30 June 2001

Source: Original carbon copy of typed manuscript.

Transcription and HTML by Cartome

[46 pp.: size, 22,0cm X 37,5cm]


(cover image)


G-2-C A.E.F.



Relief Map Department


21, December 1918.

A.M. Walker, Major Engineers. U.S.A., APO #714, A.E.F.

Col. R.G. Alexander, Chief G-2-C. G.H.Q., A.E.F.

Report of Relief Map Department


1. I have the honor to submit this report on the work of the Relief map Department, G-2-C, G.H.Q., of the American Expiditionary [sic] Forces.

2. The work of this department is to a large extent a new departure in military operations. I have endevered [sic] to make the notes and illustrations complete. It is hoped this report will be of an add in future operations of the American Army.

3. The drawings for the illustrations were made by the Relief Map Department and a few of the plates were produced by our hectograph.

4. The knowledge of the activities of the Germany Army along these lines is incomplete. I have outlined the subject to the best of my ability and am submitting copies of all the evidence of their activities which have come into our possession.

5. I would like to take this occassion [sic] to express to you my appreciation for the encouragement which you have at all times given me in my efforts.

6. I will also take this occasion to express my appreciation for the valuable assistance which Lieutenant Weeks has given to this Department and his special assistance in preparation of this report.

7. The enlisted personnel of this Department has born the monotany [sic] of much of the work and to them is due full credit for their loyalty and hearty cooperation. Special mention is made of the following men:

Sergeant, Armstrong, E.D., Co. "H".
Corporal, Maguolo, George., Co. "A".
Corporal, Simpson, Lusby., Co. "H".
Corporal, Fisher, Edwill., CO."H".
Pvt. 1st Cl. Thoranton, John., Co. "H".

A.M. Walker
Major Engineers, U.S.A.








I. Introduction
A. General Contents of Report.
B. General Uses and Advantages of Relief Models.
C. General Disadvantages of Relief Models.
D. General Uses of Relief Models in Military Operations.
E. History of the Development and Manufacture of Relief Models. By the French and American Armies During the Present War.

1. Work of Service Geographique and M.Chedanne.
a. Chedanne type model.
2. Work of Henri Arnold.
a. Arnold type.

II. Description of the Method of Producing Relief Models in Use by the Service Geographique and M. Chedanne.
A. Preparation of the Base of the Model
1. Material and method
B. Preparation of the Map Data.
1. Bristol board print.
C. Cutting of Contours.
1. Method and Apparatus.
D. Building the Model
1. Method
2. Material and tools.
3. Two methods of preparation for moulding
E. Preparation for Casting Mould.
F. Casting the Mould or Negative.
1. Method.
2. Materials and tools.
G. Construction of the Positive.
H. Finishing the Relief Model
1. Establishment of guide marks
I. Reproduction of the Cultural Detail.
1. Description and use of Japanese paper.
J. Pasting of Paper on Surface of Relief
1. Material and method
K. Material Necessary for Making Chedanne Type Relief Models.

III. Construction of Relief Models by Arnold System.
A. Preparation of the Plan Directeur.
1. Correction of Variations.
2. Size and Scale of Arnold Reliefs.
3. Table of Assemblage.
a. Work of the French Armies.
b. Description of American Plan.
4. Arraching [sic] of map to frame.
a. Precaution.
B. Apparatus and tools.
1. Description of Arnold Plateau.
2. Tools.
3. Clay
C.Process of Modeling.
1. Method and tools.
D. Indicating of Cultural and Topographic Features.
1. Dressing of model to corret [sic] size.
E.Precautions Required in Modeling.
1. Relative to plan directeur and frame.
2. Relative to Plateau and leveling frame.
3. Use of wood as filler.
4. Tracing of contour.
F. Work of Plastering Section.
1. Workmen and tools.
2. Quality of plaster employed.
3. Method of mixing and application.
4. Construction of negative or mould.
a. Method and tools.
b. Reinforcement.
c. Cleaning.
5. Relation of negative and positive, establishment of bases.
a. Detail of calculation.
b. Casting of positive.
G. Work of Painting Section.
1. Apparatus, material and tools.
2. Dressing and painting of relief.
a. Types of reliefs.
H. Storage of Reliefs.
1. Reference reliefs.
2. Backs and filing systems.
3. Necessity of proper drying conditions.
I. Shipment of Reliefs.
1. Two methods of packing.

IV. Relief Maps.
A. General Description.
B. Relief Maps in Use by American Expiditionary Force.
1. History and development.
2. Areas covered.
3. Number printed.
4. Experimentations with shading by hand.
5. Advantage

V. Use of Relief Models.
A. Examination of prisoners.
B. Staff Work.
1. Office and field reconnaisance [sic].
2. Accuracy.
a. Reference to varied scale of relief.
b. Relation of point of light to Observation Point.
c. Modeling of timbered areas.
D. Ballon [sic] Visibility.
1. Method and apparatus.
2. Advantage of camera.
a. Ballon studies.
b. Terrestrial studies.
c. Record of data.
E. Record of Visibility Data by Second Method.
1. Method.
2. Quantity and speed of operation.
3. List of special record maps, data furnished.
F. Studies of German Characters (Translation from French.)
1. Attacks with limited objectives on fixed front.
a. Preparation by use of reliefs.
b. Accuracy and value of information obtained.
2. Mobile warfare.
a. Data obtainable.
3. The use of the Profilograph.
a. Designer.
b. Description of apparatus.
c. Method of taking profile.
d. Quantity produced by operator.

VI. Relief Models and Visibility Studies in the German Army.
A. Relief Models.
1. Captured maps.
a. Description.
B. Relief Maps Made Photographing Relief Models with Shadows.
1. Captured copies.
a. Description.
C. Visibility Study and Visibility maps.
1. Description.
2. Information involved.
D. Comparison of Visibility Charts: German and Becq System.
1. Description of test.
a. Comparison of accuracy from data obtained.

VII. Chedanne and Arnold Comparisons.
A. Production.
1. Advantage of rapid production by Chedanne.
B. Accuracy.
C. Arnold Method.
1. Accuracy of method.
2. Time required.
D. Vertical Scale.
1. Flexibility of Arnold System.
E. Development of Methods by G-2-C.
1. Recommendations for futur [sic] use of reliefs.

VIII. Special Hectograph produced by the Relief Map Department.
A. History.
B. Method of Construction.
1. Description of hectograph.
2. Material used.
C. Method of Operation.

IX. Total Productions and Distributions made by Relief Map Department.
A. Production of Arnold Relief.
B. Distribution of Arnold Reliefs.
C. Distribution of Chedanne Reliefs.





The discussion of the system of manufacture, the development of the special apparatus for the manufacture of relief models and the application of the relief models to certain military studies, as well as the part taken by the Relief Map Department operating under the direction of the Chief of G-2-C, Topography, will form the principal subject of this report.

There were other systems of more or less merit developed in the armies to meet special emergencies, but because of the magnitude of the development of the methods discussed here, and because this is to form a report of the practice and operations of the Relief map Department of G-2-C, Topohraphy [sic], the discussion of the use and manufacture of relief moulds will be confined to the two methods, the Chedanne and the Arnold, which were adopted and used exclusively by this Department.



The use of the relief model for the study of land forms is not a new departure, either in civil or military practice. The relief model, because of the clearness with which it visualizes land forms, has been a popular form of map for a long time. Its great value lies in the fact that it tells more with less chance of error and in less time than any other form of map. The impression formed on the mind from the study of an accurate relief model are clear and concise, and the observer has confidence in the opinions and decisions made.

No flat topographic map either contour or hachure can be made to convey a conception of differences in elevation with such vividness and accuracy as can be shown by a relief model. Considerable training and constructive imagination are required to enable one to read quickly and correctly a map representing relief by contours. Hachure maps are given only an approximate idea of relief and lack the accuracy of vertical measurements.

Aside from the educational value of the relief model in teaching topography and map reading, it represents the best form in which to study the physical features of an area. This is especially true in studies where relief is the main feature to be investigated.



The disadvantages of the relief model are that it is bulky, heavy to handle and difficult to transport, and usually easly [sic] broken. The cost of manufacture and time required to construct the models re [sic] also elements against extensive use of this form of map.



Relief models are of greatest advantage during periods of stationary warfare. When the fighting is of such a nature as to cause retreats or advances on a grand scale, as has occurred at times during the present war, it is often impossible to furnish reliefs of an area desired due to the length of time required to make the model of a given section. However, during periods of stationary warfare it is possible to provide for future changes in position by constructing models of consecutive areas extending to considerable depth on both sides of the line. The Relief Modeling Sections of the French and American Armies have worked in liaison to carry out a program of such nature.

Relief models in general cannot be used to advantage in the field due to the weight and also to the care required in handling and packing to prevent breakage or the marring of the detail on the surface of the model. However, it is possible to make studies from these models at any convenient Headquarters, and to transpose all information obtained to Plan Directeurs or Battle Maps which can be sent to Field Units.

The present war has brought the development os [sic] several methods of obtaining essential data for field operations by use of the relief models. Visibility studies of any considerable width and depth require accurate drafting and calculations from data obtained from a contour map, or from a series of profiles. Such studies covering a considerable area can be mechanically made from a relief model. Very little special training is required. Infantry or Artillery operations on unknown territory can be worked out accurately on these models, and they are even used to obtaind [sic] data for important calculations for indrect [sic] fire of artillery and machine guns.

The relief model was found to be especially valuable in many forms of preliminary reconnaisance [sic] where time and man power is so vital an element, and as an aid in the reconnaisance for the establishment of outposts.

Relief models have become very popular with officers detailed as instructors in map reading and topographic studies. The relation of a topographical map to the terrain it represents is made clear, and the fifficult [sic] conceptions of land forms as shown by a contour map is obtained quickly from these models.


Area of Northern France Covered by Relief Models



At the time of the arrival of the American Expeditionary Force in France, the extensive use of relief models had become well established in the French Armies. From the reports of our officers who visited the headquarters of the French Armies it was soon recognized that our forces would be supplied with relief models, and should be equipped to cooperate in the manufacture and use. Accordingly arrangements were made for the instruction of certain of our officers and enlisted men preparatory to this work.

Two types of relief models made by different [sic] methods, the Chedanne and the Arnold, have been in use during the present war by the French Armies and by the American Expeditionary Force. In Paris under the direction of the Service Geographique, a department was in operation manufacturing relief models on a large scale, and covering a large territory of France.

This department was part of an old established relief model department of the Service Geographique. It occupied a large space in the Hotel des Invalides and had other branches in Paris. At the time of our visit more than forty workmen were employed by the Service Geographique in the manufacture of relief models. At the close of the war this department had produced reliefs covering about 15,000 square miles of territory along the Front from Ostend to Switzerland as shown on index map.

Operating under the direction of this department, M. George Chedanne, an architect of Paris, had a smaller section and employed ten men in the production of relief models. It was largely through the efforts of M. Chedanne that the present form of this system was established, and the possibilities of large production and the advantage of extensive use brought to the attention of the French General Staff.

The interest which M. Chedanne took in the relief model work enabled our Army to have a detail of eight men trained in this method in his studio at Paris. The detail was selected from the enlisted personnel of the 29th Engineers, and ordered to report to M. Chedanne for instruction and work on February 11, 1918. The demands which the American Expeditionary Force made on this Relief Map Department of the Service Geographique made it advisable that the detail remain as long as possible in Paris. Accordingly, the detail of eight men remained in Paris for three months and later, when the Relief Map Department was established at G.H.Q. only four of the men were withdrawn for duty in our plant.

The type of relief model made by the American detail under the direction of M. Chedanne became known in the American Army as the Chedanne or Service Geographique method. For convenience, these names will be used in the description of this method of relief model construction.

In several of the French Armies under the direction of the Canavas de Tir, a method known as the Arnold System named after its inventor, was also in extensive use. M. Henry Arnold, sculptor and a Lieutenant in the French Army, invented the apparatus used in the construction of Arnold Type Reliefs, and has spent about two years developing the production of these reliefs for the use of the French Army. It was under his direction that another American detail of four men assigned to duty for construction of relief models for the use of the American Expeditionary Forces received its first instruction and practice.

Upon completing the course of instruction in these two methods the details were returned to G.H.Q., American Expeditionary Force, and assembled to form a working personnel to produce relief models. The Relief Map Department was formed under G-2-C., Topography, and operated under its direction until the end of the war.

After operating successfully at General Headquarters from May to October 1918, the personnel and equipment was transferred to the Base Printing Plant at Langres and was installed in a specially designed shop on October 20, 1918.


Proposed plan of Turenne Barracks for Relief Map Department G-2-C Modeling Room: Apparatus for Arnold method of making relief. Painting section: Workmen dressing positive casts and painting topographical detail on surface. Plastering section: The workman on the right is in the acto of removing the finished positive cast from the negative.




The model is the original form from which the mould is cast. It requires a wooden base made of a size to correspond with the dimensions of the map to be modeled. This base is of sufficent [sic] thickness to withstand any tendency of the model to warp. The board must be of well seasoned wood, tongued and grooved, and of dimensions slightly larger than the map. The excess allows for shrinkage and trimming. The sections of the top must be make of matched lumber fitted closely together to form a smooth level plane.

In the case of relief models having a mean elevation nearly equal to that of the sea level, it is necessary to have a plus datum in order to obtain a base sufficient thickness for strength. This base should not be more that � inch thick.



The most convenient and rapid method is to have the Map printed on Bristol board or cardboard of a thickness equal to the corresponding contour interval on the proposed relief model. A separate print will be required for each contour interval interval [sic] on the map.



Each contour selected to build the relief is cut from the printed sheets by running a sharp knife along the contour line. The form of knife best suited to this work is that used by bookbinders. In cutting the contour lines great care is necessary to follow the copy exactly. Any deviation will affect the accuracy of the model. The lowest contour is cut first, staring and finishing at the margin of the map boarder [sic]. The number of each contour is placed on the section to facilitate assemblage, where possible the contour sections may be left attached to the margin until ready for use. As summits are reached, more care is necessary to see that all contour intervals of the same elevation are cut and filled with the appropriate elevation number.

At this stage each contour interval may contain may detached pieces. The numbering is of great convenience in the next stage of construction.

It is possible to employ several men at one time on the preparation of the model by this method.

The cutting of contour lines finishes this stage of the work. The time required to cut the sections will vary considerably. The ordinary contour map on the scale of 1/20,000 to 10 meter contour intervals requires an average one and a half days, of eight hours to prepare.



The relief model is built on a wooden base. A copy of the entire map is first glued to the wooden base, and on it the successive contours are glued,, and beginning with the lowest contour interval and continuing to fasten the successive layers until the highest point is reached. Hot glue of good quality is applied in a fairly thin coat. Two men are required to handle this work to best advantage. It is necessray [sic] to adjust the layers with care to insure a good register. A movable square frame is used in adjusting the larger sheets which extend to the corner of the model. The frame is placed around the corner of the base and the card board fitted to it at each gluing. It is made to cover one corner and two sides of the model. At intervals of five layers the model is placed in a press which is similar to a large size letter press. About fifteen minutes is sufficient time to keep the model under pressure under at each interval. After each pressing the surplus glue is removed. When the final layers have been added certain guides marks are placed on the surface of the model to reproduce in the plaster casting.

Having completed the building of the relief. The model will be in the form of a series of steps. The sides are too irregular to permit plaster casting, because the gluing and cutting of the contour sheets have formed a model, the surface of which contains many undercuts to which the plaster would adhere in setting.

This model is prepared for moulding of the negatives in two different [six] ways.

1st: It is covered with a coating of refined beeswax. With suitable tools the surface is made to slope correctly between contour intervals, by working the was into the steps formed by the cardboard contours. Hard steel tools shaped as shown in the cut are used for this work.

2nd: The surface of the model is given three coats of varnish, which renders the surface sufficiently smooth for the casting of the plaster negative. In this case, the steps are transferred to the negative and are modeled smooth with sculptor�s tools.


Base board, Registering frame and Casting frame.



The model is now ready for casting. The base is trimmed to give the correct elevation for the surrounding relief models and the edges planed to the correct size.



The model is placed on a smooth table with stone or marble top. Form boards which are of sufficient height to give a firm base to the negative and which may be made to standard dimensions to be used on a series of casts are placed around the model. Boards of three quarter inch seasoned pine, dressed, varnished and of a height to assure three quarters of an inch above the highest point on the highest model will be satisfactory for a series. The length of the boards is governed by the area of the relief model. A standard size should be established. The form boards are used as a frame as shown in a sketch. The rope or cord used to keep the frame rigid is held in place with wedges.

The surface of the model and the frame are given an even coating of a light vegetable oil. Olive oils or similar oils are desirable. Plaster workers differ as to the best treatment. A soft soap solution is sometimes used to prepare the surface for the oil.

The proper mixing and handling of the plaster paris is apart of the operation difficult to describe and belongs to the trade of the art plaster workers. The mould is reinforced with wire and burlap, and trimmed to the correct size. More complete directions are given in a following section on plaster work.



The construction of the positive is practically the same as described for the negative. Positives are cast from the negative mould in the desired quantity and dried for several days.



A certain number of guide marks should be established on the positive casts. It is more convenient to mark them on the model and the marks are in turn transferred automatically to the mould and the positives by casting. They will consist of prominent or other characteristic marks of the map. The cast is given a filler coat of light shellac, varnish or gelatine wash.



The desired culture may be reproduced on the model in two ways. It may be painted on the surface by hand, using the original map as a guide in conjunction with the guide lines mentioned before. The most rapid and satisfactory results are obtained by use of sheets of thin Japanese paper on which the maps detail is printed from the original lithographic plates. The paper to be used is tough and thin and can be stretched easily when wet and made to cover much of the irregularity of the relief model surface. Beyond the limits of the stretch of the paper, it is cut to fit the relief more closely.

The Japanese paper is very thin and requires care in handling, especially when wet. However, one skilled in the work can cover the models rapidly. In cutting the paper to conform to the relief the cut should follow the contour of the hill as closely as possible.



Good quality flour paste is suitable for the work. After proficiency has been acquired, large sections of paper can be handled at one time. Care should be taken to make the culture and contoures [sic] conform to the relief of the map model. The guide marks established on the model are used to insure a correct location of the paper sections. In finishing the surface of the model, an agate burnisher is used to smooth the map into position. With a low surface of relief two or three models may be covered in a day. Two or three days are required to cut and fit the map in place on models having large differences in elevation. The whole process is comparatively rapid and easy, and the production of relief models by this method can be made without excessive expense.

The Service Geographique of France fixed a price of $3.37 for each model. It is impossible to state the proportion of manufacturing cost represented by this price. The American Expeditionary Force received approximately 1,500 of these relief models.

The French Relief Model Department at Paris with a small force of workmen has covered the entire battle front, which was more or less stationary during the period of war, with the Chedanne type of relief model maps. They also made portions of other battle areas. This series of relief models covered about 15,000 square miles of territory, represented by 292 separate panels on a scale of 1:20,000. See index map.



(a). Construction of models: Bristol board, thickness 4/10 m.m. for models with the vertical scale doubled. (Quantity: 500 sheets Bristol board, thickness 2/10 m.m.; for models with the vertical scale the same as the horizontal scale. (Quantity: 500 sheets)

Cutting knives.
Glue pot
Wax pots
Wax scraping tools
Sharpening stone.

(b) Covering of the proofs:

Pasting of paper on negatives.
Japanese paper to the proper dimensions for covering the models after printing. 100 sheets.
Ordinary white paper for printing editions of the maps to be pasted in the box for reference to the corresponding relief map. 100 sheets.
Corrugated pasteboard for boxes. 1000 sheets.
Paste, brushes, and containers.
Straight edges
Water color outfits, complete with brushes and trays
Shellac, brushes and containers
Agate finishing tool]Blotters, cloths, inks, pens
(Tool Work, modeling and casting)
Wood top table
Marble top table
Plaster mixing pans
Water pails
Small trowels
Plaster spoons
Dust brushes
Modeling clay
Wire no. 17 or 19
Wide chisels
Plaster brushes
Liquid soap
Olive or peanut oil
Plaster of Paris
Containers for oil and soap
Finishing tools
Wood for platform or base of the model jointed cross pieces and moulding forms. 1 set standard form for scale 1/20,000.


Relief Model Press: Chedanne Method Modeling Tools: Chedanne Method Pasting Brushes Chedanne Method Tools and Brushes: Chedanne Method
Rough Relief Model made by building up contour intervals out of cardboard This represents the same general system used in building a model by the Chedanne Method



A relief model if made accurately is an exact reproduction of the original contour map or pan directeur. If not, it is of no value. Thus one can say that the model contains all the data furnished by the topographical map, is as accurate as that map and possesses the advantages of the third dimension. Therefore, it is evident that the first step required to produce an accurate model is the careful preparation of the original working map or plan directeur. When known variations in the plan directeurs occur, such as shrinkage of paper, poor printing, or when additional details are required it is necessary to prepare the map accordingly.

The section of the plan directeur from which a model is to be produced must be outlined to the exact size and location as determined from the general plan known as the Table of Assemblage (Tableau d"Assemblage). All Arnold relief models and except those for special school use or those to be used as fillers between two sections of maps constructed on a different [sic] base, are made 60 c.m. x 60 c.m.. All models of general use are made from plan diecteurs scale 1/20,000. The vertical scale is 1/5,000 which makes an exaggeration in relief of four times.

The model must be so built that when oriented they will co-incide on the respective borders both in size and in topographical features. Since it is required to lay out a large area of country in the relief model, the borders of those maps must be established exactly and transferred with the same degree of accuracy during the building up process of the model.

The system of Arnold reliefs was worked out from three different [sic] points by the 11, VII and VIII French armies, each working more or less independent of the others, building reliefs to conform to the requirements of the Army occupying that section of the country. When these sections were projected and worked out to meet during the cours [sic] of the war it was necessary to make small sections or fillers of odd ahspe [sic] and size. The fillers are shown on the table of assemblage of Arnold Reliefs.

The American Army reliefs were made to coincide with those of the Second French Army, the borders which conform to the geographic north. The borders of models built by the VII and VIII French Armies are established on the grid system of the Lambert Projection. Each relief covers an area of 144 square kilometers or about 56 square miles.

Although the topographical map from which the model is made is a five meter contour map, the model is made a ten meter intervals only. In order to afford ease and correctness in working, the map is prepared for the modeler by marking each ten meter contour with a colored pencil, using five colors to show the difference in elevation of fifty meters.

(Show example and scale of colors).

After marking the contours and establishing correct border lines, the section of the plan directeur is ready for mounting on the map frame of the Arnold Plateau.

(Diagram of France and Map Attached.)


Index Map for Arnold Reliefs
Edition of November 1, 1918


This is accomplished by thumb tacking the map in place on the back of the frame and pasting down the sides, drawing the map tight while it is being attached to the frame. A caution is necessary in performing this operation and care must be taken not to draw the map so tight as to stretch it and cause an elongation of the borders and an error in the location of the topographical detail. Thus, it is not advisable to dampen the map in order to insure tightness when dry. At the same time, the map must be drawn tight enough to relieve any sag. When the map is once in place on the frame, it cannot be changed during the process of modeling, since every point on the map must remain in its original position as the model is built up. Therefore any probability of error from sag or contortion of scale must be corrected before the model is started.



Apparatus know as the Plateau Altemitrique Arnold used in making Arnold type reliefs was built by engineers and mechanics at the French Headquarters. The Plateau has six principles [sic] parts; 1st, the level base upon which the model is built; 2nd, the turning base which affords an opportunity to obtain the best lighting conditions on any part of the model; 3rd, the stand upon which the bases rest; 4th, leveling frame used to determine the height and to shape the clay at each contour interval level; 5th, threaded and graduated posts at each of the four corners of Plateau upon which the leveling frame works; and 6th, the frame over which the plan directeur is stretched.




A metal straight edge of heavy material is used in conjunction with the leveling frame. Wooden tools shaped like a knife blade are used to cut away and shape the clay between each contour level. Metal tools are used in making conventional signs for roads, railroads, rivers, houses, etc. on the surface of the completed model.



A special clay known by the French trade name, Pateplastique, is used in building the body of the model. It is of such a gray color as to give a surface of uniform shade which is important to the workman. Clays of other colors change their shades in spots and cannot be used to best advantage.



Having prepared the pan directeur on the map frame, the position and area of the lowest contour level is noted. Clay is spread on the plateau over the approximate area and pressed firly [sic] onto the wood base. The leveling frame is raised to a proper level, one which gives a convenient thickness of clay for a base - about one half inch - by setting the nuts on the calibrated posts to the same level orat [sic] like marks on the scale. By drawing the straight edge across the leveling frame the surplus clay is cut away. Care must be taken to keep the straight edge always in a plane perpendicular to the horizontal in order to prevent the thick edge from cutting the clay below the level desired which is represented by the top of the leveling frame.


Relief-model Map Making: The Arnold Method

Measurements of Plateau


The map frame is put into position over the established layer or plane of clay and the lowest contour traced on the map with a pencil. An impression of the contour is transfered [sic] directly to the surface of the clay. The maps frame is removed and the contour line marked more distinctly in the clay with the edge of the clay cutting tool.

It is now only necessary to cut away the clay and shape the ground up to this lowest contour level. Knowledge of ground below this first indicated level is obtained by observation of the next lower contour which does not appear on the outlined section.

When the terrain below the indicated level has been shaped, the map frame is replaced over the model and the topographical features are indicated up to this contour by pricking thru the map on to the clay with a needle point always held perpendicular. Thus, the erection of roads, streams, railroads, location of towns, forts, cuts and fills, etc., are established on each contour interval as the model is built up.

To continue to the next contour level, the leveling frame is raised a constant height of 2 m.m. and the apparatus is thus prepared for the next layer of clay. To facilitate the work of forming the clay along the established contour, it is rolled into roads about ten or twelve inches long and one fourth inch in diameter. This roll is formed quickly along and back of the contour line and pressed flat, bringing up a thin edge of clay to coincide with the lien, the index finger is used to advantage in this operation. Care must be taken not to overlap the line with the new layer of clay.

Again the location and approximate horizontal area covered by the next or second contour is noted from the plan directeur and clay is spread over the approximate area to about the height of the leveling, frame, which represents the elevation of this second contour. The process of cutting away the surplus clay with the straight edge to form a smooth horizontal plane of clay is repeated, the map superimposed and the second contour traced thru the map onto its respective level of clay. The clay is then carefully cut away between the lowest and this second contour to form a plane contianing both of these lines.

Since the model is built up on 10 meter intervals from contour map of five meter contour intervals, much of the accuracy and true shape of the terrian would be lost if some prevision were not to consider the intermediate contour, i.e., the alternate contour which is not indicated on the model. The following provision is made to shape the clay to the exact form as indicated by the intermediate contour.

Assume that the leveling frame is set at the elevation of the 200 contour and that contour has been traced on the surface of the plane of clay. Locate on the map all portions where the 195 contour is not approximately half way between the 200 and the 190 contour, and trance in at the same level as the 200 contour those portions of the 195 contour. Where the intermediate contour is half way between the tow principal contours, it is known that the slope of the terrain is uniform.

Remove the Plan Directeur from the surface of the clay and trace along the 195 contour with the point of a specially designed tool which cuts a groove in the clay to a depth of half the vertical distance between the 190 and 200 contour levels. The bottom of the groove represents on the model the exact elevation of the 195 contour. The modeling is completed by first cutting the clay away to form a plane containing the 200 and 195 contour. Thus, the ground is formed exactly as the contours indicate.

The model is built up by contour levels, starting with the lowest elevation on the map and completing it a its highest point. Modeling is a mechanical process and requires a skilled hand to obtain results accurately and quickly. Since the work consists of making a plane between the tow contours containing both of the, the land is shaped exactly as the contour map shows it. The hand of the well trained workman has no effect upon the correctness of the model, and, as the French say, if the plan directeur is correct the completed model will be correct.


Steel Modeling Tools Arnold Plateau with map frame directly over model and in position to trace the contour directly onto the plane of clay beneath the frame Arnold Plateau containing model in early stage of construction. The starting point can be located in N.W. corner at bottom of river valley. The section not covered by clay is high ground and will be formed as the higher elevations are reached. Arnold Plateau with complete clay model. Scale of model, 1:20,000, Relief exaggerations four times. Contour map from which model was made rests on easel.



When the building of the model has been completed the plan directeur is cut away from the frame and used as a guide in marking the topographical features on the surface of the clay model. Special steel tools are used to indicate roads, woods, rivers, towns, etc. These features are placed on the surface on the marks indicated by pricking thru the original map. Therefore, the topographical detail occupies the same position on the model as on the map and is accurately located. The grid system of the Lanbert projection is transferred in the same manner from the map and placed on the model in dashed lines.

Finally, the border of the model is checked as to size and angles by use of straight edge and beam compass to check the diagonals. The surplus clay is cut away along the border line making a smooth perpendicular surface for the frame in which the plaster cast or mould is built.



Constant attention and care is required during the process of building up a relief model in order to obtain the result claimed for the relief, that it is an exact reproduction of the plan directeur or contour map. But they are only those precautions which any good workman will observe in his work.

It is necessary that the plan directeur be tight at all times, and that the map frame fits tightly and without play over the leveling frame of the table. If the frame is loose or if the paper contains sag, the various points on the map will be displaced each time the map is superimposed over the model. The frame may be corrected by tightening or adjusting the metal guides on the sides. Sag in the map may be taken up by applying a gentle heat.

Special tests of the plateau and leveling frame must be made before each new piece of work is started, Many of the plateaus have been made from poorly seasoned lumber. Under use in a dry warm room, the wood will shrink, split, and one may find that the plateau is badly in warp. The warping will increase gradually until the wood is thoroughly seasoned. A test may show that one or more corners of the plateau which bear the calibrated posts are considerably lower than they should be. Thus, the calibration on these posts is incorrect. Corrections may be made by testing the leveling frame with straight edges, raising the set screw to bring the frame level and thus correcting for warp.

It is then necessray [sic] to work with the changed calibration on the corrected corner. One can see that this correction is very important, otherwise the model would be built up of a series of warped planes and the final cast or finished map would be warped and its elevations on the lower corner would be incorrect in relation to the points on other portions of the model.

It has been the practice in some cases to use blocks of wood to fill in large spaces in high models. Such a method is only useful where an insufficient amount of clay is on hand. When wood is used it must be specially selected and be well seasoned in order to prevent a movement in that part of the work which had been completed..

The modeler must be careful at all times to trace exactly over the contour line of the map, using a hard pencil with sharp point, which will insure a clear cut impression on the clay. He must be sure to trace every portion of the same contour at the same elevation, because after ha has passed this elevation it is impossible to go back and make a correction to any degree of exactness. In modeling or cutting the clay to form a plane containing the contour at which he is working and the one ten meter below he must handle his tool carefully and avoid cutting out or over any portion of the contour line.



The plaster worker should be a man skilled in the use and handling of plaster of Paris, an ornamental plaster worker. He rquires [sic] a set of steel plaster worker�s tools, a bench three feet by six feet, a flat stone, three quarters inch by thirty six inches; mixing pans, brushed, oil, soap, wire, and burlap. (See plates)



Ordinary plaster of Paris of No. 1 grade is used throughout. On finer models, the first coat may be of No. 0. There is much variation in the grades of the plaster at the present time due to the war manufacture. In large models the base of the mould is often made with a No. 2 grade. However, it has been the practice of the Relief Map Department of G-2-C to make the models entirely of No.1 plaster. Damp storage and lumpy plaster must, of course, be avoided.



There are many individual methods of mixing, of course, but one must adhere to certain principals. [sic]. Four or five liters for the primary coats seems to be all that may be readily mixed at one time.. After a definite moment from which the plaster begins to set, any agitation will weaken and prove fatal to the final product.

Into a shallow pan partially filled with water, plaster is sifted through the fingers in order to obtain the maximum absorption with minimum agitation. After a sufficient quantity (determined by practice) is dropped into the water, a crumbling action of the contents is noted, (this point should be waited for rather than hastened by beating.) The contents are then stirred very gently, and the plaster is ready for application.

Plaster is thrown onto the mold by hand, for fear that a brush or scoop would created [sic] air bubbles detrimental to the final product. It is then worked carefully into place until a layer o some two m.m. thick is produced over the surface of the model. The second coat may be applied by hand and the third by tool, being careful that each succeeding coat is tough enough to take the following layer.



The top of the plateau containing the completed model is carried to the plaster room where the negative or master mould is cast and formed upon the surface of the clay model. To protect the wood work of the Plateau, it is first brushed with a thin coat of sweet oil, any thin vegetable oil can be used. The calibrated posts must be protected from plaster by covering with a special cardboard cylinder.

Several methods have been used in the reconstruction of the plaster of Paris moulds and positives. The final method adopted by the Relief Map Department was developed from the results of continuous experimentation. A set of boards cut to proper size is fitted accurately around the borders of the model to form a shallow box, the bottom of which is occupied by the mode. These boards are adjusted so that the top edges will be paralel [sic] to the plane of the model. This frame then represents a mould into which plaster of Paris is poured and shaped into a negative. The model is first coated with alcohol using a very soft, fine haired brush. Alcohol cuts the grease of the clay on the surface, prevents discolorization of the plaster and adherence to the clay. A first coat of very thin plaster is applied carefully to the surface of the mode. Care is taken to avoid marring the small detail. The plaster should be free from air bubbles. It should be applied quickly and worked carefully into every small crevice of the model. When this first coat has set slightly, but not too firm to prevent a mixing and cohesion of more plaster, a sheet of large mesh burlap of such size as to cover the entire area of model is placed over the plasted and a second heavier coat of plaster applied. The two coats of plaster containing the burlap form the body of the negative cast. The body must be about an inch and a half thick and be capable of withstanding the weight of additional positive casts which will be made from it.

When the body of the negative has set sufficiently to prevent any movement, small boards are placed inside the frame on top of the plaster and parallel to the sides of the frame. They form a trough or mould into which plaster is poured to form a bead around the border of the negative. Inside this bead and near the top is embeded [sic] a heavy wire (No. 6 reinforcement). The wire is cut longer than the sides of the casts and its ends are bent at right angles to overlap on the four corners when laid in place. The wire is covered with plaster which is built up slightly above the top of the board frame. A straight edge is then drawn across the top of the frame cutting away the surplus plaster and forming the level base of the negative cast.

After the plaster has set firmly the frame and boards are removed, surplus plaster cut away and the cast is lifted from the model. Small portions of clay adhering to the plaster may be picked out by pressing a small ball of the modeling clay against the plaster surface.

A dilute solution of commercial Hydrocloric (Muriatic Acid) (HCL) may be used to clean dirty models. Into a porcelain crucible it is further diluted and with a soft brush applied to the model. The dirt is quietly softened and clear pure water is used to remove the acid.

The model having served its purpose, the plateau is returned to the modeling section where the clay is broken up, rolled into cylinders of convenient size and stored for future use in model construction.


Arnold Plateau, top section, 1st process in construction of negative. The form boards are fitted to the borders of the clay model and raised to the proper height by use of straight edge in conjunction with leveling frame which is set at proper height by the calibrated posts. Arnold Plateau, 2nd process, negative casting, calibrated posts protected. First layer of plaster has been applied. Method of reinforcing with large mesh burlap. Arnold Plateau, 3rd process, negative casting. Body of negative is shown complete with burlap between first and second layers of plaster. Rough form is shown for casting of bead around border reinforced with wire.
Arnold Plateau, negative cast complete on clay model. Arnold Plateau, negative cast raised from surface of model  



A standard base or height of relief must be established for any series of models which are to be used together and which are require [sic] to match. The standard established for the territory covered by the French Armies has been adapted by the American Army for its models which coincide with those produced by the French Armies. That base is such that the 240 contour which is a mean elevation in the country under consideration will be 60 m.m. above the base of the model or positive cast. If the 240 contour does not appear on the section under construction its base is built relative to that standard.

The boards used around the clay model in the construction of the negative must be raised to such a height as to give as negative which when turned over in the standard set of boards will produce a positive the base of which conforms to this standard. For example using boards of 105 m.m. in height, the total thickness of the negative and positive when together will be 105 m.m. Locate on the clay model the 240 contour, now in order to produce a negative which will in turn produce a positive the height of which at this 240 elevation must be 60 m/m/ the base of the negative must be established 40 m.m. above the 240 contour elevation on the clay. See plate.

Positive casts or the relief models which are the final products shipped out for use at various headquarters and schools, must be light, durable and free from wind. They must be built by reinforcement to prevent warping out of shape. The method of reinforcement with wire in bead along sides and burlap reinforced body as used in the construction of a negative has been developed by experimentation with various forms of reinforcement. The same method is used in the construction of the positive cast, but it is built much lighter since it does not have to bear the weight of additional material. The body of the positive is only about 3/16 of an inch uniform thickness.

Before casting a positive, the negative is coated with a solution of vegetable oil and soft soap. The mixture is rubbed into the negative with a brush and applied to all the cracks or crevices to prevent the positive from adhering to the negative. The remainder of the process of making a positive is as described for the production of a negative. The positives produced are filed in a rack and allowed to dry about two or three days before going to the painting section for surface dressing.


French Method of Making Casts. French Method - Cont'd U.S. Army Method of Making Casts Positive cast raised on negative cast from which it has been made.
Moulding Tools Moulding Tools - Cont'd Moulding Tools - Cont'd  





Painting material may be purchased from any store selling art material oil tube paints of standard colors are used, brushes, long and short haired, of various sizes, whote [sic] and brown shellac and varnish. Special steel tools are use in dressing of surface detail and in trimming casts to size. Tables about three by six feet are of most service for handling of maps and reliefs. Each painter has his individual table and set of tools, paints and brushes. See plates.



When dry and sufficiently set, the cast is trimmed to the exact border as indicated from the clay model, the surplus clay being cut away with a chisel and finished with a common carpenters plane or plaster plane. In cases where small air bubbles in the plaster mar the surface of the cast, corrections are made by filling with plaster. Any indistinct detail is strengthened with special tools, by reference to the original plan directeur from which the model was made. The cast is given a coat of shellac which is allowed to dry and another coat applied.

The workman then colors the modeled conventional signs using the standard of colors and referring to the plan directeur for exact location of detail. When the oil colors are dry, the surface of the cast is varnished and allowed to dry about two hours before packing for shipment.

Arnold reliefs are painted for three uses and they form the Visibility type, staff type and photograph type. The Visibility type model contains only a portion of topographic detail sesential [sic] to visibility study. It is painted in simple form and signs. The photographic set varies from the visibility only in color, the base coat being pure white instead of yellow and the other colors those which will photographed distinctly and clearly. The Staff reliefs designed for use of schools and sections of the General Staff are prepared more carefully in niciety of coloring and finishing of surface topographical conventions.





Extra relief models are filed for reference and are known as reference reliefs. Should a negative or master mould be damaged another one may be made from the reference relief. Racks built against the wall, having three layers of shelves which are divided into small compartments are used as storage equipment. The relief cast rests on edge in its respective rach [sic] bearing the name of the relief. Negatives, extra reliefs and reference reliefs are filed in this manner. It is important to store reliefs in a very dry place. During the first week which is a period of drying a good system of ventilation or heat is desirable.



Preparation for short distance shipment by motor truck may be made by simply tying two reliefs face to face with paper shaving, or straw wrapped in paper to form rolls. However, a shipment to any distance requiring change of vehicles should be prepared more securely. An effective method is to pack material between the two faces and fasten them tightly together with strips of wood extending across the backs of each cast and plastered solidly into place. A solid package of two casts is obtained. The legs and bottoms may be protected by sacking, paper or straw.



Flat maps which visualize the relief by a system of shadows are called relief maps.

The relief map is made by photographing a relief model with the light coming from the upper left hand corner and at a sufficient elevation to show the relief shadows. The lighting of the model is necessarily directed from the upper left hand corner. Any other location of the light will produce the opposite effect, i.e., it causes the valleys to appear as hills. The same effect may be obtained by turning the relief model upside down.

For issue in large quanities [sic] half tone prints or other reproductive methods are used to publish the relief maps. The American Expeditionary Forces used with great satisfaction half tone copies of the relief maps.


Crating for shipment Typical Visibility type Arnold relief. Scale 1:20,000, relief Scale 1:5,000.



During the month of June 1918, experiments were made at the Relief Map Department to produce relief maps by photography. The results of these experiments were on exhibition at the Relief Map Department at General Headquarters. Relief Map Department at G.H.Q. Much favorable comment was made and gradually a demand was created for copies of certain areas.

The Relief Map Department suggested that such a map would be of value to the aviators, especially to the observer in studying the map preparatory to troop bombardment. At this time facilities for reproduction by half tone or other rapid methods were not available at General Headquarters.

From June to October, photographic copies were made from time to time of the relief models, extending from the Foret d�Argonne to and including the St. Mihiel Salient. These negatives were made on a scale of 1/80,000, but enlargements were made to 1/50,000 and 1/20,000. All of which gave satisfaction but production was limited. The 1/20,000 was the most effective map turned out by this method, but paper supplies would not permit its issue. The map on the scale 1/80,000 were not large enough to give general satisfaction, principally because they were reductions from 1/20,000 relief models. The place names and road symbols became too small in the reduction.

In October, the demand for the photographic copies of the relief maps had increased to such a proportion that other methods of reproduction became necessary and advisable. At that time the Half Tone Printing Department of the Base Printing Plant had begun operations. Samples of a relief map printed by that process were so satisfactory that production was at once begun on a large scale.

As noted in another part of this report, relief models were available for a large area along the battle front of France. Certain areas of special interest to our forces were selected and photographic copies of relief maps were assembled to correspond to the standard Battle Map, i.e.., the Plan Directeur, scale 1/20,000. The scale selected for these relief maps was 1/50,000.

The first half tone copies of the relief maps were delivered October 14, 1918, beginning with an edition of one thousand copies each.

The following maps were issued and plans were made to cover other areas.


CHAMBLEY 3,4 1,000
ETAIN, A 1,100
ETAIN, B 1,100
METZ 100
NOMENY, A 1,000
VERDUN, A 1,100
VERDUN, B 1,100

Twenty five separate maps were issued representing an area of 2200 square miles, extending from Attigny to Luneville along the battle front.

In addition to these maps experiments were made and men trained to produce relief maps by shading the maps by hand to show the relief. This plan was developed with the expectation of preparing areas in advance of the relief models.

It was carried far enough to demonstrate the practical working of the method Etain "B", Conflames, Briey, Maizieres, Vigy, a d Pange were prepared by this method and prints were made from Etain "B". Etain "B", South half had previously been made from the photographs of the relief models.

It is impossible to supply relief models or to transport them in field operations. It was with the view of overcoming this difficulty that the issue of relief maps was made. The relief maps proved to be so useful that for many purposes they supplanted the heavier relief model.

The equipment for this work is simple. The use of the commercial air brush is advised for the shading, the results being more even in tone.

Good work may be done with the ordinary crayon and stump used by artists.



Relief models have been found to be very useful in the examination of prisoners of war.

The relief models are used for solving many problems in staff work, which would otherwise have to be worked out from contour maps or by laborious office or field reconnaisance. Lieut. Col. Becq, Chief of the Telegraphic Service of the Second French Army, worked out a system by which many of the preliminary studies in field reconnaisance could be solved at Headquarters.

In field reconnaisance, the observer must visit the terrain and make observations from various possible points. Clear days must be selected and the enemy fire avoided. This baborious [sic] and often dangerous work is performed by his system at Headquarters where the facilities for study, the production of maps and publishing of results are ample. This system is especially adapted to the study of Visibility, i.e., the location of observation points of both friendly and enemy observers and the mapping of both the visible and the invisible areas from these positions.

The method plannes by Liet. Col. Becq consists of simple apparatus. The principal [sic] of this system is to substitute for the field observer, a point of light and for the terrain a relief model. The light will take the place of the observer�s eye. If a section of the terrain represented by the relief model is arranged on a suitable table in a dark room it is only necessary to put this point of light, represented by a small electric, on a selected observation point to discover the extent of territory visible in all directions from that point. The sections of relief which are lighted by the lamp are evidently visible from the position of the lamp, the territory in the shadow cannot be seen.

Vigneulles Section, photograph of Chedanne relief, scale 1:50,000



The source of light designed by Lieut. Col. Becq is the filament of a 3.5 volt electric lamp, flat bulb type. It is especially necessray [sic] that the bulbs be flat in order to bring the point of light represented by the filament as near the true vertical scale as possible, i.e., the point of light must be as near to the position and the height of the observation point as possible. It is possible by use of the flat bulb type lamp to locate the point of light only slightly higher than the exact position is should occupy, a difference of about one milimeter [sic].

The lamp is attached to a tripod and flexible cord connection to the battery box. The battery box is equipped with connections and batteries sufficient to permit the operation of three lamps at one time, As a substitute for this form of lighting, a pocked flash lamp may be used in an emergency. A convenient form of light for rapid exploration work may be made by attaching the small lamp to a tubular stick used as a handle. Around the lamp socket a close fitting movable cylindrical tube is used as a shutter to prevent unnecessary rays inconveniencing the operator and to adjust the height of the point of rays in proper relation to the vertical scale of the map.



The accuracy of the method depends first on an accurate contour map for a basis on which to construct the relief model. Without this accurate contour map the relief model cannot be constructed to approach the accuracy necessary to warrant the measurements proposed in the methods of study outlined.

It may be demonstrated that the exaggeration of the altitude of the relief model does not change the conditions of the problem. It is evident that a profile taken in any direction from the relief map having a horizontal scale of 1/20,000 and an exaggeration in relief of four times will be on a scale of 1/5,000. A similar profile taken from a relief model of the same horizontal scale containing no exaggeration in relief will be to the scale of 1/20,000. The relation of the heights above the base line of all similar points of the profiles will be as one is to four, which is also the relation of the relief scale of the two models. Therefore, a vertical distance between any two points on the model the relief of which is exaggerated four times will be to the vertical distance between the same two points on the relief model of no exaggeration as one is to four. It is common practice in the establishment of grade lines to use profiles of any convenient scale. The value of accuracy of these grade lines does not change with a change in the scale of the profile used. The rays of light from the selected observation point may be represented by lines projected to any point on the terrain of the relief model or tangent to any point.

Referring to the following diagram, assume the point of light to be established in the similar position on the three profiles of scales 1/20,000, 1/10,000, and 1/5,000 at A,A: and A". The line AC is a line of sight or ray of light tangent to the two selected points B and C. It is evident that the points B� and B" occupy the same relative position as B and C� and C" are located relative to C. Therefore, the line A�B� constructed tangent to point B, when projected will be tangent to point C� and the lines or rays radiating from the observation point will cover similar territory on a relief model of any exaggeration. (Include diagram).

Therefore, the same results may be obtained by use of relief models of any convenient horizontal or vertical scale as long as the scales are constant for any particular study and the point of light is set to correspond to the vertical scale, since the location of the point of light in reference to the elevation of the observation point is one of the controlling factors of accurate work.

The minimum distance which ca [sic] be established by use of the special flat bulb lamp between the correct observation point and the elevation of the filament or point of light is about one milimeter. However, it is possible to reduce this distance and bring the point of light to the elevation of the observation point by moving the light down the slope until it coincides with the horizontal line of sight. It is then necessary to move the light along this level, that is, to follow this contour around the point of observation to obtain the correct visibility data in all directions from the observation point. This method is of advantage in increasing the accuracy of the study of difficult visibility problems. The following data shows the correction made for the three scales of relief under consideration.

Assume the height of the filament above the observation point to be 1 m.m.

Relief scale or measurement is proportional to land measurement. 1 m.m. map measurement on scale 1/20,000 is equal to 20 meters land measurement.

1 m.m. map measurement on scale 1/10,000 is equal to 10 meters land measurement.

1 m.m. map measurement on scale 1/5,000 is equal to 5 meters land measurement.

It is evident that this difference decreases in value with an increase in the scale of the relief.



Timber areas are a large factor in visibility they should be carefully represented in all cases and the factor of obstruction considered in outlining the zones of visibility. The advisability of modeling the forest areas, i.e., indicating the height by relief is subject to question. The factor is an uncertain one in most cases as it is difficult to obtain information of sufficient accuracy to establish as average height for the various sections of wooded areas. Timber is, of course, a great obstruction to visibility but in areas of much activity it has often disappeared. In either case, whether modeled on the cast or omitted, where it is possible a field reconnaisance should be made from special points to varigy [sic] the studies.

Where known, the height of timber may be approximated for visibility studies in various ways. It may be modeled in the negative so that the wooded areas will appear raised above the form or terrain which it covers and be transferred to each cast, or layer of clay, made to the proper thickness to represent the average height of the forest as obtained from available information, may be spread over the proper area of the visibility working model. In active areas such as the Verdun sector this factor has become very doubtful. No attempt has been made to indicate the height of the wooded areas in such territory.

Relief models have been made of the Vosges section where the forests have not been destroyed on which the height of the wooded areas have been indicated approximately to the scale of the relief by dressing the negative as previously described. The conditions of the area and the information available regarding the wooded areas and the time allowed for production are factors governing a decision in regard to the advantage of indicating the timber by the above method.



Maps may be prepared showing the portions of the surrounding country which are visible from a balloon at various altitudes. It is only necessary to substitute the point of light for the balloon and locate it at the desired height above the terrain I [sic] reference to the vertical scale of the relief model.

The information obtained may be recorded in two forms, by photography, or by painting on the surface of the cast the areas in the shadow and transferring the location of the area covered to a flat topographic map which may be issued to field officers.

In making a photographic copy of the area visible from a balloon, the camera is placed vertically over the section of terrain represented by the series of relief models and is operated on a special frame which may be raised or lowered to bring the photographic reproduction to any desired scale. The relief model may be formed under the frame on the floor or in special cases on a large table. The illumination should be as strong as possible without losing a clear definition between the areas in the light and shade. A large bulb similar to those used in automobile lights may be used. Considerable improvement may be made in this method of reproduction.

The photographing of terrestrial observation studies is more difficult and not practical. The light which is very small is used close to the model on the observation point as explained before, and the illumination is not suffisient [sic] to obtain a photographic reproduction without a very long exposure which is not satisfactory. An exposure of about thirty minutes is required in the case of balloon observation studies which have much better illumination.


Lamp Tripod Visibility Outfit Battery Box.



In either the case of the balloon observation or terrestrial observation, the source of light remains constant at the point selected during the study. A mixture of burnt sienna, glue and water is used as a paint which is spread over that portion of the cast which is unlighted from the source of light. The paint is readily soluble in water, and it can be washed from the relief model when the stuty [sic] has been completed.

The reliefs are taken to the drafting room where the data is transferred to a plan directeur by reference to the grid lines, contour, etc. The plan directeur used may be of the same scale as the relief, but it is more convenient for studies covering considerable areas to transfer the data to a map of smaller scale,

Tracings may be made of the essential detail and blue prints used for distribution. It is possible by use of the Dorrel process to overprint the Visibility data on any Plan Directeur of proper scale. Copies have been produced by oiling a plan directeur containing all the visibility data and using it as a tracing from which to make blue prints. It is not easy to get a clear impression by this method. For production on a more elaborate scale, any of the lithographic processes are available.

The entire method of producing visibility results for publication is mechanical and very accurate. A study covering an area of several hundred miles may be made and published in one day by two or three operators. The special maps produced furnishing military information obtained by the use of the relief model may be grouped under the following titles.

Visible and invisible areas.

2. Simultaneous Visibility from several points or from all points of our own or enemy territory.

The location of Observatories and preparation for reconnaisance.

4. Preparation of fire.
Auxiliary Targets.
Advance of Observatories.
Observation post for High Command and Special Observatories.
Records of Visibility.
Official Liasion.[sic]
Balloon Observation � Visibility at varied altitudes.
Studies of general character.
Attacks with limited objectives on a fixed front.
Mobile warfare.



The following information relating to the military uses of the relief models has been compiled from a a translation of the French, "Bulletin de Renesigements [sic] de l�Artillerie", published at the French General Headquarters.

The methods of publishing the results of a series of problems, the rapid and accurate solution of which facilitates the work of the performers has already been described. But the method of suing the point of light in conjunction with the relief model, which indicates in a striking manner the properties of the terrain, leads to studies of a more general character the results of which ought to form a firm basis for provisions and orders of the High Command.



In order to repulse the enemy, it is necessary to deprive him of this observatories, to dominate him through ours and to advance to a position which possesses immediate and continuous views over enemy territory. In a few moments by use of the relief model it is possible to determine the location of the objectives to be reached, which fulfil all the requirements pointed out. Having outlines [sic] the above information, one can establish by a few hours work the following requirements for an attack on a given front, the principal observation zones of the enemy, our defilades (narrow passages forming territory invisible from enemy observers) before attack and the improvement of defilades necessary when the objectives have been reached, and information can be obtained regarding any improvement which may be obtained in our observation of the enemy activity.

If the result of the study indicate insufficient defilades which are small and far from the front, the conquered territory will be difficult one to organize and it will be a "long lasting" sector. On the contrary, if the defilades are compact and near the front, if the observation of enemy activity is considerably improved affording a sufficient view of the enemy zone, the operation will be a profitable one. The first situation will only be a temporary one; it will be necessary to advance in order to prevent undue loss. In the second case it is possible to remain without too many risks.

In considering the line of objectives and the immediate views, special attention must be given to the points where hostile defilades are the nearest and of the greatest importance. Where these defilated zones are easily accessible from the rear it can be assumed that they could be used for counter attacks and that they correspond to weak pints in our line. It will be a better practice to avoid putting the sector limits of a division in that locality, and the troops in that sector should be reinforced.

A map showing defilades, with an indication of our zones of observation will be of value in making provisions for telephonic liaisons, all defilate itineraries, the establishment of narrow gauge railroads and the progression of the position of the artillery.

The Commander, in possession of such informations obtained from the relief maps, my adopt a plan for military operations accurately and in a short time. He can quickly arrange the attack with the means at his disposal and has a basis from which to make demands for reinforcements. Those maps, therefore, point our very clearly the properties of the terrain and are quite sufficient for the establishment of all the rational initial arrangements required for offensive or defensive action.


Calculations may be made from relief models of scale 1/20,000, representing a suffisient [sic] area on both sides of the current front line, to obtain essential information for plans of advance or retreat, for operations in case of an advance to position from which it may be necessary to fall back if repulsed and for plan requiring an establishment of new lines in the rear. Under such circumstances the solution of liaison problems comes first, and it is necessary, aside from the establishment of the axis of communication between the artillery and the infantry, to rebulid [sic] promptly the general scheme of observation and fire adjustment in order to profit to the full extent by the action of the artillery.



The Profilograph, an instrument made by the Service Telegraphique of the Second French Army, was designed by Lt. Col. Becq. By means of this apparatus, it is possible to establish and draw a series of profiles quickly and accurately from the surface of the relief model. See Plate Fig.

The French describe the use of the Profilograpg [sic] as follows:

With regard to firing problems, particularly when indrect [sic] machine fun fire is concerned, for the verofication [sic] of special visibility studies, there are some advantages in rapidly obtaining profiles of the terrain to which sights of groups of trajectory lines or straight lines drawn to convenient scales are superposed in order to obtain a scheme of curves of equal elevation or equal site. The ordinary process of establishing profiles from the contour map is generally slow. Considerable time is required when a large number of profiles are to be established. The Profilograph establishes quickly and mechanically a satisfactory profile from the surface of the relief model.

The Profilograph is composed of a series of rigid needles of equal length, set in a flat wooden frame and held in position by two guide bars extending along the length [sic] of the frame. The guide bars form a pressure device making it possible to hold or release the needles at will. To take a frofile [sic] the board is placed in a vertical position over the relief model and along the line of the desired profile, one of the two end pins of heavier material is set equal to the length of the needles, and established at the origin of the profile to be drawn. The top edge of the board is placed in a horizontal plane with a small hand level, and secured by adjusting the second end pin. Then, the pressure is removed from the needles by raising one guide bar a three points by the special device at the back of the Profilograph and the needles fall vertically onto the surface of the relief model. The heads of the needles then represent a profile in the line occupied by the instrument. The profile line is then drawn on the paper which has been inserted behind the needles by following along the line of needle heads with a pencil. The sheet of paper is replaced by another, the board inverted to bring the needles to the original position and the instrument is set for another profile. A skilful operator ma draw 25 to 30 profiles per hour with this instrument.



Information on this subject is a t present incomplete. There is sufficient on hand to show that the Germans manufactured relief models; published relief maps by photogravure process, studied visibility problems from relief models and issued results in the form of visibility maps printed by the lithographic process.



The Relief Map Department has one relief model made by the German Army. The area covered is south of Laon and contains a section of the valleys of the Ailletts and the Aisne rivers, the villages, Veuville, Chermiay, Maizy and Oeully. The scale of the model is 1/25,000 and the relief scale is 1/25,000, making a ration of the horizontal scale to the vertical scale as 1:1. The following legend appears at the bottom of the model: "Hochbild IG., Mabstab 1:25,000 Uberhchung etwa 1:1; Verm. Abt. 14".

Additional evidence of the manufacture of relief models by the same method is available from the captured relief maps of the sector near Verdun. Those are evidently photographs of relief models with additional shading with an air brush. These two exhibits prove the use of relief models in the two sectors Soissons and Verdun.

The models are to scale of 1:25,000 and made from German maps, names and quadrillage are mostly German.

The form of construction in both areas is similar to the method described under Chedanne method of that manufactured by the French at the Service Geographique. The captured model is crude in construction, heavy with plaster of Paris and a wooden frame. It is evidently built from a cardboard model, small attempt has been made to remove the roughness left by the cardboard layers in the moulding. The overprint used to show the cultural and other topographic detail is the ordinary map print cut along the contour lines to better fit the relief. The model shows the German works in blue and Allied works in red. There is nothing to show the date of manufacture or wheather [sic] it is an early attempt a modeling or that it is the present method.


Method of taking Profile with Profilograph.



From a few captured maps the knowledge of the use of relief maps is established. They are evidently made by photographing their relief models and strengthening the shadows by additional shading. This shading appears very even and was probly [sic] applied with an air brush.

The legend on one series of four calls the maps Verdun. At the bottom is this inscription:"Vermessungs-Abteilung 3 Kartenmaterial der Vern. Abt. 15 (Beyer) Mabstab 1:37,500 der baturlichen lange", and a diagram to show the location of each one in the series. One of this series is covered by a map of the same area on the scale of 1/35,000 and the legend at the top calls it, Verdun, Vermessungs Abteilung No. 15 (Bayer) Photograph Verkleinerung aus dem Reliefplan 1/25,000. It would appear that this map belonged to an earlier stage in the relief model system. It was evidently made by building a model up by cardboard layers and is possible photographed from the original model.



A few copies of visibility maps have been seen. The most complete copy shows the section visible from Dead Man�s Hill to Montfaucon. It is printed in three colors by lithography. The color zones are brought out by ruled patterns in color. It shows the area visible from Hill 304, Deadman�s Hill and Hill 344. The visibility maps are superior in finish and accuracy to any attempt by our forces. Visibility studies judged by the maps captured were an important part of the German Army�s plans of maneuvre. The careful execution both of the preparation of the study and the final overprint on the map by lithographic press work shows how highly this feature, the use of relief models ans [sic] visibility studies may have been in their army.


Photograph of Captured German Relief Map, scale of model 1:25,000. Relief scale 1:25,000 Photograph of hal tone copy of captured German relief map. This form probably represents an early type of relief. Photograph of half tone copy of relief map captured from Germans.



To check the captured German visibility map a study from the same points, namely, Le Mort Homme and Hill 304 was made by the map room of this section. These two charts compare very closely up to a distance of about six kilometers from the O.P�s, but differ considerably as the distance increases beyond this. The difference was very marked in the region North and Northeast of Sivray as observed from Hill 304; here the German map showed an invisible area which was indicated as visible by our study. (Sivray 15 kilometers fro Hill 304).

Therefor a profile was taken from the Plan Directeur from Hill 304. Through the Northern end of Sivray, and plotted with ten fold vertical exaggeration. A profile was also taken along the same line from the German 1/25,000 map and enlarged to the same scale. Finally by means of the profilograph a profile was make along the same line on the Arnold relief map.

The profiles from the French and German maps compare closely, and indicate nothing more than a personal error in making the profiles. The profile from the relief map seems to indicate the same thing with perhaps slight error due to the use of worn casts, rough handling, etc. It should be noted, however, that in no case does the line of sight, drawn from the O.P. Hill 304 (el 297), strike the profile anywhere in the neighborhood of Sivray. But if the straight line representing the line of sight be drawn from a slightly higher elevation than the O.P it will meet the profile much nearer to Sivray and approximate the result obtained in our visibility study.

It would seem, therefore, that the divergence between the German study and ours is due primarily to the inaccuracy introduced by the use of the light, which cannot be placed absolutely on the point of observation. This gives us an actual O.P about a millimeter above the true O.P and introduces also by reflection and refraction a penumbrs [sic] which gives a zone of doubtful visibility. This is an additional argument for using the very flattest bulb obtainable and placing its point of light on the line of sight.

It might also be noted that the German Chart seems to have the accuracy of the map profile. The legend also declairs [sic] that the visibility chart was prepared from the profiles and panoramic views (Rundbildern). It seems likely from present evidence that the system of lights for preparing visibility charts is not used by the Germans. We on our part have apparently sacrificed a point of accuracy at long distances in order to greatly expedite the execution of our work.


33. Photograph of captured German plan directeur containg visibility data.



The Chedanne method because it utilizes the print of the original map produces as extremely detailed model map. For purposes of staff study it is superior for this reason to the Arnold method on which detail is painted by hand.



The methods of production are simple, the system is arranged easily for rapid and large production. The models are light and strong. Methods of packing and shipping easily arranged. It is estimated, from comparison of production that this method is about three times more rapid than by the Arnold system.

The Chedanne system has the advantage over the Arnold system in the organization of labor for production. It is possible to make a greater concentration of labor on each stage of the work. Several morkman [sic] can be placed on the same model preparing the parts, assembling and finishing. For this reason speed can be attained on the early stage of the work. For example several workman [sic] can cut contours on the model. The building of the contour relief also lends to concentration. In this way a model may be finished in a few days, usually from five to seven in a well organized shop. The moulding, casting, positives and finishing by the two methods does not differ so much. However, the difference is in favor of the Chedanne method for speed and also for cost of production. No exact figures of coast can be furnished owing to military conditions of manufacture and organization. It is dependent on a map producing plant possessing the original printing plates, because an edition must be printed on bristol board also an edition is printed on the Japan paper for covering the cast with the surface detail. Both of these facilities are available in war time production and the element of rapid production is the most important factor.



The Chedanne method does not appear to have exactness of vertical scale which is possible to maintain by the Arnold method. In the Chedanne method the vertical scale is controlled by alternate layers of glue and caedboard [sic] neither of which are at any one period capable of exact measurement. Also the surface detail is supplied by a sheet of very elastic, thin paper on which is printed the detail . However, this may not be as great a source of error as is apparent from the description. In the first place it may be demonstrated that the measurement for profile will produce the same result on any vertical scale, therefore, if the vertical scale is relatively true for all sections of the model, the visibility studies will be accurate regardless of whether the vertical scale is exactly the same as represented or whether it is slightly over or under true scale.

The fact of overprint not being in exact position is also not vital because measurements for horizontal distances are not taken from a relief model without special apparatus.. The principal use of the relief model where exact profile is necessary will be in constructing profiles by use of the Profilograph. Here exact vertiacl [sic] scale is necessary. It is possible with suitable precautions to regulate the building of the vertical element by the Chedanne system. This precaution should be attended to.



It is evident that the Arnold method of producing relief models permits the greater accuracy in modeling the terrain and in controlling the vertical element. Each contour interval is not only modeled to its exact scale of relief by the land form is preserved in most minute detail. There is little chance for the personal opinion of the modeler to be exercised. The machine used for modeling is designed to produce an exact copy of the original contour map. There is slight chance for the modeler�s skill in shaping the drainage between contour intervals but that is all. The form of the contour is cut at each layer. The Arnold model machine permits any vertical exaggeration of relief to be modeled. For this reason it is very flexible.

The Arnold method is also preferable where the Becq system of military reconnaisance is the only object of the modeling. With the Becq system large vertical system is to be preferred, because of mechanical difficulties and the advantage of sharp definition between light and shadow zones. In the use of the profipograph [sic] it is also essential that the exactness of vertical scale be maintained.



The making o the usual model in clay by the Arnold system for a m ode 60 x 60 c.m., scale 1/20,000 is 25 working days. The time for casting the negative and positives and finishing the painting of the cultural detail does not differ greatly from the Chedanne system. It requires one an one half days to paint a positive which will correspond closely to the average production by the Chedanne method.



There is no definate [sic] rule to establish the correct relative difference between the vertical and horizontal scale for relief model maps but practice that the choice of a correct vertical scale is important. The relief model should give a correct impression to the eye of the area modeled. It should also be large enough to make profile studies and visibility studies accurate. If possible these two requirements should be embraced in the same mode, otherwise it is necessary to duplicate the series. The proper proportin changes with the horizontal scale. The principle seems to be established that the larger the scale the less necessity for vertical exaggeration. In a model of 1/5,000 no exaggeration is necessary while with a scale of 1/10,000 an exaggeration of 2 to 1 will be ample to give a correct impression of the physical features. For models of a scale 1/20,000 it is advisable to use a vertical scale of 1/10,000 or 1/6,000. The vertical scale of 1/5,000 is in excess of the best proportion to give a true apparent value to the slopes. So far as mechanical results are concerned any working scale will be proper. Since the exaggeration of the relief does not produce any change of the position of one point relative to another but only a relative change in profile lines.



The part which the American Expeditionary Force took in the present War in France required that they adopt methods and apparatus found to be available. It was therefore expedient to use both forms of models and work in liaison with the French Army methods.

Should relief models be attempted in a new field it is most desirable that but one system be adopted for making relief models. It is not advisable to attempt two systems of relief model maps. Two systems are carried on in the French Armies because the development of the use of relief models had not become standardized. At the beginning of the war the manufacture of relief models to cover the whole north boundry [sic] of France was not contemplated. The whole scheme was largely a development of this war. Now that the two systems have been in successful use, and the merits of both proved, it would seem advisable to attempt a combination of the merrits [sic] of both systems and standardize on one system of relief models which would be adaptable to general use.

The necessary time to cover a terrain with relief models is at best considerable. It is necessary that the work be organized on a good commercial basis. In choosing between the two methods here outlined it would seem that the advantages for war production favored the Chedanne method. It may also be advised that imperfections in methods of manufacture of this system as outlined may be overcome. Special apparatus can be designed to use the systems of visibility studies with this method and with these improvements, the advantages of rapid manufacture which the Chedanne offers will recommend it for extensive production. The Arnold system is advised for isolated areas and special detailed studies because it works independent of a Base Printing Plant, in that a single copy of the map to be modeled is sufficient copy.

For special studies it is superior and more economical.

The great interest the American Expeditionary Force took in the use and development of the relief model warrants the continued study of the subject by our officers. It is strongly advised that the necessary equipment be obtained to carry on similar studies by our army. The G-2-C of the American Army had at the close of the war a good working plant for the production of relief models by both systems. This section of G-2 also had the necessary officers trained to carry on studies for the production of relief models by both systems. There were also officers trained to carry on the necessary studies to produce the various forms of special maps which are a result of studies from the relief models.

In the United States relief models should be available for all training areas where officers are instructed in the methods of visibility studies, map reading and reconnaisance.




Because this department was using modeling clay we were called on to furnish clay for a hectograph. The modeling clay was a failure. However, the subject interested one of the enlisted men in the department and he located a clay at the Quarter Master�s warehouse which they were using for stove backing and repairs. It was a form of fire clay and one of the many French clays. His efforts produced a very satisfactory clay Hectohraph [sic]. Three Hectographs were make and these proved very useful. A description of the process of manufacture is inclosed [sic] in this report.

Two special clay hectographs have been produced by this Department and furnished to the 1st. and 2nd. U.S. Armies. From special experimentation made previously in co-operation with the Drafting Department at General Headquarters, if was discovered that fifty of more prints could be pulled from one impression made on the prepared clay. It was suggested that the hectograph would prove useful where the facilities for overprinting by the Dorrel Process were not at hand. Accordingly, the two hectographs were built to cinform [sic] to the dimensions of the regular Plan Directeur, scale 1/20,000.



The basis of the hectograph is a special white clay in finely divided form properly mixed with glycerine and formed into a metal tray of proper dimensions. The tray and cover may be made of sheet zinc, galvanized iron or tin and should be deep enough to hold an inch layer of clay spread uniformly over the surface.

The clay is broken and thoroughly pulverized with a mallet. It should be sifter through a fine screen to break up the small lumps. A special solution of nine parts pure glycerine to one part water is mixed with the clay by kneading it with the hands until every portion is completely moist. One half pound solution is required for ten pounds of very dry clay. By practice one can judge the quantity to use when the clay contains some moisture. A text can be made to determine the proper consistency by pressing out a small section of the clay to form a smooth surface over which a sheet of paper is pressed firmly. If the paper sticks and cannot be removed readily, the clay is to [sic] moist and more dry clay should be mixed with it.

The hectograph tray is filled by hammering sections of the moist clay into place with a wooden mallet until the entire surface of the tray is covered with a uniform layer about one inch thick. The surface is planned or smoothed by drawing across it any form of thin edged metal scraper.



Prints are pulled from an impression made on the surface of the clay from a specially prepared tracing. Every line or letter to be reproduced must be trace with a regular hectograph ink of good quality. Any number of colors may be used on one tracing. The lines should be broad and as heavy as possible to insure a clear reproduction.

Before applying the tracing to the clay, the surface should be tested for moisture. Moisture may be removed by pressing a sheet of soft paper uniformly over the surface allowing it to remain in place a few seconds. The operation should be repeated until the clay appears dull and shows no trace of glycerine.

The tracing is placed face down on the layer of clay and pressed firmly onto it so that every point comes in contact with the clay by rubbing the back gently with the palm of the hand. The tracing should be left in place about two minutes, then removed slowly and evenly to prevent blurring of the lines and detail. Blurred lines may also indicate that the ink on the tracing was not sufficiently dry or that moisture was present on the surface of the clay.

At least forty prints can be taken from the first impression. The first dozen prints should be pulled immediately, but each succeeding print should remain in contact with the clay for a few seconds. A second impression from which about ren [sic] or a dozen additional prints may be pulled can be made by applying the same tracing to a fresh portion of the clay or on another hectograph. More time is required to transfer the second impression to the clay.

The surface of the hectograph is prepared for further work by sponging off the ink with a damp cloth and drying with sheets of paper. When not in use the surface of the clay should be protected by a cloth sheet well saturated with a solution of eight parts water to two parts glycerine. To avoid breaking off clay layer, the hectograph should always remain in the horizontal position.



This department with a small force commenceing [sic] with four men reached a maximum of fifteen men by November 11, produced the following results by the Arnold system.

New Models made by G-2-C 19
Positives made from these models 71
Negatives received from French Armies 32
Negatives made by G-2-C (From French Positives 119
Total positives made from G-2-C and French Models 436



Number of Arnold relief models shipped to:

1st Army 20
2nd Army 84
G.H.Q. 128
Miscel. 42

Total 274

Remaining on hand Nov. 11, 162

Total 436



(Chedanne Method)

Number of models shipped to:
1st Army 266
2nd Army 152
G.H.Q. 244
Miscel. 158

Total 820




Beuvardes 1:20,000 July 22 Sept. 6 39
Chalindrey 1:20,000 April 23 May 30 31
Chatillon sur Seine 1:10,000 June 7 June 17 8 (half size)
Camp de Coetquidan 1:20,000 Nov. 19
Camp de la Coutine 1:20,000 Nov. 20
Camp de Meucon 1:20,000 Nov. 20
Chateau Thierry 1:20,000 June 30 July 13 Not noted
Crebange 1:20,000 Oct. 31 Nov. 17 17 (half size)
Chaumont 1:20,000 Aug. 22 20 1/2
Conde-en-Brie 1:20,000 July 17 Sept. 28 30
Didenhofen 1:20,000 Sept. 17 Nov. 1 26
Fort de Peigney 1:10,000 June 17 July 2 19 1/2 (half size)
Gondrecourt 1:10,000 Aug. 5 Sept. 16 32
Langres 1:20,000 April 24 May 30 31
Luttinger 1:20,000 Sept. 30 Nov. 22 53 [?]
Maizieres 1:20,000 Sept. 16 Oct. 27 30
Metz, S. half 1:20,000 Aug. 7 Sept. 12 19 (half size)
Metz, N. half 1:20,000 June 17 July 29 20 (half size)
St. Front
1:20,000 June 30 July 12 Not noted
Remilly 1:20,000 Oct. 3 Nov. 7 27 1/2
Tennschen 1:20,000 Sept. 18 Dec. 1 56
Verny 1:20,000 June 3 June 21 13
Xousse 1:10,000 June 19 July 15 Not noted





Earl D. Armstrong, Sgt., Modeler
George Magnolo, Corporal, Modeler, draftsman
Lusby Simpson, Corporal, N.C.O. in charge
A.S. MacLeod, Corporal, Painter
Raymond Herbert, Corporal, Painter, draftsman
Ed Will Fisher, Corporal, Painter
J. Leonard Boynton, Pvt. 1st Cl., Modeler
Fred Heyd, Pvt. 1st Cl., Painter
Joseph A. Mitchel, Pvt. 1st Cl., Modeler
John Thornton, Pvt. 1st Cl., Plaster worker
John M. Tucker, Pvt. 1st Cl., Shipment and record clerk
Louis E. Nora, Pvt., Modeler
Frank Bonini, Pvt., Plaster worker
H. A. Smith, Pvt., Painter
William Bell, Pvt., Contour map shading retoucher

The following men were on duty in Paris under the direction of

M. Chedanne;

Joseph R. Fay, Pvt.
Lewis P. Noon, Pvt.
Wallace K, Spooner, Pvt.
Edward T. McNierney, Pvt.

Base Printing Plant,
Hd1. 1st Bn. 29th Engrs.
Relief Map Department
December 21, 1918

Submitted to Col. R.G. Alexander, Chief G-2-C, G.H.Q. American Expeditionary Force.


A.M. Walker,
Major, Engrs.
Chief of Relief Map Department.

30 June 2001
Transcription and HTML by Cartome