and relieving attacking troops on the badly exposed terrain was very defective, and the losses in a few hours were so heavy that the operation was regarded as a disaster. Not until command of the plain had been secured by full Allied control of Vimy Ridge could operations on the plain below be carried on to advantage. The battles for Vimy Ridge and the Arras bastion are treated in a later chapter. Throughout the rest of 19 15 and all of 1916 the battle line in Flanders remained practically stationary. There were local struggles at various points on the front as each side endeavored to secure some topographic ad vantage, some point of better obser- vation, in preparation for larger operations in the future. The war on the surface was now accompanied by the war of sub- terranean mines on a large scale. So valuable were hill positions 66 BATTLEFIELD OF FLANDERS on the level plain, and so difficult to take by surface operations alone, that extensive tunnels were run under the hills and the summits, in some cases, literally blown off. In the confusion the hill might be seized by the attacking forces, the lines of the former holders deluged with artillery fire accurately directed from the point of vantage, enemy movements in the plain so well observed as to make surprise counterattacks difficult or impos- sible, while the new possessors of the elevation would enjoy comparative immunity from direct observation in their back areas. It was during the first half of 1916 that the bloody struggles for individual hills, mentioned on earlier pages (pp. 35- 37), took place. Many of these local operations were hampered or defeated by the water and mud of the Flanders plain. Sir Douglas Haig's despatches covering this period abound with references to the extraordinary difficulties. During the fighting for the Bluff "heavy rain turned the ground into a quagmire so that progress was difficult for the attacking force." At St. Eloi "the work of consolidating our new position . . . proved extremely diffi- cult, owing to the wet soil, heavy shelling, and mine explo- sions; though pumps were brought up and efforts at draining were instituted, the result achieved was comparatively small. By dint of much heavy work the brigade holding these trenches . . . succeeded in reducing the water in the trenches by two feet by the morning of the 5th. This state of affairs could not, even so, be regarded as satisfactory; and during the 5th the enemy's bombardment increased in intensity, and the new trenches practically ceased to exist." 8 As one soldier is reported to have said, when told to "consolidate" his position: "It is im- possible to consolidate porridge." Men were swallowed up in the mire and suffocated, while rifles became so caked with mud that firing was impossible. Bayonets and bombs alone could be used in some of the assaults. 9 The fields became bogs and the trenches canals. Little progress could be made in such a terrain. 8 Sir Douglas Haig's Despatches, London, ioio, pp. 6, 8. 9 A. Conan Doyle: The British Campaign in France and Flanders: 1916, London, 1917. THIRD BATTLE OF YPRES 67 The Third Battle of Ypres (Battle of Messines Ridge) In the meantime the British were at one place carrying on mining operations on a scale never before attempted in warfare. The Ypres bastion was again to become the center of a violent struggle, in which the British would seek to dislodge the enemy from its eastern wall, the Messines-Passchendaele ridge. The positions held by us in Ypres salient since May, 191 5, were far from satisfactory [writes Sir Douglas Haig]. They were completely overlooked by the enemy. . . . They were certain to be costly to maintain against a serious attack, in which the enemy would enjoy all the advantages in observation and in the placing of his artillery. Our positions would be much improved by the capture of the Messines- Wytschaete Ridge*, and of the high ground which extends thence north- eastwards forsomeseven miles and then trends north through Broodseinde and Passchendaele. . . . The village of Messines, situated on the southern spur of the ridge, commands a wide view of the valley of the Lys, and enfiladed the British lines to the south. Northwest of Messines the village of Wytschaete, situated ... on the highest part of the ridge, from its height of about 260 feet commands even more completely the town of Ypres and the whole of the old British positions in the Ypres salient. . . . The natural advantages of the pos.tion were excep- tional, and during more than two years of occupation the enemy had de- voted the greatest skill and industry to developing them to the utmost. 10 Ludendorff emphasizes the importance of the ridge to the Germans, referring particularly to the value of direct observa- tion of enemy positions from the ground and of having his own back areas shielded from the view of the enemy. 11 But experience had shown that any operation against the formidable obstacles of the ridge must, in order to be successful, be conducted on an elaborate scale, with large forces, and only after the most careful preliminary preparation. Accordingly early in 1916, nearly a year and a half before the blow was launched, the British began the building of the network of roads * Southern part of the Messines-Passchendaele ridge. 10 Haig, pp. 82, 105. 11 Erich von Ludendorff: Ludendorff's Own Story, August, 1914-November, 1918: The Great War from the Siege of Liege to the Signing of the Armistice. As Viewed from the Grand Headquarters of the German Army, 2 vols., New York, 1919; reference in Vol. 2, p. 101. 68 BATTLEFIELD OF FLANDERS and railways necessary to move troops, guns, munitions, and other supplies in enormous quantities with great speed, and the excavating of a series of tunnels under the ridge in order to blow the Germans out of their strong position by a series of mine explosions. The work of preparation was peculiarly difficult on account of the nature of the terrain. On the flat plain there was nothing to conceal the new roads and railways undergoing con- struction, and enemy observers on the ridge crest directed the accurate registering of these for artillery fire when they should come into use. Yet without an abundance of good roads as well as railways the attackers would find their task impossible in a region of mud and marsh intersected by countless streams, canals, and ditches. There was therefore no possibility of a surprise attack. The enemy would know what was coming, and prepare for it. Only in the case of the mines could a wholly unexpected element be injected into the attack, and there was doubt as to whether the Germans might not even be forewarned of this danger. Sir Douglas Haig speaks of the countermining carried on by the Germans and says they realized the peril, 12 but Ludendorff claims that mining operations had largely ceased, that no sound of underground work on the part of the enemy could be heard, and implies that the explosions were totally unexpected. 13 In any event, so saturated with water were the underground formations and so much trouble did the Germans themselves encounter during their tunneling, that they probably had no conception of the stupendous scale of the British opera- tions, even if they suspected mining. The difficulties were in- deed enormous; but they were surmounted through the scientific skill of the chief geologist of the British Expeditionary Force, Lieutenant Colonel T. Edgeworth David, who made a careful study of the alternate layers of sand and clay in the ridge and of the variations of the water levels in each, with the result that tunneling on a grand scale was successfully carried out. After fifteen months of titanic labor twenty-four giant mines were 12 Haig, p. 104. "Ludendorff, Vol. 2, p. 31. THIRD BATTLE OF YPRES 69 placed under the ridge, charged with over one million pounds of high explosives. In the water-soaked plain one of the difficult problems to be solved was the supply of sanitary water to the attacking forces. Advantage was taken of natural lakes and of catch pits con- structed on Mont Kemmel, from which water was piped to the front. Sterilizing barges were established on the Lys, and materials assembled for extending the pipe lines up the ridge as the attack progressed. With the aid of pack animals and carrying parties good water reached the troops within twenty to forty minutes after the capture of new positions, while a week after the assault began water was being piped to the crest of the ridge at the rate of between 450,000 and 600,000 gallons daily. 14 Full appreciation of the formidable character of the ridge obstacle was evidenced not alone by the grand scale on which preparations for the attack were made but also by the care with which the troops were instructed as to the form of the terrain on which they were to operate. Relief topographic models, showing every detail of ridge, spur, ravine, valley, wood, as well as all artificial features, were prepared and carefully studied. The different operations of the attack were rehearsed on the models; and when officers and men knew well both the parts they were to play and the stage on which they were to play them, all was ready for the opening act. On June 7, 1917, at ten minutes past three in the early dawn, Lloyd George, sitting at his telephone in London, heard the simul- taneous explosion of nineteen mines under Messines Ridge. The battle was on. Nineteen volcanoes vomited fire, mud, and men, the country shook with a great earthquake, craters as much as 140 yards in diameter opened in the crest of the ridge and engulfed the defenders. In the awful stillness which succeeded the sub- terranean upheaval observers at a distance stood appalled, and so unearthly was the scene before their eyes that strong men suffered the impression of having awakened in the hereafter to behold a glimpse of the inferno. Then began the victorious " Haig, p. 103. 70 BATTLEFIELD OF FLANDERS advance. Terrified and disorganized, smothered by a terrific bom- bardment, and drenched with liquid fire pouring on them from airplanes in the sky, the Germans were no match for the British forces which swept over them with the precision of clockwork. Ludendorff admits that "the moral effect of the explosions was simply staggering." In a few hours Messines Ridge for a distance of some seven miles, from its southern extremity to the vicinity of Gheluvelt, was once more in Allied hands. The consequences of the victory were clearly expressed by two British officers who, according to one account of the battle, stood on the captured crest, the first looking back into the plain which had been a literal hell for him and his comrades during all the long months that Germany had held the ridge, the second looking eastward to where the Germans in the plain below would now become the chief sufferers. "My God," cried the first, "it's a wonder they let us live there at all!" "It's great to look eastward !" said the other. The Fourth Battle of Ypres With the southern end of the Messines-Passchendaele ridge in Allied hands, preparations could be pushed for the attack on the northern end. Again the preparations on the flat plain were fully exposed to German observation from the ridge north of Gheluvelt, and from minor elevations, like Pilckem ridge,* within the salient; while the water-soaked clay permitted no such deep dugouts and subterranean cellars and caverns as could be used in the chalk upland of the Somme plain to shelter troops assem- bling for attack. Sir Douglas Haig complained of both these difficulties of the terrain. "On no previous occasion," he writes, "not excepting the attack on the Messines-Wytschaete Ridge, had the whole of the ground from which we had to attack been so completely exposed to the enemy's observation. Even after the enemy had been driven from the Messines-Wytschaete Ridge, he still possessed excellent direct observation over the salient from the east and southeast, as well as from the Pilckem ridge to the * Three miles north of Ypres. Spelled Pilken on PI. I. FOURTH BATTLE OF YPRES 71 north. Nothing existed at Ypres to correspond with the vast caves and cellars which proved of such value in the days prior to the Arras battle, and the provision of shelter for the troops pre- sented a very serious problem." Preparations were completed by the last of July, and on the morning of the 31st the fourth great battle for possession of the Ypres bastion, often called the "Third Battle of Ypres" because the Messines operation was given a special name, opened on a front of fifteen miles. The main attack was directed northeast- ward toward the ridge, from that part of the front lying between the Gheluvelt region and the Yser barrier. Fortunately the Germans for some unknown reason, possibly from fear of new mine explosions, had withdrawn some distance from the east side of the Yser Canal barrier north of Ypres, thus permitting the Allied forces holding that part of the line to debouch from behind the obstacle and assist in driving the enemy northeast. The British Commander-in-Chief points out that "this operation greatly facilitated the task of the Allied troops on this part of the battle front, to whose attack the Yser Canal had previously presented a formidable obstacle." As the attack progressed, the greatest opposition was encoun- tered east of Ypres, where the enemy enjoyed the great advan- tages of the main ridge position and where forest patches on the sandy formation had been organized into strong points of resis- tance. Tanks were employed, but encountered great difficulty from the clay soil. Their employment in the preceding battle had not been a great success, partly because they could not advance rapidly enough to be of service. The Allied hold on the ridge crest was extended a little, but only at the cost of heroic exertions. Then intervened in full force the worst enemy of an offensive. Rain began falling that night and continued for four days without intermission, turning the battlefield into a quagmire. The low-lying, clayey soil, torn by shells and sodden with rain, turned to a succession of vast muddy pools. The valleys of the choked and overflowing streams were speedily transformed into long stretches of bog, impassable except by a few well-defined tracks, which became 72 BATTLEFIELD OF FLANDERS marks for the enemy's artillery. To leave these tracks was to risk death by drowning, and in the course of the subsequent fighting on several occasions both men and pack animals were lost in this way. In these conditions operations of any magnitude became impossible, and the resumption of our offensive was necessarily postponed until a period of fine weather should allow the ground to recover. . . . This una- voidable delay in the development of our offensive was of the greatest service to the enemy. 16 In the middle of August the battle was resumed and some ground gained despite great difficulties encountered in the charac- ter of the terrain. But again the rain fell, the ground became im- passable, and the offensive was reduced to a series of small local operations. It was not until September 20 that major opera- tions were once more feasible. Heavy attacks on an eight-mile front on that day gained additional ground on the ridge and wid- ened the salient farther north. The battle raged for several days, and on the 26th another grand assault on the ridge captured the famous Polygon Wood stronghold and Zonnebeke. Here the struggle raged with the utmost fury, and the ground gained was only held at a heavy cost. On October 4 the British battering ram once more launched its powerful blows against that part of the ridge east of Zonnebeke and despite serious obstacles gained an additional section of the crest. On the plain to the northwest a little mound called 19 Metre Hill caused the attackers consider- able embarrassment. Rains were now frequent, and the ground was in terrible con- dition. Nevertheless, another assault was attempted October 9, and progress was made under most painful conditions. Fighting in the mire continued, and on the 12th the indomitable British again plowed forward. But they were attempting the impos- sible. "The valleys of the streams which run westward from the main ridge were found to be impassable. It was therefore deter- mined not to persist in the attack, and the advance towards our more distant objectives was cancelled." Natural barriers were bringing the great drive to a halt. "It was the difficulty of move- ment," reported the British Commander-in-Chief, "far more 15 Haig, p. 116. FOURTH BATTLE OF YPRES 73 than hostile resistance, which continued to limit our progress, and now made it doubtful whether the capture of the remainder of the ridge before winter finally set in was possible." However, the struggle was not given up without further efforts to gain all of the vital crest position. Local battles of the greatest intensity were waged almost daily, and, every time the sun or a favorable wind dried the ground even a little, new assaults on a larger scale were attempted. The men plowed on, "in spite of immense difficulties from marsh and floods in the more low-lying ground," and of vigorous resistance from the enemy posted in excellent positions on the spurs and in the undulations of the main ridge. Ground gained was sometimes yielded because in the advance the rifles of the infantry became so choked with mud that it was impossible to use them when the enemy counterattacked. This was what happened at Gheluvelt. "We went with our rifles and Lewis guns bound up with flannel," wrote one officer according to Conan Doyle,"so as to keep the mud out, and with special cleaning apparatus in our pockets; but you can't clean a rifle when your own hands are covered an inch thick." Nor can one keep the flannel wrappings on guns and at the same time use them in pushing an attack. Ludendorff paints the picture from the German side : "The horror of the shell hole area of Verdun was surpassed. It was no longer life at all. It was mere unspeakable suffering. And through this world of mud the attackers dragged themselves, slowly but steadily, and in dense masses. Caught in the advanced zone of our hail of fire they often collapsed, and the lonely man in the shell hole breathed again. Rifle and machine gun jammed with the mud. Man fought against man, and only too often the mass was suc- cessful." 16 Several successive assaults on the ridge at Passchendaele gave the Allies possession of that village on November 6, but the forces of Nature finally triumphed over the forces of man. The lowland had long been one vast morass. Marshy beeks, or brooks, such as the Steenbeek, the Brombeek, the Watervlietbeek, and 16 Ludendorff, Vol. 2, p. 105. 74 BATTLEFIELD OF FLANDERS others, were flooded marshes transformed into an awful m'.re by heavy shelling. In the seemingly bottomless mud the men could not outmaneuver the concrete "pill boxes" and, fully exposed to their murderous fire, fell by the thousands. On the ridge, where the ground was a little less difficult so far as concerns its physical condition, the form of the surface afforded compensating advan- tages to the defenders. Artillery support of attacks was difficult because guns stuck in the mud and because shells bursting in the plastic material did less than the normal damage. "Throughout the major part of the Ypres Battle, and especially in its latter stages, the condition of the ground made the use of tanks dif- ficult or impossible." As the mud became worse the task of keeping weapons clean and serviceable became more and more difficult and was an important element in bringing the offen- sive to a close. The long battle, which had raged for three months and a half with a persistence and a fury theretofore equaled only by the 1916 Battle of the Somme (p. 144), was ended by the inter- vention of insuperable natural obstacles before its objective had been fully attained. Much had been won at great cost. Most of the Ypres bastion was in Allied hands; the line of the Yser, now flooded from Nieuport to Ypres, impregnable to frontal attacks and solidly buttressed on the sea at one end, was at last reason- ably secure from the danger of a turning movement at the other end. But the northern end of the Messines-Passchendaele ridge, the eastern wall of the Ypres bastion, still remained in enemy hands. The whole bastion would, in the opinion of the British Commander-in-Chief, have been captured in a few weeks had not the character of the terrain offered greater difficulties than did the military power of the German armies. Looking back on the whole history of the operations and giving full credit to the fighting ability of the enemy, he could say: "Despite the magni- tude of his efforts, it was the immense natural difficulties, accen- tuated manifold by the abnormally wet weather, rather than the enemy's resistance, which limited our progress and prevented the complete capture of the ridge." n 17 Haig, p. 133- BATTLE OF THE LYS 75 The Battle of the Lys While the fourth battle of Ypres was was still in progress, Canadian troops in a brilliant local operation on August 15 seized and held Hill 70, dominating Lens on the north. This low elevation rose enough above the general level of the plain of I he Gohelle in the transition belt to give good observation over a wide range of country. With it in German hands the Allies had suffered severely from accurate artillery fire; now that it was in their own hands they could direct a well-controlled fire upon the defenses of Lens. Other local operations occupied the armies on the Battlefield of Flanders during the remainder of the year 1917, but it was not until April 9, 191 8, that the plain was to witness another battle on the grander scale. The great German advance across the plain of the Somme, be- gun March 21 as the first move in a desperate effort to achieve victory before American strength should be fully developed, was wearing itself out, and Ludendorff was ready to launch a new attack. He had rejected the Flanders plain as the scene of the first offensive operation because in the south the Lys barrier, against which the main force of the blow would strike, was at that season too formidable an obstacle. The possibility of a direct attack against the still more formidable barriers of the Yser floods and the Ypres bastion was not even considered. But in April excep- tionally favorable weather had dried the muds and marshes of the Lys valley to some extent, and it seemed feasible to launch an offensive across that part of the plain. 18 Ludendorff fully realized that the operation could not achieve any real success unless the two strong natural bastions defending the plain on the north and south were captured. He had just assaulted the Arras bastion in vain (p. 188), but would try again by a flank attack from the north as soon as he had pushed west- ward across the plain some distance. "To have the high ground in our possession," he writes, "was bound to be decisive in any fighting in the plain of the Lys." He would likewise assault the Ypres bastion from the south as progress was made in the plain. 18 Ludendorff, Vol. 2, pp. 220, 238. 7.6 BATTLEFIELD OF FLANDERS . i - r I' j 1 — »4«» ■»" . a - r ■ . .^y OJ J^ 0{£ 6S S | O C o •- BATTLE OF THE LYS 77 ft is possible, indeed, that the bastion, being the key to the whole situation in Flanders, was the main objective of the operations, and that the Battle of the Lys was as a whole merely another Battle of Ypres. Early on the morning of April 9, under cover of a dense mist or fog, the Germans struck on a ten-mile front from the strong point on slightly higher and drier land in the La Bassee-Givenchy region to a point just southwest of Armentieres. Portuguese troops holding several miles of this front were routed, the attack- ers poured through the gap, and the defenders fell back at once on the natural barriers of the Lys River and its tributary, the Lawe. Bridges over the streams were unfortunately not every- where completely destroyed, and the pursuers were able to cross at several points. Nevertheless the battle continued furiously along the rivers, and parties of Germans reaching the west bank were repeatedly thrown back. Ludendorff complained that "toward Estaires we did not penetrate far enough, but stuck fast in the Lys position; toward the Lawe, also, our progress was but slight." Other features of the terrain were giving him trouble: "The ground was still soft in places, and this made it very difficult to bridge the enemy's trench system . . . The detachments of tanks we had employed had proved a hindrance. It took a long time to get up guns and ammunition;" 19 while lack of obser- vation on the flat plain made proper artillery support difficult. Gradually, however, the outflanking of their defenses by the German advance across the southern edge of the plain and at the crossings which they had been able to retain compelled the British to abandon the line of the Lys, and the Germans pushed deeper into the pocket. The defenders found it difficult to follow closely the enemy's movements and intentions in a flat country where numerous trees cut off the view, especially in the middle of the plain, far from the rimming heights which alone could offer good direct observation; while their field of fire was frequently cut short by willows and other growth along the watercourses. Fortunately, Nature had set a limit to the depth of the advance 19 Ludendorff, Vol. 2, pp. 240, 245. 78 BATTLEFIELD OF FLANDERS possible. As the pocket was deepened and widened it was at every point coming closer to the rimming heights which narrow the plain westward until the bounding highlands meet near St. Omer. Soon the advancing enemy would begin to fight with direct observation favoring the defenders (Fig. 23), and then affairs would wear a totally new complexion. Already, indeed, small outlying elevations in the plain were beginning to tell their story. There were fierce combats at Mont de Lille, Windmill Hill, and Mont de Merris, three low hills on an east-west line from 10 to 15 miles southwest of Ypres. These checked the assaults and destroyed the vigor of the offensive. The bottom of the pocket was approaching the eastern edge of the great Forest of Nieppe south of Hazebrouck. It was high time that the two bastions threatening the Ger- mans on the north and south should be conquered, for the pocket was already so deep that their rear and flanks were seriously menaced. To continue westward into the trap would be folly unless its jaws could be broken. The attack on the Ypres bastion had begun April 10 but was making slow progress. On the 18th a violent assault directed at the Arras bastion was launched on a front extending from Givenchy westward. Here the bastion was protected by the outlying barrier formed by the Aire-La Bassee Canal. At Hinges, northwest of Bethune, the canal has a north- ward bend around a hill which commands long stretches of the barrier to the east and west. If the Germans could seize this hill they might breach the preliminary defense and attack the main obstacle. Eighteen gray-clad battalions leaped to the assault, but the line held. A well-directed British fire mowed down the enemy troops where they attempted to cross bridges over the canal, and the strong point at Givenchy resisted every effort at capture. Along the whole barrier the attack met a bloody repulse. The southern bastion was unshakable. The Fifth Battle of Ypres On the north the preliminary attacks were more successful. As already stated, the Ypres bastion was assaulted April 10, the FIFTH BATTLE OF YPRES 79 Germans sweeping forward on a wide front from Ploegsteert Wood north of Armentieres to Hollebeke. Filtering along the valleys of the Douve River and other streams heading against the ridge, under cover of a mist which made defense difficult, parties of the enemy outflanked the British positions at Ploeg- steert Wood and Messines and secured a foothold on the southern end of the barrier. The lower crest at Neuve Eglise, parallel to, and an outer defense of, the main east-west Mont Kemmel-Mont des Cats ridge, was bitterly contested for two days; and when the British were forced from it they had to yield at the same time other portions of their lines dominated from it. The south- ern wall of the bastion was being gravely menaced. Day after day the struggle raged with unabating fury, the enemy increasing his hold on the southern end of the eastern wall of the bastion inch by inch until both Messines and Wytschaete were in his possession. Then on April 17 the attack was con- centrated on the adjacent Mont Kemmel (Fig. 3), forming the eastern end of the southern wall. Two heavy assaults were re- pulsed on that day. French troops now took over the defense, while the Germans prepared to renew their attempts to capture the commanding position. A week later the slopes of Kemmel were flowing blood as German Alpine troops, skilled in hill fight- ing, were storming the position, supported by a grand assault against the whole eastern end of the ridge from north of Bailleul to west of Wytschaete, and by "a frightful charge of airplanes like tempestuous cavalry" which rained fire from the sky. 20 With prodigal disregard of the cost in lives, Von Arnim flung his masses forward again and again until their very numbers smothered the defenders and Mont Kemmel passed into German hands. The situation was now serious indeed. The apex of the bastion, including the ends of both the southern and eastern walls where they joined, was held by the enemy. From Mont Kemmel the Germans overlooked the entire system of defense and all com- munication lines behind the eastern wall, where the British still 20 Louis Gillet: La bataille des Monts de Flandre, Rev. des Deux Mondes, Vol. Si, 1919. PP- 640-670; reference on p. 663. 80 BATTLEFIELD OF FLANDERS held their advanced positions along the crest, won at such a terrible price the preceding year. Their position was too perilous to endure. Already the British Commander-in-Chief had reduced the risks by withdrawing from Passchendaele behind the marshy valley of the Steenbeek, thus yielding the more northern portion of his ridge holdings and decreasing the number of troops in the apex of the Ypres salient. But so valuable to the enemy was the dominating height of Mont Kemmel that surrender of the entire eastern wall became a painful necessity. With heavy hearts the British retired from the Messines- Passchendaele ridge April 26 and 27, and drew their lines tighter about Ypres. It remained for the Germans to take the rest of the southern wall of the bastion. With the Mont Kemmel-Mont Cassel ridge wholly theirs, the Allies would be compelled, as Ludendorff points out, to abandon the impregnable Yser barrier. All Bel- gium would be conquered, the Channel ports gained, and the Allies forced back on the Artois upland with their front in the form of a dangerous salient which they could hardly hope to hold. Without delay the Germans pressed forward to complete the great task. French and British counterattacks to recapture Mont Kemmel were first beaten off, the new defenders of the hill using the marshy Kemmel Brook to excellent advantage. The French were completely held up at this small but difficult ob- stacle. British troops crossed waist deep in the water but could not effect a permanent lodging on the slopes beyond. Now the Germans in their turn rushed to the assault. But the little valley could not be passed. Four times they swept down the smooth slopes to drive the Allied forces from behind the barrier, and four times a murderous fire mowed them down as with an invisible scythe. Only a heavy blow on a large scale could shatter the defense, and Von Arnim's army was incapable of striking such a blow. Too large a proportion of it lay dead on the sloping walls of the bastion, and the survivors were too exhausted for any more supreme efforts. If the bastion had not been strong enough to withstand entirely the blows of the Ger- WITHDRAWAL FROM LYS SALIENT 81 man battering ram, it had at least smashed the machine beyond repair. Ludendorff might dream of renewing the offensive in this region and completing his difficult undertaking; but the dream would never come true. The Ypres bastion, its walls broken and bloodstained, had saved the whole left wing of the Allied armies from incalculable disaster. Withdrawal from the Lys Salient It was Ludendorff's intention, after his offensives on the Aisne and Marne had weakened the Allies' power of resistance and had led them to reduce their forces in Flanders, to strike again on the Flanders front and attempt the complete reduction of the Ypres bastion, the key to the Allies' whole position on the low plain. 21 But Foch's great counteroffensive blasted all such hopes. Eight days after Foch launched his attack on the Marne plateau (p. 312) the German leaders knew their Flanders offensive was doomed. The German armies must economize men by shortening their front. About July 26 they began the evacuation of the vast accumulations of munitions and stores from the dangerous salient which they had made in the hope of conquering the Ypres bastion and in which they had maintained themselves only at a heavy cost in casualties inflicted by Allied artillery fire concentrated from all sides and accurately directed from the rim- ming heights. Early in August local withdrawals of troops in the salient began, and at the end of the month the retirement on a grand scale was under way. German rear guards made excellent use of the topographic features of the plain, a few men protected by natural obstacles holding the Allies at bay from point to point so that the retreat was able to proceed in an orderly fashion. Along the Neuve Eglise ridge and at Hill 63 the resistance was particularly hard to overcome. But the enemy was not trying to stay in the salient, he was trying to get out; and for this reason topographic advantages were utilized for the moment only, being yielded up whenever the pressure of the pursuers became uncomfortably strong. Early in September the Ger- 21 Ludendorff, Vol. 2, pp. 253, 278. 82 BATTLEFIELD OF FLANDERS mans were out of the trap, Mont Kemmel was surrendered, and the battle line ran across the plain from Givenchy to the apex of the Ypres bastion near Messines. The Germans now held the eastern wall of the bastion, the Allies all of the southern wall. The Sixth Battle of Ypres On the 28th day of September, 19 18, four British divisions sprang out of their trenches at the foot of the Messines-Passchen- daele ridge east of Ypres, without preliminary bombardment, and in a few hours swept the German defenders from the crest and down into the plain to the east. To the south the crest of the ridge about Messines and Wytschaete was attained, while Belgian troops farther north cleared Houthulst Forest and reached the northern continuation of the ridge west of Staden. The King of the Belgians was leading an Allied army of Belgians, French, and British out of the Ypres bastion to final victory. Opposed to them was a weakened and demoralized German army. Less than five divisions were found defending the vital ridge crest east of Ypres. Their quality was far below that of the German army of former days. For more than two months an unbroken series of disasters had been sapping their confidence and lowering their morale, while the disintegrating influence of a skillful propaganda was beginning to destroy their discipline. No topographic barrier, however strong, could be held by such an army. In a single day the British won at slight cost what they had fought long bloody months to attain a year before. A few days later King Albert's forces held all the Ypres bastion. On the north the Belgians had passed beyond the ridge from Dixmude to Staden and were facing the marshy lowland of Handzaeme. Farther south the British were far out over the plain to the east and approaching the lower Lys. Between them the French were beyond the ridge and advancing on Roulers. On October 2 the Germans began a further retirement on the southern Flanders front, from the edge of the chalk upland south of Lens to the Lys at Armentieres. RETREAT FROM FLANDERS 83 The Retreat from Flanders Henceforth it was an unbroken story of retreat, sometimes voluntary as to the given sector because of disaster elsewhere, sometimes forced at that point by a new Allied offensive; always pressed heavily by the armies of His Majesty the King of the Belgians, which now included two divisions of American troops. Throughout the retreat it was the series of more or less parallel rivers, whose northeastward courses were determined by the initial slope of the plain, which the Germans utilized most extensively as temporary lines of defense. By the middle of October the Ger- mans were back behind the Lys barrier on a twenty-five mile front, from near Armentieres to northeast of Courtrai. Farther north the retreat was in rapid progress ; Ostend and Bruges were evacu- ated, and in a few days the northern sand belt west of the Eecloo canal was free of an enemy who had not paused to take much ad- vantage of such topographic defense lines as existed. South of the Lys the retiring Germans stood for a time along the Deule, offer- ing strong resistance behind its marshes and canal. A week later it was the Scheldt from Valenciennes to Avelghem which protected the defeated hordes of the Kaiser on a front of 40 to 50 miles. The Scheldt was part of the famous "Hermann Stellung," which near Avelghem crossed to the Lys, followed it northward to the Eecloo canal, then followed the canal to the Dutch border (Fig. 55). South of Valenciennes the Hermann Line followed up the Selle River as we shall see in a later chapter (p. 211). This great defensive position, based almost entirely on natural obsta- cles, was reconnoitered by the Germans, and the work of strength- ening it begun as soon as it became evident that a retreat from Flanders was unavoidable. The entire Hermann Stellung, across the Flanders plain and into the chalk country farther south, was now occupied by the Germans, who were greatly relieved to have a practically unbroken barrier of marshy valley, river, and canal, 150 miles long, between them and their enemies. When the Hermann Line was finally shattered along the Lys, the Scheldt, and the Selle, the Germans fled eastward toward the Antwerp- Meuse Line, their last important defensive system west of the Rhine. The war in Flanders was over. CHAPTER III THE BATTLEFIELD OF THE SOMME: THE DRY CHALK PLAIN BATTLEFIELD In the early summer of 1914 a dusty traveler on the highway from Amiens to St. Quentin paused in the open expanse of the rolling plain to look about him. North, east, south, or west, everywhere the same monotonous landscape met his gaze, the same dreary waste of country stretching away in gentle undulations to the level horizon. The main highway descended a gentle slope at his feet and in a straight line crossed a shallow depression hardly deserving the name of valley, to mount the gentle slope beyond, disappearing and reappearing as suc- cessive low hills were crossed, like a narrow white ribbon laid down upon the uneven surface with geometrical precision. Near by a minor road curved gracefully around a projecting hill spur and disappeared into a shallow ravine soon lost in the maze of undulations. Far to the north a double row of poplars, silhouetted against the sky line, showed where a more distant highway took its course across the plain, while two dark patches of trees, tiny remnants of once greater forests, rose faintly above the southeastern horizon. An occasional isolated cluster of peasant homes with green shade trees spreading above thatched or tiled roofs, the tall chimney of a sugar refinery, and the smoke of a distant locomotive were the only signs of life in the sleepy landscape. Elsewhere treeless low hills and treeless shallow valleys succeeded each other in endless procession as far as the eye could reach. Such in 1914 was the stage upon which was soon to be played one of the most tragic acts in the World War drama. Could our traveler have visited the plain of the Somme in Note — For Chapters III and IV the reader should constantly consult the detailed map of the battlefield in the pocket (PI. II) and the block diagrams (Figs. 14, 36, and 64). GENERAL ASPECT 85 the days before the hand of man changed the features which Nature gave it, he would have found it almost entirely covered by vast forests. Where now he could sweep with his eye a range of country almost equal to that visible on the rolling Great Plains of western America, then he would perforce have Fig. 24 — Generalized sketch map of the Battlefield of the Somme. Ruled areas are higher than adjacent unruled areas, and especially so where the ruling is con- tinuous. For topographic details and place names referred to in the text, see PL II and Figs. 14, 36, and 64. had to thread his way through a woodland so dense that it concealed even what was close at hand. But very early in his- tory the destruction of the forests was begun in order to make way for a more productive agriculture. When the Roman legions first entered this part of Gaul they found large clear- ings already effected and proceeded to extend their limits. With the coming of the Benedictines in the seventh century and the growth of their agricultural communities the destruc- 86 BATTLEFIELD OF THE SOMME tion of the forests was carried on more vigorously than ever. In the tenth and eleventh centuries there began another period of rapid deforestation, coinciding with the new period of colo- nization and agricultural activity by the monks which followed the era of agricultural stagnation due to the Norman invasions. % Fig. 25 — Chalk area of northern France, on which the Battlefield of the Somme is located. The white portion of the map shows the extent of the chalk, which is, however, often partially concealed by overlying deposits of loam, loess, and clay. Ruled areas represent other rocks. Note that the forests (dotted areas) have been almost entirely cut away from the chalk area, because of the value of the land for agriculture. (Modified after Demangeon.) The rulers of state and of church, the great lords and the pro- prietors of vast domains, eager to increase their wealth by turn- ing forest land into productive fields, encouraged their subjects and serfs by rich rewards to cut down the trees and entrusted the exploitation of the cleared lands to the energetic occupants of the monasteries. By the opening of the eighteenth century the forests were reduced to scattered remnants separated by GENERAL ASPECT 87 broader areas of cultivated fields. Still the process of destroy- ing the woodlands continued, especially when in the nineteenth century the number of small farmers seeking new lands to cultivate greatly increased. In the Department of the Somme there were cut down in the four decades between the years 1792 and 1833 between 7,000 and 8,000 acres of forest. But it was the introduction of sugar beet culture which finally gave to the woodlands their coup de grace. In the feverish desire to gain every square inch of land for the new source of wealth, infertile tracts which could support nothing but trees were robbed of their only possible produce. Such is the history which transformed the vast forests of an earlier day into that dreary expanse of open, rolling plain (Fig. 25) over which swept the waves of German invasion in 1914. During the war what was merely dreary became the acme of desolation. Four times the contending armies carried their work of destruction forward or backward over the plain. The productive fields were trampled under foot, seamed with an endless network of trenches, and pitted with the shell craters of countless bombardments. Shapeless heaps of stone, brick, mortar, and tiles reveal the site of former villages; the sugar refineries and other evidences of human industry are gaunt ruins: stumps show where the double rows of poplars lined the roads and where occasional orchards added their wealth to the scattered communities; while the remaining patches of woodland, some the product of reforestation on infertile tracts, are only marked by shattered trunks and splintered branches of the few skeleton trees which still rise from the ruin about them. Strategic Position of the Somme Battlefield A glance at a map of northern France shows that the Somme is only one of a series of rivers which flow in remarkably straight courses northwestward to the sea (Fig. 26). The Seine, Bethune, Bresle, Authie, and Canche, as well as the Avre branch of the Somme and a large number of smaller streams, have valleys 88 BATTLEFIELD OF THE SOMME as rigidly geometrical and as distinctly parallel as if Nature had laid down their courses with the same giant ruler. Even the Therain, which flows in the opposite direction southeast to the Oise, has a valley parallel to those just mentioned. The secret seems to be that the rocks of the plain are faintly folded, the axes of the shallow folds trending northwest-southeast. Although so faint as usually to be unnotice- able to the eye, these folds were sufficient to guide the running waters and so to determine the location of parallel val- leys. The rocks were sometimes broken as well as bent, and the northwest-southeast breaks, or faults, likewise determined the course of stream erosion, thus emphasizing the parallel topographic features dependent upon rock structure. Geological structure, by determining surface form, exercises a profound influence upon the military history of a region. It is evident that the valleys just described must form a succession of natural defense lines against an enemy advancing north or south across the chalk uplands of Artois and Picardy. Indeed, it has well been said that these valleys give to northern France a distinctly military character. History offers ample confirmation of this conclusion. During the War of the Spanish Succession the French armies took their final stand in the cam- paign of 171 1 behind the Canche, Marshal Villars establishing his famous ne plus ultra lines on the southern side of this natural Fig. 26 — Parallel valleys of the Somme and neighboring rivers, due to parallel folds and faults in the underlying rocks, and constituting natural defense lines of great historic importance STRATEGIC POSITION 89 trench from its mouth to its source, whence they continued eastward behind the marshes of the Scarpe and Sensee. The Duke of Marlborough found the French position along the Canche practically impregnable and directed his strategy against a portion of the line farther east. The Authie next south repeatedly served as an important line of defense, and the fortified town of Doullens, which guarded the eastern end of the valley barrier, so often suffered the horrors of fire and sword that certain etymologists would derive its name from vallum dolens, the vale of sorrow. Even the small valley of the Maye takes a significant rank in military geography, for it was along the northern slope of this depression, where it makes a re-entrant into a small side ravine, that the English army de- ployed to receive the shock of the French attack at the historic battle of Crecy. The soldiers of France, advancing from the south, were forced to cross the natural trench under fire from the English archers. These were placed in rows one above the other on the terraced northwestern wall of the ravine, which was almost impossible of ascent by the French cavalry. Military critics have ascribed the overwhelming victory of the English not merely to the indiscipline of the French forces but in consider- able part also to the magnificent natural position chosen by King Edward for the stand of his army. The River Somme has always been one of the most important military barriers of northern France. It served as a formid- able line of defense for the Roman Empire during the time of its decadence, and the valley walls are still dotted with traces of Roman defensive works for a distance of more than fifty miles. In the wars of France against the English, against the Burgundians, and against the Imperialists first one army and then another hurled itself against this natural moat with its steeply sloping walls and its floor of river and marsh. When Edward III invaded France and moved north toward Flanders, it was on the line of the Somme that the French under King Philip first tried to stop him; and Froissart 1 records in detail ^eanFroissart: Chroniqucs (edit.de Lettenhove), Vol. 5, pp. 1-22, Brussels, 1870. 9 o BATTLEFIELD OF THE SOMME the repeated failures of the English soldiers to cross the obstacle, over which Edward assured his disheartened followers that "God, the Mother of God, and St. George have provided a passage, I know not where." When, seventy years later, Henry V repeated the invasion and turned north toward Calais, it was again the line of the Somme that the French forces elected to defend, and with such effect that the English army marched along the southern bank some sixty or seventy miles, from near the sea to a point between Peronne and Ham, before finding a crossing over the marshy barrier. In each case it was only through the treachery of a French peasant that a little-known passage, insufficiently guarded, was discovered to the English; in the first instance over shallows at low tide near the sea, in the second by a pathway through the marshes to a ford across the river. And just as Edward's passage of the Somme was the prelude to the English victory of Crecy, so Henry's passage of the same obstacle was the prelude to the English victory of Agincourt. In 1536 Henry of Nassau moved first on St. Quentin, then on Peronne, with the object of forcing the Somme barrier, which blocked the way to Paris. Philip II besieged St. Quentin in 1557 with the same design of breaching the line of the Somme, but the barrier held him at bay long enough for the King of France to reconstitute his army on the Oise and prevent the advance on his capital. Near the end of the sixteenth century the Spaniards were fighting to force a passage at Amiens. The names of St. Valery and Le Crotoy near the mouth of the Somme, of Abbeville, Amiens, Corbie, Peronne, Ham, and St. Quentin, places which guard the more important strategic points along the barrier, awaken the memories of many a siege and battle. The first-named suffered sixteen different sieges in four centuries, while Le Crotoy was taken, retaken, and burned twenty times within the same period. So also for Abbeville, Amiens, and the rest of the list one might catalogue abundant proofs that the line of the Somme looms large in military history. Even in the Franco-Prussian War, after the military power of France was STRATEGIC POSITION 91 practically crushed and Paris was invested, the French Army of the North assembled behind the Somme barrier and disputed with the Germans for its possession until the capture of Amiens, Ham, and Peronne by the invaders sealed its fate. Thenceforth the Germans utilized it as a protective barrier against the rem- nants of the French army which still held the field farther north, strongly guarding every passage across the river and its marshes while their own ranks were being reinforced. The rectilinear lower portion of the Somme trench is continued southeastward by the marshy valley of the Avre. At its eastern end the Somrae- Avre line is protected by the town of Roye (Fig. 24), the impor- tance of which may be gauged from the fact that it sustained no less than thirteen sieges between the tenth and seventeenth centuries. Enough has been said to demonstrate the enormous military significance of the parallel valleys of Picardy and Artois. It should be noted that these valley trenches are most impressive in their lower courses, in part because the rivers cut deeper and have wider flood plains as they approach the sea, and in part also because the upland rises gradually to the northwest, especially in Artois. Add to this the further facts that the rising upland is more thoroughly dissected into rugged hills as the sea is approached and that it terminates toward the northeast in the formidable barrier of the Vimy Ridge escarpment (p. 101), and one can readily appreciate why invading armies might in general seek to avoid a terrain so difficult to cross, and to choose for their operations the more gently rolling plains farther east, where the shorter valleys have disappeared and where the Somme alone presents a barrier of primary importance. Now it is pre- cisely this latter region that is included in the modern Battlefield of the Somme. The significance of the position occupied by the Somme battlefield becomes clearer when one notes that to the eastward the land rises to merge in the rough country of the Ardennes Mountains. Between the difficult terrain to the west in Artois and Picardy, already described, and the difficult terrain of the 92 BATTLEFIELD OF THE SOMME Ardennes Mountains and foothills to the east there is a low saddle or gap. This is in effect a natural gateway connecting the Paris region on the south with the low country of Flanders on the north; and, since the region of the Somme includes this gateway (Fig. 27) with parts of its northern and southern ap- proaches, one is not surprised to hear it spoken of by the French as the "Seuil de (threshold of) Vermandois." It was inevitable that this threshold or gateway should become the theater of a historic struggle be- tween the French on the one hand and, on the other, the powers con- trolling or seeking to con- trol Flanders. The great battles of the Somme during the World War were but episodes in the latest chapter of that long, historic struggle. But it is not merely in its military aspects that the strategic position of the Somme area becomes apparent. In its economic and cultural relations it has always been little influenced by the lands to the east and west, but profoundly affected by the products and peoples of French Flanders, Belgium, Holland, and even England to the north, and similarly by those of the Paris region to the south; for it was the gateway through which peoples and products moved from one of these regions to the other, and both areas eagerly looked to it for the rich harvests gathered from its fertile plain. In manufactures it was now the Paris district, now the Anglo-Flemish, which predominated in pushing the sale of its raw materials to the factories of Picardy and Artois or which invaded the towns and villages of the plain in the form of colonies of artisans or individual workmen who brought with them the Fig. 27— Location of the Somme battlefield at the natural gateway (between the Artois and Ardennes barriers) connecting Flanders with the Paris region. STRATEGIC POSITION 93 industries of their respective lands. The monotonous surface of the Somme country, interposing few obstacles to ready migration, favored economic penetration from both north arid south; and the quiet struggles for economic and cultural control which inevitably resulted from the geographic situa- tion were but preludes to those clashes of arms by which it was sought to crown the work of peaceful penetration by full political control. The strategic position of the Somme battlefield becomes even more apparent when one considers the great lines of movement across the region, the principal roads, canals, and railways. On first thought one might suppose that the nearness of the sea on the west would result in a princi- pal movement of people and products along east-west routes. The contrary is the fact. The great movement is north and south by roads and railways and by canals which occasionally must pierce the divides between different drainage basins by long tunnels. From the southwest, south, and southeast in France, from Great Britain on the northwest, Belgium and Holland on the north, and Germany on the northeast many of the more important transport lines converge to pass through the gateway between the barrier of Artois and the barrier of the Ardennes and hence to cross the historic Battlefield of the Somme. When the war of movement in the summer of 1914 changed to the war of position in the autumn and the battle lines extended southward from the dunes near Nieuport to the vicinity of Noyon, there to make a right-angled bend toward the east, special elements of strategic and tactical importance character- ized the region of the Somme. It was henceforth one of those „ ■ ■ ' ."' . oCambrai .- oDoullens '■••••" riM . ''••' ■'!'. * ''V>' ■ ' C:} :* oAmiens ■ ■ ' ■• •* :-- . as. .. .. StOuentid ■ £, , G» '• ' ■•' v -o - • -'••% y ,,-,> ■■■„•■ ?t • ' ' /? s./t^fsreo ] Areas higherthan | | Areas la 1 M00meters(328ft) ' ' ID0.meters(328ft.J Fig. 28 — Map showing the in- fluence of the shallow northwest- southeast folds on the Somme battlefield. The slightly upfolded belts (anticlines) remain on the average a little higher than the down-folded belts (synclines). 94 BATTLEFIELD OF THE SOMME .5 c STRATEGIC POSITION 95 o 5 3 a V a a o 0) w 03 a J ~ 96 BATTLEFIELD OF THE SOMME sectors of the front against which the Germans, profiting by the advantage of interior lines, could hurl the overwhelming forces they had concentrated at some central point within the bend before the Allies could determine whether those forces were to be moved west, southwest, or south. It was a sector which tempted such an attack, since it offered large rewards for success. An advance down the Somme would not only cut the lines of communication connecting the northern battle front with the rest of France, but if pushed beyond Amiens would drive a wedge between the British and French armies, thus destroying their unity of action and throwing the British forces back upon the coast, where naval communications would form their only effective connection with their allies. On the other hand, a successful attack by the Allies in the Somme region would have the double advantage of compelling the enemy to retreat from the dangerous Noyon salient and of closing the strategic gateway into Flanders, thereby forcing the Germans to depend upon the more limited lines of transport and com- munication passing through the narrow valleys of the Ardennes Mountains. Thus the precise form of this battle front gave to the historic field of the Somme a temporary and local sig- nificance additional to that which it already possessed for broader strategic reasons. Surface Features of the Somme Battlefield The most striking characteristic of the Somme battlefield is its monotonous succession of low, rolling plain. It is true that the faint northwest-southeast folds, which have produced in the region immediately to the west a marked parallelism of valley trenches and upland strips, are continued into the area now under discussion. A layered map of the battlefield (Fig. 28) shows a suggestion of the northwest-southeast alignment of topography in the form of alternate belts of slightly higher and lower land — the higher belts representing up-folds, or anticlines, and the lower belts down-folds, or synclines. In the 1916 Battle of the Somme the French and British were SURFACE FEATURES 97 fighting to get from the lower land of the Somme syncline to the crest of the anticline next north and to force the Germans down into the next syncline beyond. But while slight varia- tions of altitude are of critical military importance, these un- dulations are so faint and the differences in average elevation Fig. 31 — German camouflage of a road across the Somme battlefield. Roads on the white chalk of the barren plain are readily visible for great distances; hence the necessity for concealing them by strips of this type, which prevent effective observation from enemy captive balloons or other points near the horizon. (French official photograph.) are so small that even the trained geographer would scarcely remark them. He would probably describe the Battlefield of the Somme as a plain of low relief, dissected by the branches of the Somme River and neighboring streams to a late-mature stage of erosion, in which valleys with gently sloping sides are separated by low, rounded hills or by remnants of the flat upland surface which descend gradually toward their margins to merge with the valley slopes (Fig. 29V The topography is not 98 BATTLEFIELD OF THE S'OMME unlike that of the Great Plains in the vicinity of Bismarck, North Dakota (Fig. 30), which are there in a similar stage of erosion. THE LOW HILLS It is clear from what has just been said that there are two elements of prime importance in the topography of the Somme region: the low hills into which the plain has been dissected by stream erosion and the valleys of those streams. Let us first examine in some detail the form and character of the hills, to learn in what way they must have affected the military operations, after which we will turn our attention to the valleys. Perhaps the most obvious result of the low elevations and gentle slopes of the so-called hills is the facility they afford to travel in every direction. No hills are so high and few slopes are so steep as to offer any real obstruction to road making. The main highways, like those from Amiens, Roye, and St. Quentin to Cambrai, and from Ham to Bapaume, could be built in perfectly straight lines for long distances with a minimum of cutting and filling, while less important roads avoided the necessity of any grading by very moderate adjustments of course to the gentle contours of the land (Fig. 31). Parts of the plain are so little dissected that the upland surface remains practically level, and here even the secondary roads have straight courses for remarkably long distances. In the very heart of the Somme battlefield is such an undissected remnant of the level plain, known as the Santerre, across which the road from Amiens to St. Quentin, although one of the east-west roads and hence not of the first rank of importance, runs for nearly forty miles in a straight line. It was along this road that the Australians battered in the German front in August, 1918. In a region where road making is so easy the inhabitants find that it is cheaper to build new roads than it is to take roundabout courses over roads previously existing. As a con- sequence the plain is covered by an intricate network of national highways, departmental highways, good country roads, and passable lanes, such as is seldom found in other parts of France. THE LOW HILLS 99 Railway construction is likewise comparatively easy and cheap, with the result that in addition to the main through-going rail- way lines there are numerous local railways which serve the multifarious needs of this rich agricultural region. Add to this the system of canals which follow the valley bottoms, where the slopes are uniform and the rivers easily controlled, and one can truthfully say that nowhere else is there a region more abundantly provided with cheap and easy means of rapid communication. The Battlefield of the Somme was ideally circumstanced for the transporting of those enormous quantities of men and materials which are essential to modern military operations of the first magnitude. In other respects the form of the terrain in much of the Somme battlefield favored repetition there of military opera- tions on an extensive scale. The general lack of marked topo- graphic inequalities which might give to one side overwhelming defensive advantages, such as existed in the plateau scarps east of Verdun and Nancy, and the absence of such strong and rugged relief as discouraged major operations in the Vosges and the Trentino, tempted first one army and then the other to seek victory by a sudden massing of men, or by an unexpected concentration of fire, or by the use of novel methods of warfare, under conditions which might reasonably be considered as equal except for the advantage to be gained by the element of surprise. When the tank became an important weapon of offense, it was evident that, whereas it might not be utilized to the best ad- vantage on the marshy soil of Flanders nor on the steep, forested slopes of the Vosges, the open, gently rolling plain of the Somme region lent itself admirably to the new method of attack. Hence it was here that tanks were first employed in battle (Battle of the Somme, 1916) and here that they won their most striking victory (Battle of Cambrai, 1917). The flat uplands and gentle valley slopes which invited armies to move at will in every direction also invited the French farmer to clear the forests of old and to extract from the soil abundant harvests; and armies, which fight on their stomachs, saw in the productive ioo BATTLEFIELD OF THE SOMME region of the Somme a battlefield where their stomachs could always be full. In a score of ways Nature had so fashioned the plain of the Somme as to make of it an ideal stage upon which the drama of war should be enacted. There is, however, one element of a natural fortification very noticeable in many parts of the Somme battlefield. In defend- ing important positions by artificial works the military engineer has often constructed low earth ridges or embankments called rideau (French for "screen" or "curtain"). Now, the hill and valley slopes of the chalk country are often interrupted by a series of terraces or ridges, sometimes faint and irregular, sometimes from 5 to 25 feet high (or even more) on their steep, downhill sides, and so closely spaced as to turn the slope into a giant stairway. These natural rideaux, as the French term the terraces, appear to be due in part to long-continued cultiva- tion of the slopes, during which the soil has for centuries been worked down toward the lower part of each cultivated strip, and in part, especially in case of the higher ones, to fractures in the chalk. 2 To prevent the encroachment of higher on lower fields the crests of the rideaux were often required, by custom or by law, to be left uncultivated. Thus, with their steep slopes and crests covered with bushes, trees, and stones cleared from the fields, they form natural defensive positions of considerable value. The northern termination of the Somme type of topography, where it gives place to the low, marshy plain of Flanders, is strikingly abrupt. Northwest of Arras, in particular, there is no transition zone from the rolling chalk upland to the flat clay marshland below. Instead, one of the northwest-southeast trending fault lines described on page 88 cuts across this part of the country, and the area to the northeast has been dropped downward so as to bring the sand and clay formations of Flanders opposite the chalk. As we should expect, erosion has washed away the sand and clay to produce a broad lowland, 5 Albert Demangeon: La Picardie et les regions voisines — Artois, Cambresis, Beauvaisis. Paris, 1905, p. 44. THE LOW HILLS IOI while the more resistant chalk has merely been dissected into rolling hills and valleys (Fig. 32). Vimy Ridge The combined result of the faulting or breaking of the rocks, plus the later erosion, has been to produce a topographic feature of the highest military importance known as Vimy Ridge. A traveler advancing northeastward toward the crest of the ridge finds that the country rises very gradually as he proceeds. ArtoiS *-sw y Y2x 1:5 J^artL-Lr-L Flanders ,- S/Lnd and C/oys U E ■* J^J^^~7~^^^::~^~~^i~=^ -t,LV^/ >tz t~i — '—t— l^n — ,ii 1 Fig. 32 — Eroded fault cliff forming the steep northeast face of Vimy Ridge and separating the rolling chalk upland of Artois from the fiat clay plain of Flanders. Indeed, so gentle is the ascent that he might fail to give it par- ticular notice, unless he turns and looks back. Then he cannot fail to realize that he is rising to an elevation which gives him a commanding view westward to the ruins of Mont St. Eloi and beyond (Fig. 33) and southward far past the towers of Arras. Continuing on his way, he reaches the crest. Instantly there bursts upon his vision a magnificent panorama of the Flanders plain (Fig. 35). The northeastern side of the ridge, the "fault scarp," as the geologist would call it, is steep, the ground dropping abruptly from the traveler's feet to the edge of the plain below. To the northwest, beyond the Souchez River, the ridge is somewhat higher, forming the heights of Notre Dame de Lorette; to the southeast it sinks lower and dies away in the plain of Arras. But toward the northeast nothing blocks the vision until the eye faintly glimpses on the horizon the dim outlines of Mont Kemmel far away in Belgium (Fig. 36). 102 BATTLEFIELD OF THE SOMME THE LOW HILLS 103 io4 BATTLEFIELD OF THE SOMME STRUCTURE OF THE HILLS 105 It is not difficult to understand the great military value of Vimy Ridge. So long as the German trenches lay on the gentle southwestern slope of the ridge, German observation posts commanded an extended view behind the Allied lines (Fig. 33). They possessed the inestimable advantage of direct control of artillery fire, and a surprise attack by the Allies was difficult to prepare when roads and railways were under continuous observation and subject to instant and accurate bombardment. The Allies, on the other hand, from their trenches on the north- eastern slope, had all direct observation cut off by the rising crest immediately in front (Fig. 34). The Germans could maneuver at will over a broad stretch of the Flanders plain beyond the ridge, providing the weather or their own airmen kept the aerial eyes of the Allied armies partly or wholly closed. Not until the Allies should drive the Germans beyond the crest would the situation be reversed and the Allies enjoy the topo- graphic advantages, the Germans suffer the topographic dis- advantages, inherent in the Vimy Ridge position. To make an advance of but a few hundred yards the Allied commanders in a series of terrific battles sacrificed their men literally by the tens of thousands. Let the reader compare Figures 34 and 35 and he will see that the sacrifice was justified; for the gain is to be measured not in linear yards of advance, but in the increased depth and breadth of observation behind the enemy's lines. GEOLOGICAL STRUCTURE OF THE HILLS In a war of movement it is the surface forms of the land alone which play a principal r61e in the military operations. But in a war of position, where the opposing armies "dig in" and for long periods remain rooted to a given piece of ground, the character and structure of the soil and rocks beneath the surface exercise a profound influence upon the condition of the armies and the nature of the fighting. Throughout the entire area now under discussion there ex- tends a layer of chalk (Fig. 37) practically horizontal and so thick that ordinarily its bottom is neither exposed in the deepest io6 BATTLEFIELD OF THE SOMME Fig. 36 — Block diagram of the Mont Kemmel-Vimy Ri STRUCTURE OF THE HILLS 107 NEUVE(£GLISE 0*ESS1NES iKgTARMENTIERES LA BASSEE .:-:s..-' "^^.'^SOU CH E Z "^ eft Vc^feL? -~, ' ~-