When the fighting stopped our military s_America's Munitions 1917-1918

cooperating with them were developing a type of ground sound-ranging

apparatus which it was hoped could be utilized to give troops warning

of the firing of heavy artillery shells in their general direction.

Preliminary experiments show that at a distance of 4.1 miles this

mechanism could record the firing of a gun some 19 seconds before the

arrival of the shell. Under proper circumstances this elapsed time

would enable troops properly warned to seek cover from the explosion

of the projectile. This development of sound-ranging apparatus and its

application to the protection of personnel was made possible by the

far greater speed with which shock vibrations travel through a dense

medium like the earth than through the usual sound-conveying medium,

the atmosphere.

SEARCHLIGHTS.

The searchlight equipment of the United States Army prior to 1914

consisted chiefly of lights located at our coast defenses. In 1916

we began the development of mobile searchlight-and-power units for

field-army work, four horse-drawn equipments, with 36-inch lights,

being ordered first, and later eight other sets, with extensible towers

and gasoline electric generators. When the war was approaching we

ordered 85 sets of the limber-and-caisson type. The caissons of these

sets carried 24-inch lights on extensible towers. In January, 1917,

we ordered 50 high-intensity lights to replace as many low-intensity

lamps at our seacoast fortifications. The first war order was placed in

April, 1917, and consisted of 20 additional searchlights of the 60-inch

dimension, the largest light ordered by the War Department.

After the entrance of America in the war the Engineer Department

began studying the requirements abroad for searchlights used in

defense against hostile aircraft; and in September, 1917, this

investigation resulted in orders for 360 high-intensity searchlights,

693 high-intensity arc mechanisms, and 1,000 glass mirrors of standard

design.

About this time we began looking to the improvement of existing

searchlight equipment. The cooperation of leading scientists,

manufacturers, and Government bureaus was obtained, and the product

of exhaustive experiments was 18 different new kinds of searchlights

either partially or wholly developed.

The first of these were produced, shipped, and were in operation with

the Second Field Army in France on October 1, 1918. This was a new form

of searchlight more powerful than any that had been produced before

that time. It weighed one-eighth as much as lights of former design,

cost only one-third as much, was about one-fourth as large in bulk, and

threw a light 10 per cent stronger than any other portable projector in

existence.

Without going into the details of this mechanism its most striking

innovation, from the standpoint of the nontechnical observer, was the

absence of the front glass through which the beams of the older type

lamps are sent. The absence of the glass, while reducing the weight

and cost of a light, also increased the intensity of the beam of the

searchlight, since any glass, no matter how conducive to rays, absorbs

light to a considerable extent.

In the first part of the war we took the 36-inch lights which the

Government had on hand and mounted them on motor trucks. For generating

power for the lights, motor trucks were equipped with electric

generators operated by the crank-shaft of the truck engine. In moving

about each truck carried not only the light and power unit and

accessories, but provided space for the crew and their equipment.

When we went into the war there was only one firm in the United States

that could make the large searchlight mirrors, but two other concerns

developed the art and the faculties during the hostilities. These

mirrors were of glass and cost about $1,000 at prewar prices. The

maximum output in the United States before the war was three 60-inch

mirrors per week. As the result of governmental encouragement the

production of the 60-inch mirrors increased until it reached the stage

of 15 a week in November, 1918; and the price was reduced to about $900

per mirror, even under war-time conditions with respect to labor and

material. This was equivalent to a price of about $700 per mirror under

normal conditions, or a saving of 30 per cent.

A remarkable contribution of the United States to searchlight science

was the production of a satisfactory metal mirror for projecting the

beam. The metal mirror not only weighed a little less than the glass

mirror, but it cost only one-third as much as the glass one, could be

produced in one-fifth the time, was much less fragile, and extended the

possibility of manufacture to a wide number of industries. The metal

mirror possessed 97 per cent of the reflectivity of the glass mirror.

This slight dullness is inappreciable in searchlight work and more than

compensated for by the other qualities of the metal reflector. This

type of mirror, however, had not yet been put in production when the

war ended.

Our inventors during the 19 months of hostilities succeeded in

reducing the size of carbons used in 200-ampere lamps from 2 inches in

diameter to 1-1/8 inches. This cut the cost of carbons in two, but the

improvements tripled the amount of light developed.

In November, 1918, we were working with assurance of success to develop

a simple system whereby field searchlights could be pointed and

controlled from a distance. Such controls had been used in experimental

work prior to 1917, but the mechanisms were complicated and not

suitable for field service.

The searchlight section of the Corps of Engineers also developed

optical finding devices, which doubled the range of all searchlights

without requiring any modification of the lights themselves. Neither

the ordinary telescope nor night glass is suitable for target finding

by searchlight. The result of our investigation was the development

of a combined observer's chair, eye protector, and searchlight target

finder, the new equipment adding only 10 per cent to the cost of the

searchlight unit.

The range of our modern high-power searchlight, whose target is a ship

at sea, is about 15,000 yards; the range of this searchlight when its

target is an airplane is about 15,000 feet.

BOOK IV.

CHEMICAL WARFARE.

CHAPTER I.

TOXIC GASES.

The first recorded use of suffocating gases in warfare was about 431

B. C., sulphur fumes having been used in besieging the cities of

Platea and Belium in the war between the Athenians and the Spartans.

Similar uses of toxic substances are recorded during the Middle Ages.

In August, 1855, the English Admiral Lord Dundonald, having observed

the deadly character of the fumes of sulphur in Sicily, proposed to

reduce Sebastopol by sulphur fumes, and worked out the details of the

proposition. The English Government disapproved the proposition on the

ground that "the effects were so horrible that no honorable combatant

could use the means required to produce them."

That the probable use of poison gases was still in the minds of

military men is evidenced by the fact that at The Hague conference in

1899 several of the more prominent nations of Europe and Asia pledged

themselves not to use any projectiles whose only object was to give out

suffocating or poisonous gases. Many of the Powers did not sign this

declaration until later. Germany signed and ratified it on September 4,

1900, but the United States never signed it. Further, this declaration

was not to be binding in case of a war in which a non-signatory was or

became a belligerent. Admiral Mahan, a United States delegate, stated

his position in regard to the use of gas in shell, at that time an

untried theory, as follows:

The reproach of cruelty and perfidy addressed against these

supposed shells was equally uttered previously against firearms

and torpedoes, although both are now employed without scruple.

It is illogical and not demonstrably humane to be tender about

asphyxiating men with gas, when all are prepared to admit

that it is allowable to blow the bottom out of an ironclad at

midnight, throwing four or five hundred men into the sea to

be choked by the water, with scarcely the remotest chance to

escape.

The Second Hague Peace Congress in 1907 adopted rules for land warfare,

and among them was article 23 which read as follows: "It is expressly

forbidden to employ poisons or poisonous weapons."

The use of toxic gas in the great war dates back to April 22, 1915, on

which day the Germans employed chlorine, a common and well-known gas,

in an attack against the French and British lines in the northeastern

part of the upper Ypres salient.

The methods of manufacturing toxic gases, the use of such gases, and

the tactics connected with their use were new developments of this war;

yet during the year 1918 from 20 to 30 per cent of all American battle

casualties were due to gas, showing that toxic gas is one of the most

powerful implements of war. The records show, however, that when armies

were supplied with masks and other defensive appliances, only about 3

or 4 per cent of the gas casualties were fatal. This indicates that gas

can be made not only one of the most effective implements of war, but

one of the most humane. It will, of course, be necessary to remove the

noncombatant population from a greater depth of country immediately in

the rear of the fighting lines than formerly, in order that women and

children may not be gassed. This additional sacrifice of territory for

war uses is another element of effectiveness in the weapon.

Since Germany had chosen to utilize toxic gas in warfare, the allied

nations were compelled to adopt like tactics; accordingly England and

France, faced with the desperate situation resulting from advantages

secured by the Germans through the employment of these new weapons,

immediately turned their attention not only to devising methods

for protecting their own troops, but also to securing supplies and

equipment necessary for the utilization of toxic gas as an agent of

warfare against the German Army.

Germany originated thereafter the use of most of the new forms of gas,

but the allied nations and America were actually producing, at the time

of the armistice, gases on a much greater scale than Germany was ever

able to attain. In fact, America itself was producing gases at a rate

several times as great as was possible in Germany.

Prior to the entry of America in the war our overseas observers had

been collecting information bearing upon gas warfare, referring the

facts so obtained to the Ordnance Department in Washington, where the

information was turned over to Lieut. Col. E. J. W. Ragsdale, who was

then in charge of the Trench Warfare Section.

In the early days of our belligerency it was seen that we should need

a plant for filling artillery shell with toxic gases. The Government

in the fall of 1917 bought a large tract of land near Aberdeen, Md.,

to be an artillery proving ground. Approximately 3,400 acres of this

reservation, about one-tenth of it in area, was set aside as the

site for the gas shell-filling plant. This reservation was known as

Edgewood, and the plant erected on the site was called the Edgewood

Arsenal. Work started on the construction of the arsenal on November 1,

1917.

None of the toxic gases in use in Europe, except chlorine and small

amounts of phosgene, had ever been commercially prepared in the

United States. It was the original intention to interest existing

chemical firms in the manufacture of these gases; but there were many

difficulties in the way of such a project, not the least of which was

the ruling of the Director General of Railways that such products as

poison gas be transported only on special trains.

Also we discovered that the private chemical companies were loath to

undertake such manufacture. The exhaustive investigations necessary

before quantity methods of manufacture could be devised would be

uncertain and expensive. There would be great danger to the lives of

those employed in such work. Many of the private concerns were already

crowded with war work. Finally, the new plant equipment which must be

set up would be worth nothing when the war ended, since the manufacture

of such gases would be limited to the period of hostilities. These and

other considerations explain the reluctance of the commercial chemical

industry to undertake the production of war gases.

Consequently the Government was forced to adopt the plan of building

various chemical plants at the Edgewood Arsenal in connection with the

filling plant. By December 1, 1917, it had been decided to build at

Edgewood a chlorpicrin plant and a phosgene plant. The contracts were

immediately let, and the work was pushed through the rigorous winter of

1917-18.

In March, 1918, the Edgewood project was taken from the Trench Warfare

Section of the Ordnance Department and made an independent division

under the command of Col. Wm. H. Walker. In June, 1918, the Chemical

Warfare Service was organized, and the Edgewood Arsenal was transferred

to it. Gen. W. L. Sibert, Director of the Gas Service, took charge of

the activities of the arsenal in May prior to the official transfer.

Chlorine, the raw material for the manufacture of which is common salt,

was one of the principal materials required in the gas-production

program. Although chlorine was a standard product in the United States

prior to the war, it was soon seen that we had an inadequate commercial

supply to meet the requirements of our proposed gas offensive. Chlorine

was used not only by itself, but it was also the active agent in the

manufacture of nearly all the other toxic gases which we required.

Consequently we decided to build a Government chlorine plant with two

50-ton units, giving a daily capacity of 100 tons of liquid chlorine.

The ground for this plant at Edgewood was broken on May 11, 1918, and

the actual production of chlorine begun on September 1.

In July, 1917, the Germans introduced the so-called mustard gas. It

was immediately realized that for certain purposes of fighting this

chemical was the most effective product so far employed, and a large

number of Government experts here at once concentrated their energies

in developing methods for its manufacture on a large scale. Not only

were the uniformed experimenters busy at the Gas Service's American

University Camp, at Washington, D. C., but experimental units were

established at the plant of the Dow Chemical Co., at Midland, Mich.,

at the plant of Zinsser & Co., Hastings-on-Hudson, N. Y., and at the

Government plant which had been started by the Trench Warfare Section,

at Cleveland, Ohio.

Eventually it was decided to erect a large plant at Edgewood for the

manufacture of mustard gas. Not until April, 1918, however, did we

feel that we possessed sufficient knowledge and information to justify

the construction of a mustard-gas plant on a large scale. France

and England also were long in working out satisfactory methods of

mustard-gas production. We began to make mustard in June, and continued

with rapidly increasing output until the signing of the armistice.

It soon became evident that we could not depend upon civilian labor in

the operation of the various chemical plants at Edgewood because of the

danger involved. It was decided, therefore, to utilize enlisted men in

the working crews. As the projects at Edgewood increased in size and

number, the forces at the arsenal grew, until at one time there were

7,400 troops at this point.

Meanwhile the Government had at last been able to persuade a number

of private chemical firms to manufacture toxic gases. The Government

agreed to finance all new construction, but the operation was to be in

the hands of the contracting companies. At each plant the Government

stationed one of its representatives with necessary assistants. In the

spring of 1918, these scattered factories by official order were made

part of the Edgewood Arsenal, each plant being designated by the name

of the city or town where it was located. Thereafter in Army usage the

term "Edgewood Arsenal" embraced not only the group of factories on the

Edgewood reservation, but also included the following projects:

Niagara Falls plant, operated by the Oldbury Electro-Chemical Co.

Project--the manufacture of phosgene.

Midland, Mich., plant, operated by the Dow Chemical Co. Project--the

sinking of 17 brine wells for the purpose of securing adequate supplies

of bromine.

Charleston, W. Va., plant, operated by the Charleston Chemical Co.

Project--the manufacture of sulphur chloride.

Bound Brook, N. J., plant, operated by Frank Hemingway (Inc.).

Project--the manufacture of phosgene.

Buffalo plant, operated by the National Aniline & Chemical Co.

Project--the manufacture of mustard gas.

In addition to these, the Edgewood Arsenal built at points advantageous

to supplies of raw materials four other plants, and operated them as

well. These were as follows:

Stamford, Conn., plant. Project--the manufacture of chlorpicrin.

Hastings-on-Hudson, N. Y., plant. Project--the manufacture of mustard

gas.

Kingsport, Tenn., plant. Project--the manufacture of brombenzylcyanide.

Croyland, Pa., plant. Project--the manufacture of diphenylchlorarsine.

In constructing and equipping the Edgewood Arsenal we laid 21 miles

of standard-gauge railway and 15 miles of narrow-gauge railway, built

nearly 15 miles of improved roadway, and set up two water systems, one

with a capacity of 1,500,000 gallons per day for the manufacturing

purposes of the chemical plants, and the other providing a fresh-water

supply pumped 4 miles with a daily capacity of 2,000,000 gallons. In

all 558 buildings were put up on the grounds of the arsenal. There

were 86 cantonment buildings, with a capacity of 8,400 men, as well

as adequate quarters for officers and civilian employees. Three field

hospitals, a complete base hospital, and separate buildings for Y. M.

C. A. and K. of C. activities indicated the extent of the building

equipment. Three power houses were provided, with a total capacity of

26,500 kilowatts.

In the construction of buildings every precaution was taken to avoid

accidents from the handling of toxic gas, the ventilating systems being

as near perfection as human science could make them. It is notable that

out of the thousands of men employed only four met their death by gas

poisoning. Three of these casualties were due to phosgene and one to

mustard gas.

To show that all of the danger of the war was not confined to the

front, the following table of casualties in 1918 at the Edgewood

Arsenal proper is here given:

-----------------+------+------+-----+------+-----+-----+-----+-------

Toxic agent. |June. |July. |Aug. |Sept. |Oct. |Nov. |Dec. |Total.

-----------------+------+------+-----+------+-----+-----+-----+-------

Mustard gas | 14 | 41 | 190 | 153 | 227 | 47 | 2 | 674

Stannic chloride | | 3 | 8 | 15 | 21 | 3 | | 50

Phosgene | | | 3 | 7 | 2 | 17 | 1 | 50

Chlorpicrin | | 14 | 18 | 9 | 3 | | | 44

Bleach chlorine | | 2 | 39 | 2 | 1 | | | 44

Liquid chlorine | | 1 | 3 | 2 | 7 | 5 | | 18

Sulphur chloride | | | 2 | 1 | 6 | | | 9

Phosphorus | | 2 | 7 | 5 | 1 | | | 15

Caustic soda | | | 3 | | 3 | 4 | | 10

Sulphuric acid | | | 4 | 3 | 1 | | | 8

Picric acid | | | 2 | | | | | 2

Carbon monoxide | | | | | 1 | | | 1

+------+------+-----+------+-----+-----+-----+-------

Totals | 14 | 63 | 279 | 197 | 293 | 76 | 3 | 925

-----------------+------+------+-----+------+-----+-----+-----+-------

As has been noted, chlorine was the only war gas produced on a

commercial scale in America prior to the war. At the ordinary

temperatures chlorine is a greenish-yellow gas of strong, suffocating

odor. Through the combined effects of cold and pressure it is readily

condensed to a liquid and is ordinarily shipped in this form, stored in

strong cylinders.

Chlorine is prepared commercially by the electrolytic process. A

current of electricity is passed through a solution of common salt. The

greenish gas at once arises, leaving behind it a residue of caustic

soda. The apparatus in which the salt is decomposed by the electric

current is known as a cell. The Government plant used Nelson cells,

each with a capacity of 60 pounds of chlorine and 65 pounds of caustic

soda per 24 hours.

The Government chlorine plant at Edgewood was ready for operation in

August, 1918, but was not actually started until September 1. The plant

consisted of (1) a cell house, which had a total capacity of 100 tons

of chlorine per 24 hours; (2) an electric substation for supplying the

current; (3) a brine building, where the salt was mixed with water and

the resulting brine purified; (4) a boiler and evaporation building,

for concentrating the caustic soda from the cells; (5) a caustic fusion

building, for drying the caustic soda and fusing it into solid form for

shipment; and (6) a liquefying plant to condense and liquefy 50 tons of

chlorine per day.

With the exception of chlorine, chlorpicrin was the first war gas to be

manufactured on a large scale in this country. When pure, chlorpicrin

is a colorless liquid which boils at a temperature approximately of

112° C. The compound has been known since 1848. While not so poisonous

as some of the other products used in gas warfare, it is, nevertheless,

an active poison, and has the additional advantage of being a fair

lachrymator, or tear producer.

Chlorpicrin is made by the reaction between picric acid and chlorine.

The chlorine is best supplied in the form of so-called bleaching

powder, which is ordinary chloride of lime. In the manufacturing

process as originally carried out, free picric acid was mixed with

bleaching powder held in suspension with water. Later it was found

advantageous to use calcium picrate instead of picric acid.

Accordingly, the final process was as follows: The bleaching powder

was creamed with water and mixed with a solution of calcium picrate in

large stills holding 5,000 gallons or more. A jet of live steam was

then introduced at the bottom of the still, and the reaction began

at once, the rapidity depending upon the amount of steam introduced.

The resulting chlorpicrin, together with a certain quantity of steam,

passed out of the still and was liquefied in the condenser. The

resulting mixture of chlorpicrin and water was run into tanks, where

the chlorpicrin, being insoluble in water, gradually settled to the

bottom and was run off and used directly in gas shell.

In developing this process the Government was assisted by the Dow

Chemical Co., the Semet-Solvay Co., and the American Synthetic

Color Co., of Stamford, Conn., the principal work being done by

representatives of the Bureau of Mines at the Stamford plant.

[Illustration: AIRPLANE VIEW OF CHLORINE PLANT, EDGEWOOD ARSENAL.]

[Illustration: CHLORINE PLANT, EDGEWOOD ARSENAL.

This is the largest single chlorine and caustic soda plant in

the country. Its capacity, when entirely completed, is 100 tons

of chlorine and 112 tons of fused caustic soda per day.]

[Illustration: CHLORPICRIN PLANT AT EDGEWOOD ARSENAL.

Bleaching powder, lime, and picric acid are received by rail.

In the mixers appearing in the right foreground lime, picric

acid, and water are mixed to form a solution of calcium picric,

and bleach and water are mixed to form a cream. These solutions

are pumped together into any of the several stills, where they

react to form chlorpicrin. This plant was rated at 12½ tons of

chlorpicrin a day, but reached a production as high as 31 tons

on one day.]

[Illustration: MIXER BUILDING OF PHOSGENE PLANT AT EDGEWOOD ARSENAL.

The capacity of this building is 20 tons of liquid phosgene per

day. Dry 98 per cent gaseous chlorine, as obtained directly

from the cells of the chlorine plant, and pure carbon monoxide

obtained from the producers, are mixed in approximate equal

volumes and the mixture passed through catalyzers, where the

two gases combine to form phosgene. The resultant gas is

liquefied in the condensers, appearing in the left.]

[Illustration: CHLORINE PLANT, EDGEWOOD ARSENAL.

One of eight cell rooms, capacity 12½ tons gaseous chlorine per day.

Each cell room consists of six circuits--74 cells per circuit, or a

total of 444 cells per room.]

[Illustration:

IN THE FOREGROUND THE CHLORINE PIPE LINE FROM CHLORINE PLANT

PASSING TO CHEMICAL PLANTS IN RIGHT CENTER OF THE PICTURE. ON

THE EXTREME RIGHT THE MUSTARD GAS PLANT. IN UPPER LEFT AND

CENTER, VIEW OF FILLING PLANTS AND SHELL DUMPS.]

[Illustration: FILLING 1-TON CONTAINERS WITH PHOSGENE.

Each empty cylinder weighs 1,300 pounds and will contain 1,650 pounds

of liquid. The plant shown fills 25 cylinders per day.]

[Illustration: MACHINE FILLING 75-MILLIMETER SHELL WITH MUSTARD GAS.]

[Illustration: FILLING HAND GRENADES WITH WHITE PHOSPHORUS.

Empty grenades are first completely immersed in a shallow

hot-water bath, shown on extreme left in photo. In a tank that

is not shown in picture white phosphorus is melted under water,

and this molten phosphorus is pumped by a small centrifugal

pump into a system of distributing pipes. Through a flexible

tube and by hand, each grenade is completely filled with

molten phosphorus, displacing the water in them. While the

grenades are still immersed in the water bath, a suction tube

is inserted in each grenade to remove the molten phosphorus

to a certain depth below the top of the grenade, this molten

phosphorus being displaced by water in the bath. The operation

shown in the photo depicts the grenades thus filled with molten

phosphorus to a definite heighth and with the remaining heighth

filled with water, having the water removed from the top of the

phosphorus by suction, after being taken out of the bath.]

[Illustration: FILLING MUSTARD GAS SHELL AT EDGEWOOD ARSENAL.

Inspected empty shell, as shown inverted on the left in the

foreground, are placed on small filling trucks, shown in the

right middle ground, and run under filling machine. Filled

shell with boosters screwed down leave the tunnel, as shown

in center of picture, where any possible mustard gas liquid

on them is vaporized by gasoline torch. A draft from this

operation into the tunnel prevents the distribution of mustard

gas vapor throughout the plant. The loaded shell are then

placed on trucks, as shown in foreground of photo.]

[Illustration: FILLING LIVENS DRUMS AT EDGEWOOD ARSENAL.

This photo shows the Livens drums being filled with phosgene.

The range of this special type of projectile, known as the

Livens drum, is about 1,500 yards. Its empty weight is about 30

pounds, and it contains a charge of about 30 pounds of gas.]

[Illustration: PAINTING AND STRIPING FILLED SHELL AT EDGEWOOD ARSENAL.

After leaving the filling plants, shell are classified by

weight into four groups and each group maintained separately.

The shell then are stored in an inverted position to detect

leaks. After testing for 24 hours, the shell are buffed,

painted, and striped by spray painting, as shown on the

endless conveyor, and then are ready for packing. In the left

background will be noted Livens drums being similarly painted.]

[Illustration: SHELL DUMP AT EDGEWOOD ARSENAL.

This picture shows filled shell being stored for leakage test before

being painted.]

[Illustration: FILLED CONTAINERS OF PHOSGENE, READY AT EDGEWOOD ARSENAL

FOR OVERSEAS SHIPMENT.

Each container holds approximately 1 ton of liquid.]

[Illustration: PHOSPHORUS CLOUDS FROM BURSTS OF 75-MILLIMETER SHELL AT

LAKEHURST, N. J., PROVING GROUNDS.]

[Illustration: GAS CLOUD FROM 4.7-INCH GAS SHELL EXPLODING 8,533 YARDS

AWAY FROM THE GUN AT LAKEHURST.] America's whole supply of chlorpicrin

during the war came from the American Synthetic Color Co. and the

Edgewood Arsenal. The Stamford plant was the first to reach large-scale

production.

The contract with the American Synthetic Color Co. was dated December

13, 1917; and the company shipped over 111,853 pounds of the gas to

Edgewood on March 11. This, when mixed with the necessary stannic

chloride, supplies of which were already on the ground, was sufficient

to fill approximately 100,000 75-millimeter shell. In the spring of

1918, due to certain internal troubles at the Stamford plant, it was

agreed that the United States should lease this factory and operate it

as a Government plant. Under Government operation the total production

of chlorpicrin at the Stamford plant amounted to 3,226,000 pounds, of

which 2,703,300 pounds were shipped overseas in 660-pound drums.

The chlorpicrin plant at Edgewood went into entire operation on June

14, 1918. Up to the signing of the armistice this plant had produced

2,320,000 pounds of chlorpicrin.

Phosgene was one of the deadliest gases employed in the war. Numerous

other gases were used to annoy the enemy and force the wearing of

masks, but phosgene was a killer employed to produce as many casualties

as possible. The gas did not persist long in the air or on the ground

after the shell had exploded, so that it was an ideal chemical for use

in an attack. The gas would clear away by the time the troops following

reached the place of gas concentration.

Phosgene at ordinary temperatures is a colorless gas, but it condenses

to a liquid at 8° C. It is formed by the combination of two gases,

chlorine and carbon monoxide, in the presence of a catalyzer. The

reaction is best conducted in iron boxes lined with lead and filled

with charcoal of proper quality, into which boxes a stream of the

reacting gases, mixed in proper proportions, is introduced. The

reaction creates heat, and means must usually be taken to keep the

reaction boxes cooled. The resulting phosgene is condensed to a liquid

by passing the gas through a condenser which is surrounded by brine

kept cold by refrigeration. The liquid is then stored in strong steel

containers or run directly into Livens drums or artillery shell.

Prior to 1917, the Oldbury Electro-Chemical Co., of Niagara Falls, N.

Y., had set up a small experimental phosgene plant in the hope that the

experiments might lead to the commercial utilization of carbon monoxide

which was obtained by this company as a by-product in the manufacture

of phosphorus. When we entered the war the company had developed its

process to such efficiency as to warrant the construction of a large

phosgene plant, and the Government entered into a contract with the

company for the creation of facilities with a capacity of 10 tons of

phosgene per day. Also, because of the great importance of phosgene in

warfare, it was decided at the same time to build a Government phosgene

plant at Edgewood. A little later the Government financed a phosgene

plant at the factory of Frank Hemingway (Inc.), at Bound Brook, N. J.

The total output of the original small experimental plant at Niagara

Falls, which was later leased by the United States, was 83,070 pounds

of phosgene, of which 24,800 pounds were shipped overseas. The contract

with the Oldbury Chemical Co. for its main phosgene plant was signed on

January 15, 1918. Production here began on August 5 and by August 20

had reached a daily average of 5 tons. On November 1 the average daily

production was 7 tons. The total quantity produced at this plant was

435 tons. The plant loaded 18,768 Livens drums with phosgene, each drum

holding about 30 pounds. This plant was operated by enlisted men.

The contract with Frank Hemingway (Inc.) called for a factory producing

5 tons of phosgene per day by a secret process controlled by the

company. The construction of the plant was begun on February 2, 1918,

and phosgene was first manufactured on May 17. This concern reached its

maximum of 5 tons per day by August 1, and produced in all 205 tons of

phosgene.

Construction of the phosgene plant at Edgewood was begun on March 1,

1918. The plant consisted of four catalyzer buildings, each building

having four units, each unit possessing a projected capacity of 5 tons

of phosgene per day. The total capacity, therefore, was designed to be

80 tons per day. The carbon monoxide used in the process was produced

by passing a mixture of oxygen and carbon dioxide over heated coke in

a gas producer, the oxygen being supplied by a Claude machine with a

capacity of 100,000 cubic feet of oxygen every 24 hours. The chlorine

used came partly from the Edgewood chlorine plant and partly from

outside sources.

The actual production of phosgene at Edgewood began on July 5, 1918,

and worked up to an output of 20 tons per day by the date of the

armistice. The total production of phosgene at Edgewood was 935 tons.

The total output of phosgene from all three plants, Edgewood and the

Bound Brook and Niagara Falls operations, at the date of the armistice

was 35 tons per day; and this was increasing to reach 95 tons per day

by May 1, 1919. The total phosgene produced by all the plants before

the armistice was 1,616 tons.

The Germans, in spite of their attainments in chemistry, were never

able to improve their clumsy and expensive methods of producing mustard

gas. The best reports we have show that at the time the fighting ended,

all of Germany's chemical warfare facilities could not produce more

than 6 tons of mustard per day. The United States alone had ten times

that capacity on the same date, while France and England both reached a

heavy output. So concerned was the German high command because of the

fact that Germany was being outdistanced in the production of mustard

gas that the ablest spy of the German Empire was sent into France in

October, 1918, to find out the French method of making mustard. One of

the Chemical Warfare officers who accompanied our forces into German

territory reported that the Germans had decided to adopt the American

method of making mustard gas and to stop their former process.

Mustard gas was by no means a child of the great war, having been

prepared in experimental quantities since 1886. It is a colorless,

slightly oily liquid, boiling at 220° C. with some decomposition. When

perfectly pure it freezes at 14° C.; but, since it usually contains

small percentages of impurities, it usually remains liquid at 0° C, or

even below that. In chemistry the substance is known as dichlorethyl

sulphide.

The first commercial process proposed for the manufacture of mustard

gas depended upon the use of ethylene chlorhydrin; and on April 13,

1918, a contract was made with the Commercial Research Co., Flushing,

Long Island, for the manufacture of 10 tons per day by this process.

In the spring and summer of 1918 a new process was developed both

abroad and in the United States, one which used sulphur monochloride.

Accordingly, the contract with the Commercial Research Co. was

canceled, and efforts were concentrated on the later process.

This process consisted in blowing gaseous ethylene into liquid sulphur

monochloride in large iron reaction vessels. The reaction developed

much heat. Sulphur is set free by this reaction, and the temperature

must be controlled in order to prevent the formation of solid sulphur

in the reaction machine.

At the date of the armistice three mustard gas plants were either

completed or nearing completion. The construction of the Edgewood plant

was begun on May 18, 1918, and the first mustard was produced exactly

a month later. The changing of processes, however, hampered production

somewhat, but by September 20, the arsenal was producing 10 tons per

day, and by November 11 had increased this to 30 tons per day. The

total production of mustard gas at Edgewood during the war period was

711 tons, of which approximately 300 tons went into shell.

On July 8, 1918, the Government began the construction of a mustard gas

plant at Hastings, N. Y. This factory was to have a capacity of 25 tons

per day, afterwards increased to 50 tons per day. The first unit of

this plant was ready to operate when the armistice was signed.

On July 6, 1918, the Government signed a contract with the National

Aniline & Chemical Co., Buffalo, N. Y., calling for a mustard gas plant

with a capacity of 50 tons daily. On November 11 this plant was 80 per

cent complete. The cost of the plant was met by the Government, but

the operation was to be in the hands of the Buffalo concern. The total

daily capacity of all three plants when complete was estimated to be

200 tons.

To insure an adequate supply of sulphur monochloride for its mustard

gas production the Government built a special plant at Edgewood with a

capacity of 300 tons of sulphur monochloride per day.

As soon as toxic gas warfare had developed to a considerable extent,

the perfection of gas-absorbing masks had given almost a complete

protection against this new weapon, if the soldier put on his gas mask

in time. But the mask, especially the earlier forms of it, was not easy

upon the wearer, due to the difficulty of breathing through it and also

because it restricted the soldier's vision. It was soon discovered

that a force compelled to wear its gas masks for any considerable

period lost in efficiency. The employment of gas by both sides for the

purpose of forcing the opposite sides to wear masks continually was an

important element in war at the close of hostilities.

For this purpose the so-called tear gases were produced. Gassing the

enemy with tear gas was much cheaper than with poison gas, yet it

forced him to remain masked. The tear gases were highly effective.

Even a trace of tear gas in the air would in a few moments blind a man

temporarily. A single tear-gas shell could force the wearing of masks

over an area so wide that it would require from 500 to 1,000 phosgene

shell to produce the same effect.

Most of the tear gases had bromine bases; so it was early determined

that we should have to increase the American supply of bromine

considerably if we were to meet our gas-warfare requirements. Bromine

is a deep red liquid which boils at 63° C. The domestic source of

bromine is principally in certain subterranean brines found in the

United States, these solutions containing bromine in its compounds. The

brines obtained in the vicinity of Midland, Mich., are especially rich

in bromine, and by far the largest amount of bromine obtained in this

country comes from that locality.

In December, 1917, at a conference with Mr. Dow, of the Dow Chemical

Co., Midland, Mich., it was decided that the Government should finance

the sinking of 17 brine wells near Midland, the Dow Chemical Co. to

supervise the work and to produce the bromine from the brine. The work

on this project was not begun until March, 1918, but the entire project

was practically completed when the armistice was signed. This plant is

a future war asset of the United States. It is capable of yielding

approximately 650,000 pounds of bromine per year.

The tear gas which we prepared to manufacture was brombenzyl cyanide.

It is a brownish oily liquid which solidifies to white or brownish

crystals at 29° C.

The production of brombenzyl cyanide involves a fairly intricate

chemical process. The first step is to chlorinate ordinary toluol, one

of the coal tar bases, to produce benzyl chloride. This chloride is

then mixed with sodium cyanide in alcoholic solution and distilled,

benzyl cyanide being the result. It is then only necessary to brominate

the benzyl cyanide by treating it with bromine vapor.

The first manufacture of brombenzyl cyanide in the United States was

conducted at an experimental plant at the American University Station

at Washington. After this a large scale plant was authorized at the

plant of the Federal Dye & Chemical Co., at Kingsport, Tenn. The

construction of this factory began on July 8, 1918, and operations

started on October 29, the total production of brombenzyl cyanide being

a trifle over 5 tons. In November the plant reached a capacity of 3

tons per day.

The bromine gases were not poisonous in the sense of being killers,

but were merely highly irritating to the membranes of the eye. The

killing gases were phosgene, chlorpicrin, and chlorine. Mustard gas in

sufficient amount was also fatal, its effect being identical to that

of a deep burn. It attacked the lungs, the eyes, the skin, and even

the intestines if food contaminated with mustard gas were swallowed.

An insidious feature of mustard gas is the fact that its action is

practically always delayed. It might be several hours after a man was

gassed, even fatally, with mustard before he became aware of it, and

then it was too late to administer the treatment that might save his

life. Goggles alone would have been sufficient protection against tear

gas, except for the fact that it was invariably mixed with the deadlier

gases.

The various experiments preliminary to our production of gases were

conducted in provisional laboratories at the Bureau of Standards,

Washington, D. C., Bureau of Mines, Washington, D. C., the Geophysical

Laboratory, Washington, D. C., the Ohio State University, Columbus,

Ohio, and Johns Hopkins University, Baltimore, Md. A control

laboratory for the solution of problems arising in manufacture was

eventually established at Edgewood. A total of 167,092 single chemical

determinations were made at these laboratories under the direction of

20 commissioned officers, 45 noncommissioned officers, and 204 privates.

The production of gases and other chemicals was only part of the work

of the Edgewood Arsenal and its subsidiary plants. The other chief

activity was that of filling artillery shell with the toxic substances.

The description of the plant which filled shell with phosgene will

indicate the scale upon which this operation was conducted.

The empty shell, after being inspected, were loaded on trucks, together

with the proper number of loaded boosters. The booster was the device

which exploded the shell and scattered the gas. Electric locomotives

then pulled the shell trucks to the filling buildings. There were four

of these to a single shell-filling plant, radiating at right angles

from a common center. From the trucks the empty shell were lifted by

hand to a belt conveyor and the conveyor carried the shell slowly

through a room kept cold by artificial refrigeration. Although the

shell moved only 70 feet through this room the conveyor traveled so

slowly that they were 30 minutes in transit, and during this time they

were cooled to a temperature of about 0° F. This chilling was necessary

because phosgene has a low boiling point, and it was necessary to

keep the temperature of the metal of the shell considerably below the

boiling point of phosgene in order that the gas might remain in liquid

form while the filling was going on.

The chilled shell cases were next transferred to small trucks, each

carrying six of them. The loaded truck was then drawn through a filling

tunnel by means of a chain haul. This tunnel was so ingeniously

contrived that the human assistance to the filling and closing

machinery could all be conducted from the outside. The phosgene, kept

liquid by refrigeration, was run into the shell by an automatic filler.

The truck was then moved forward a few feet to a point where the

boosters were inserted into the noses of the shell by the hands of

the operator reaching in through an aperture in the tunnel. The final

closing of the shell was then accomplished by motors. The air in

the filling tunnel was constantly withdrawn by strong ventilation,

the exhaust air being washed in stone towers by chemical agents to

neutralize any gases that might be present. The filled, inclosed shell

were next conveyed to a dump, where they were classified and then stood

nose down for 24 hours to test them for leaks. Then they were painted,

striped, and stenciled by air paint brushes. The final process was to

pack them in boxes and store them for shipment. This was done in large

storage magazines on the grounds of the Edgewood Arsenal.

A similar method was used for filling shell with chlorpicrin, except

that refrigeration was unnecessary. Mustard gas required another sort

of filling machine.

Several filling plants were designed and constructed for filling

grenades with stannic chloride and with white phosphorus, and also one

for filling incendiary drop bombs.

The capacity of each of these plants per day was as follows:

Stannic chloride plant, hand grenades, 25,000. White phosphorus grenade

plant, 30,000. White phosphorus smoke-shell plant, 155 millimeter

shell, 2,000; or 4.7-inch or 5-inch shell, 4,000; or 75-millimeter

shell, 6,000. Incendiary drop-bomb plant, 2,000.

The following sentences summarize the production and expectations of

the Edgewood Arsenal:

(1) The gas program as of March 1918 called for approximately 545 tons

of toxic gas weekly.

(2) The Chemical Warfare Service program of August 12, 1918, called for

a much larger amount, viz, about 4,525 tons per week.

(3) The approximate filling capacity of the Edgewood Arsenal plant from

August to November, 1918, was nearly 1,000 tons per week.

(4) The toxic gas production during this same period increased from 450

to 675 tons per week.

(5) The capacity of all projectiles received, unlimited by boosters,

varied during the same period from 125 to 450 tons per week.

(6) The maximum capacity corresponding to boosters received was less

than 100 tons per week.

In these facts it will be seen that the numbers of empty shell

delivered to the plant was far less than the number required to

accommodate the gas production. Many of the shell received were without

boosters and therefore without value until boosters were provided,

so that the limiting factor was really the supply of boosters. The

booster supply was sufficient to take care of only a relatively small

fraction of the toxic gas actually produced. The filling capacity of

the plant was also in excess of the delivery of shell and boosters. The

75-millimeter shell-filling plant had a capacity of 1,200,000 shell per

month, eventually double that, while delivery of shell was slightly

over 300,000 per month and of boosters less than 200,000.

Because of the nature of toxic gas it is impossible to store it up in

any large quantities. Early in the summer of 1918 large amounts were

shipped in bulk overseas and there loaded into shell. Later we received

instructions to stop all shipments in bulk except a limited amount of

chlorine, and thereafter our production was limited to the number of

shell and boosters available.

In June, 1918, we shipped in bulk 15 tons of mustard gas, 705 tons of

chlorpicrin, and 48 tons of phosgene. This was to be exchanged for

gas shell produced by the French. In late July the French had no more

extra shell to be filled with American gas and this fact terminated

the arrangement. However, we sold excess gas both to England and to

France. England received 900 tons of our chlorpicrin and 368 tons of

American phosgene. France took 300 tons of chlorpicrin and 1,408 tons

of chlorine, equivalent to 1,226 tons of phosgene, since phosgene is

80 per cent chlorine including allowance for wastage in manufacture.

France furnished phosgene shell to us in exchange for chlorine. In

addition 200 tons of mustard gas were shipped to England and utilized

by the English.

We therefore shipped to Europe in bulk 3,662 tons of gas or its

equivalent, which gas was largely loaded in shell and used by the

United States troops or those of the allies. This quantity was

sufficient to load 1,600,000 shell, two-thirds of them being of the

75-millimeter caliber and the other one-third 155-millimeter, the total

number being thought to be at least equal to the total number of gas

shell fired by American troops in action. Thus while American gas was

not actually fired in American shell against the Germans, American gas

was used against the enemy and America furnished at least as much gas

as she fired.

In addition to this we shipped 18,600 Livens drums loaded with

phosgene. These contained 279 tons of gas, and some of them were fired

at the enemy. We began producing loaded gas shell in the summer of 1918

and by August 9 had shipped 75,000 loaded 75-millimeter shell. These

shell were unassembled for firing in the guns, the Ordnance Department

having decided in June to assemble gas shell in their cartridge cases

in France.

The Chemical Warfare production organization developed and manufactured

a large number of special containers for the shipment of toxic gases.

These were of special construction in order to guard against dangers

that would result from leaks, and all had to stand the tests required

by the Bureau of Explosives before they would be received for railroad

shipment. The 1-ton containers, all of which would hold 1 ton of liquid

chlorine, were designed by the Ordnance Department and would withstand

a pressure of 500 pounds per square inch. The 300-pound phosgene

cylinders, designed by the Ordnance Department, were made to withstand

a 500-pound hydrostatic pressure and a 250-pound air test.

We purchased standard 55-gallon acid drums and standard-pattern

cylinders for holding 75 pounds of chlorine.

We constructed chlorine tank cars, each tank with a capacity of 15 tons

and a strength that could withstand a pressure of 500 pounds to the

square inch. We also designed a tank car originally for the shipment of

chlorpicrin and later used it for shipping sulphur monochloride.

CHEMICAL WARFARE PRODUCTION DATA.

_Production of toxic material._

[All figures are for tons of 2,000 pounds.]

---------+-------+-------+-------+-----+--------------+-------+------+-----

| | | | | | | | ide |ide.

---------+-------+-------+-------+-----+-----+--------+-------+------+------

January | | | 10 | | | | | |

February | | | 27 | | | | 34| |

March | 40| | 59 | | | | 74| 38|

April | 176| | 33 | 15| | | 59| 116|

May | 378| | 130 | 18| | | 70| 51| 50

June | 546| | 263 | 23| 6| | 60| 95|

July | 512| | 499 | 100| 21| | 80| 112| 27

August | 243| | 646 | 314| 36| | 162| 94| 53

September| 438| 191| 564 | 327| 144| | 125| 96| 26

October | 242| 649| 445 | 664| 361| | 265| 75| 25

November | 148| 264| 100 | 155| 143| 5 | 77| 18|

| | | | | | | | |

+-------+-------+-------+-----+-----+--------+-------+------+------

Total | 2,723| 1,104| 2,776 |1,616| 711| 5 | 1,006| 695| 181

| | | | | | | | |

Amount | 1,488| | 1,903| 420| 190| | 171| 106|

Total | 895| 1,500| 1,500|1,050| 900| 90| 100| 91| 30

Nov. 1, | | | | | | | | |

1918 | | | | | | | | |

Estimated| 1,100| 2,250| 1,500|1,650|4,000| 90| 100| 91| 30

Jan. 1, | | | | | | | | |

1919 | | | | | | | | |

---------+-------+-------+-------+-----+-----+--------+-------+------+------

[33] Procured from commercial agencies.

[34] Manufactured at Edgewood.

_Shell, grenades, Livens drums, and drop bombs filled._

---------+----------------------+-----------------+--------+----------------

| 75-mm. shell. | Grenades. | Livens |Incendiary drop

| | | drums. | bombs.

1918 +-------+-----+--------+-------+---------+--------+-------+--------

---------+-------+-----+--------+-------+---------+--------+-------+--------

July | 62,866| | | 8,696| 1,639| | |

August |125,951| | |170,160| 56,763| 1,738| 350|

September|110,358|1,988| 75,529| 51,421| 127,319| 6,355| | 1,998

October | 09,704| 12| 79,272|110,928| 147,669| 12,026| 184| 100

November | 15,892| 9| 224| 98,948| 30,386| 5,570| 8| 6

+-------+-----+--------+-------+---------+--------+-------+--------

Total |424,771|2,009| 155,025|440,153| 363,776| 25,689| 542| 2,104

| | | | | | | |

Total |300,000| | 150,000|224,984| 175,080| 18,600| |

---------+-------+-----+--------+-------+---------+--------+-------+--------

_Total monthly capacity of filling plants on date of armistice

(Stokes shell, drop bombs, and other special containers not

included)._

75-mm. shell (ultimate capacity) 2,400,000

4.7-inch shell 450,000

155-mm. shell 540,000

6-inch shell 180,000

Gas grenades 750,000

Smoke grenades 480,000

Livens drums 30,000

CHAPTER II.

GAS DEFENSE EQUIPMENT.

During the spring and summer of 1917 two marked tendencies were to

be observed in the fighting in France. One of these was the greatly

increased use by both sides of poisonous gases and chemicals, frightful

in their effect; the other the almost complete censorship that hid

the knowledge of this tendency not only from the people of Europe but

particularly from those of the newest belligerent, America. The French

and British Governments, who then controlled all news from the front,

feared, and perhaps with reason, that if the picture of gas warfare, as

it was then developing, should be placed before the American people,

it would result in an unreasonable dread of gases on the part of the

American Nation and its soldiers.

One year later, with tens of thousands of American troops facing the

Germans, there was almost no censorship upon the details of fighting

with chemicals. The mysterious gases of 1917 were then known to almost

every reading individual in the civilized world. The once secret

formulas were published in the technical journals. Non-censored

photographs of defensive equipment were freely published, and masks and

other paraphernalia were exhibited for the public interest. Except for

secret plans for the future and the various surprises being prepared by

one or more of the belligerents, the whole subject of chemical warfare

had become an open book.

What occasioned this change in policy on the part of governing

authorities? The reason was that the American troops brought with

them to France the best and most protective gas masks the world had

seen; and they brought these with them by the millions. Starting a

mask-production effort in May, 1917, America turned out a total of

5,250,000 gas masks before the armistice was signed, and sent more

than 4,000,000 of them overseas. As to the quality of these masks, it

is only necessary to say that they gave twenty times the protection

afforded by the best German gas masks. In other words, we protected our

soldiers against the poisons which Germany had brought into warfare,

and protected them completely. No American soldier was ever gassed due

to the failure of an American gas mask, and such gas casualties as

did occur were due to the fact that the masks were not quickly enough

utilized when gas was thrown over, or because the soldier was unaware

of the presence of gas. With such protection there was no longer

reason to fear that the frightfulness of chemical warfare would reduce

American morale.

The production of gas masks was one of the most picturesque and

successful phases of our entire war preparation. It engaged the

attention of some of the principal chemical engineers of the country,

and millions of men, women, and children in the United States

contributed something to the success of the undertaking, if only to

obey the "Eat More Coconut" slogan or to save peach stones for the

benefit of the production of the charcoal essential to efficient gas

masks.

The problem of making masks in such quantity and under such supreme

demands for perfection was one which might well stagger manufacturers

accustomed to large-scale operations. We started in with practically

no knowledge whatsoever of the fundamental principles of a perfect

mask. Yet the apparatus was as difficult to build as a rifle. It must,

perforce, be made of perishable materials, and this fact brought the

question of durability to the fore at the very start. It was evident

that no chemical substances known in our past commercial life would

give protection against the new poisons which had been developed in

Europe. With the exception of phosgene and chlorine, the various war

gases which had been brought out prior to our entrance in the struggle

were completely unknown in our trade or commerce and had existed only

in our experimental laboratories. Then it was discovered that as these

toxins increased in power they could penetrate the ordinary fabrics

known in commerce, and this necessitated the creation of new types

of materials to be used in the masks. Finally the increasing use of

gases forced the soldiers to wear their masks for much longer periods

than had been necessary at the beginning of gas warfare; so that

the problem of comfort became one of great importance. All of these

basic considerations indicate to some extent the difficulty of the

undertaking.

The chlorine, which floated in a pale greenish-yellow cloud down upon

the defenseless Canadian troops at Ypres, with such terrible effect

upon the men, was, as has been said, the first gas used. Chlorine,

though easy to obtain, the principal source of supply being common

table salt, was, from the standpoint of strategy, far from being the

ideal gas of warfare. Troops could be quickly and easily protected

from it. But even as it was, only lack of faith in their new weapon

prevented the Germans from winning the war with it then and there. Had

they brought into the fighting a sufficient supply of this chlorine,

they might have gassed their way to Paris in short order. In fact, they

brought to the line an almost negligible supply and they themselves

were insufficiently protected to go through their own gas and follow up

the attack. By the time they were able to renew gas warfare the French

and British had equipped themselves with masks which were sufficient

to protect men against chlorine.

Thereafter the tendency was toward new and strange gases which were

heavy in weight and highly toxic in their physiological action. This

development led to new, slightly volatile liquids, the so-called

mustard gas being the best example. Mustard gas (properly called

dichlorethyl sulphide) is similar to lubricating oil in many of its

physical characteristics but smells like ordinary mustard. Ground

soaked with the mustard gas remains impregnated for days, the vapor

rising continually.

A perfect mask is one which will remove completely every trace of gas

or poisonous vapor before the air can reach the eyes, nose, or mouth of

the soldier.

The first masks adopted by the allies were simply gauze pads saturated

with neutralizing chemicals. These became unsuitable as soon as new

varieties of powerful poisons were brought out. The mask development

thereafter progressed to the box respirator type. This consisted

of a mask or helmet connected to a box filled with absorbing and

neutralizing chemicals which purified the air for the mask wearer. This

was the type of respirator in use to the end of the fighting.

It is quite clear to us now that only such a mask could be efficient

in chemical warfare, but in the early part of 1917 the matter was not

clear either to us or to the allies. The first requisitions from the

A. E. F. called for masks of two types, each soldier to be supplied

with one of each. The reserve mask was to be of the gauze type and the

regular mask of the box respirator type, affording protection from

the more powerful poisons that were then just coming into use. We

wasted considerable energy at the beginning in our attempt to produce

both types. Eventually, however, when we were just ready to start

manufacturing the gauze-type mask, orders came to abandon the effort,

since it was even then apparent that our soldiers must be prepared at

all times to withstand all gases.

The box respirator equipment, the general principle of which was

finally adopted by all the nations at war, fell into two classes. In a

single-protection mask the wearer breathed air from inside of the face

piece, so that any leakage around the edges of the face piece would

result in a casualty when the wearer was in a strong concentration of

gas. The other sort, known as the double-protection mask, consisted of

a gas-tight face piece, similar to that of the single-protection mask.

In this type, to guard against any possible leakage around the edges

between the mask and the wearer's skin, the breathing system was sealed

away from the air inside the face piece by means of a rubber mouthpiece

and a nose clip, the wearer inhaling through the mouthpiece.

[Illustration: VARIOUS TYPES OF GAS MASKS.

Top row, left to right.--First type U. S. Navy mask, now

obsolete; U. S. Navy mask as finally developed; U. S. C. E.

respirator (production started October, 1917); U. S. R. F. K.

respirator (production started February, 1918); U. S. A. T.

respirator (production started August, 1918); U. S. K. T. mask

(production started August, 1918); U. S. model 1919 mask (ready

for production when armistice was signed).

Middle row, left to right--British black veil mask (first

mask used after initial gas attack in April, 1915); British

P. H. helmet (stops phosgene but not tear gases); standard

British box respirator used by all British forces after 1916;

French M-2 mask used by the French; French Tissot mask used by

artillerymen; French A. R. S. mask.

Bottom row, left to right.--Late type of German mask;

Experimental mask; Italian mask (similar to French M-2 mask);

British Motor Corps respirator; U.S. rear area emergency

respirator; U. S. Connell mask (never passed the experimental

stage).]

[Illustration: AMERICAN C. E. TYPE OF BOX RESPIRATOR.

This side view shows face piece, harness, hose, flutter valve, and

knapsack. This is the mask most used by our troops.]

The United States and English double-protection masks consisted of 11

principal parts as follows:

1. A knapsack slung from the shoulder or neck. This contained the

canister and a pocket for storing away the mask when not in use.

2. A metal canister in which was contained the absorptive neutralizing

chemicals.

3. A flexible hose reaching from the canister to the face piece.

4. A flutter, or exhalation, valve, which opened when the wearer

exhaled his breath and closed when he inhaled, thus bringing the

inhalation through the canister but allowing the exhalation from the

lungs to pass out without polluting the chemicals of the canister.

5. The face piece, or hood, fitting snugly around the edges and

covering the eyes, cheeks, lower forehead, nose, mouth, and chin.

6. The eyepieces, or lenses, through which vision was maintained.

7. An elastic harness for the head, to hold the face piece in place.

8. A body cord to tie around the chest and hold the knapsack firmly,

so that the mask could be seized in both hands and pulled out of the

knapsack.

9. A metal flange connection or angle tube which carried the hose

through the face piece to the mouthpiece.

10. A rubber mouthpiece through which the wearer breathed and which

helped to hold the mask in place.

11. A wire nose spring and rubber nose pad to hold the nostrils shut

and force breathing through the mouth.

The first order for gas masks was issued on May 16, 1917, when the

Chief of Staff asked the Surgeon General to supply 1,100,000 masks

before June 30, 1918, or within about one year. Meanwhile 25,000 masks

were needed at once in order to equip Gen. Pershing's first division,

then about to sail overseas. There was but one man in the Army who knew

anything at all about the subject and who could even attempt to produce

this quantity in three weeks. This was Maj. (later colonel) L. P.

Williamson, of the Surgeon General's Department, who had been spending

some months at the Army War College at Washington studying as a side

issue such papers on gas warfare as came from abroad. It was due to his

knowledge and the volunteer staff of the Bureau of Mines that we were

able to begin the actual manufacture of masks within a few days after

the requirements were fixed, and actually to turn out 25,000 masks in

but little more than three weeks' time.

Col. Williamson's first step was to consult with Dr. Van. H. Manning,

the Director of the Bureau of Mines, and with his assistant, Mr.

G. A. Burrell. Since February, 1917, the Bureau of Mines had been

experimenting with gas masks and had built up a corps of scientists for

this work. Within this organization was Mr. Bradley Dewey, a chemical

engineer, who, though then director of the research laboratory of the

American Sheet & Tin Plate Co., of Pittsburgh, had been loaned to the

Bureau of Mines. To Mr. Dewey was turned over the job of directing

the production of the first 25,000 masks for the American troops then

sailing.

To produce 25,000 gas masks in three weeks meant to compress England's

two years of experience into 21 days. The military authorities of this

country at that time could plead entire ignorance of the qualifications

of an efficient gas mask. The prevailing idea seemed to be that you

could go out into the market and buy them by the hundreds of thousands,

as you might buy Halloween masks. But this was not any ordinary poison

which we were to fight. These powerful chemicals attacked the human

tissues as would acid. As the result of the effort, we did supply

the first division going overseas in July. However, the masks were

inferior to the British and were quickly replaced in France by British

equipment. It was not until the following January that we developed an

apparatus which we regarded as satisfactory to undergo the supreme test

of battle.

To indicate some of the difficulties overcome between May and December,

1917, there are here set forth some of the features of an effective

mask.

In the first place, the face piece must fit perfectly; it must not leak

gas around the edges. It must fit into the hollows of the temples and

must give the jaws a free space in which to work, and yet not slip back

and press against one's Adam's apple. The pressure of the mask on the

forehead must come above the supraorbital nerves which are just above

the eyebrows, or else intense headaches will result from a few moments'

wear. Moreover, to fit all faces and heads, several graduated sizes of

masks are required. We first attained the gas-tight fit with a padded

band around the edge of a flexible rubber-cloth face piece. Later we

developed a thicker, stiffer face piece, but maintained a gas-tight fit

by the elasticity of the face piece and the head harness.

Then the material of the face piece must be gas-tight in itself. At

first we manufactured a fabric made by spreading rubber on cotton

sailcloth; and, after testing it, we found that the smallest molecule

known, that of hydrogen, would not pass through it in large amounts.

This seemed to be a suitable fabric, until tested by the newer gases.

Then we found that some of these gases were soluble in rubber compounds

and could dissolve their way through thin rubber so quickly that the

face piece cloth offered practically no protection at all. Another

difficulty with the rubber fabric was that it was likely to absorb and

hold certain of the poisons, so that a man might be gassed by the mask

itself. The rubber companies, principally at Akron, Ohio, experimented

continually until they discovered a coating that would not only

withstand gas concentrations for a sufficient time, but would also

aerate promptly and lose as much gas as it had absorbed.

The eyepieces or lenses offered another problem. Celluloid is strong

but it is not so transparent as glass. It ignites easily and is easily

scratched. Glass is ideal in transparency and will not burn, but is

fragile. It was evident that we must provide eyepieces which would not

break easily, since even so slight an accident as the breaking of a

lens might cost a soldier his life by admitting concentrated gas to the

mask. A material known as triplex glass had been experimentally made.

This consisted of a thin celluloid strip sandwiched between two layers

of glass, all three welded together. This glass would not splinter, and

even if cracked or broken, would still be gas-tight. However, this had

never been made in quantity and it was necessary to work out many kinks

and to start a large plant to provide the necessary millions of lenses.

Then there was also to be overcome the tendency of the eyepieces to

dim, particularly in cold weather, as the wearer breathed moist breath

into the mask. The answer to this problem was a soapy compound which

put a slippery surface on the glass and avoided the droplets of mist.

The first masks were also equipped with deep plaits so that the wearer

could wipe off the lens with the interior of the face-piece itself,

though the final development (the invention of a Frenchman by the name

of Tissot) was to bring the cold air into the mask so that it flowed

directly against the lenses and evaporated any condensed moisture. This

kept them clear under all ordinary circumstances.

It was evident that the metal tube passing through the face piece must

not contain pinholes and must be able to stand rough handling without

pulling loose. The harness must maintain a gas-tight connection between

the wearer's face and the face piece, but not at the cost of pain or

chafing of the face or head. The flutter valve must fit with absolute

tightness and must work perfectly and instantaneously at all times.

The flexible hose leading from the canister to the face piece must be

strong and without flaws or leaks, and yet flexible in the extreme.

A stiff hose would be likely to swing and displace the face piece

whenever the wearer moved. The mouthpiece must be comfortable and must

be built along lines to prevent irritation to the gums or lips, yet it

must be reinforced so that in his excitement the soldier can not bite

down and shut off his air supply.

The canister must withstand corrosion and must be gas-tight. Smooth

sided canisters can not be used, for the gas would slip up the sides

without coming in contact with much of the chemical filling. The sides

of the canisters were, therefore, ribbed so that the charcoal and

other ingredients working into these ribs baffled the gas and threw

it out into the body of the chemicals. The canister, moreover, must

be equipped with a perfectly working check valve which will stop

exhalation through the canister and force the air to pass out through

the flutter valve.

The web sling of the knapsack must not curl and chafe the neck or

shoulders of the wearer. The knapsack must be waterproof and must have

easily and quickly workable fastenings.

The canisters were filled with charcoal and with cement granules.

These were crushed into carefully sized small bits about the size of

a pinhead and packed in layers in the canisters. The air could pass

through them easily and the particles of both substances absorbed gas.

The chief quality requirements for the carbon and the cement were that

they must have long life and great activity.

Of the canister ingredients the charcoal offered the more difficult

technical problem. It had long been known that charcoal was highly

absorptive of certain gases, but except in rare instances no thorough

study had ever been made of the subject. It was evident, however, that

the more charcoal or carbon which could be packed into the canister and

still allow the free passage of air the greater the amount of gas that

would be absorbed. Consequently a search was made for carbon existing

in the natural state in the most compact form. This search is described

later.

Each canister also contained concrete granules in a definite proportion

to the carbon pieces. These granules were made of cement mixed with

strong alkalis and oxidizing agents to digest the poisons as they

passed through the canister.

It will be seen that the manufacture of good gas masks was a highly

technical undertaking, one calling for the best talents of eminent men

of science. The mask was not something that could be improvised on the

spur of the moment, but each part of it must be worked out after the

most painstaking research. The Gas Defense Division of the Chemical

Warfare Service never at any time approved a type of mask which

its own officers or men did not themselves wear in the most deadly

concentrations of gas.

To get back to the chronological order of development, on May 21,

1917, the making of the first 25,000 masks was started with frantic

haste; though, as it developed later, there was no need for such an

effort, since there were available in England and France plenty of

masks for the first American troops. Working to produce in the shortest

possible time some sort of protection for the first overseas division,

the officers in charge were forced to adopt methods which, had they

been followed throughout the manufacturing program, would have been

extremely costly. There was no time then to stop and study the problem

either here or abroad. Before the end of June 20,088 masks had been

started overseas, and 5,000 more were ready a little later. The most

that can be said for this effort was that it gave our officers the

experience which was the groundwork of the solid development later on.

The production of these first 25,000 masks called upon the services of

various manufacturers. The assembling of the masks was conducted by the

American Can Co., at Brooklyn, N. Y. The B. F. Goodrich Co., of Akron,

manufactured the face pieces with the eyepieces inserted, also the

connecting hose, the check valve of the canister, the flutter valve,

and the rubber mouthpiece. The American Can Co. produced the canisters.

The Day Chemical Co., of Westline, Pa., gave the charcoal its first

burning. The Ward Baking Co., of Brooklyn, patriotically baked the

charcoal--to activate it--in their bread ovens free of charge. The

General Chemical Co., of New York, supplied the soda-lime granules. The

Doehler Die Casting Co., of Brooklyn, manufactured the angle tubes. The

Simmons Hardware Co., of St. Louis, produced the waterproof knapsacks.

The Seaver Howland Press, of Boston, printed the cards of instructions

that went with the mask outfit; and the Beetle & MacLean Manufacturing

Co., of Boston, printed the record tags.

Though Maj. (now colonel) Williamson was formally in charge of this

emergency work, he requisitioned the masks from the Bureau of Mines,

which took entire charge of the first contract. Following this, on

August 31, 1917, the Gas Defense Service of the Surgeon General's

Department was established by official order, and Mr. Dewey, who had

been working as a volunteer in the Bureau of Mines, was commissioned

major and put in charge.

The next step was to prepare for the permanent development and

manufacture of gas masks. Contracts were let for the manufacture of

320,000 component parts of masks as we then knew them, and a price

was fixed for the assembling of the entire original requirement of

1,100,000 masks. The assembling contract went to the Hero Manufacturing

Co., of Philadelphia, which remained until the end of the war the sole

private contractor assembling our gas masks.

The spirit of cooperation and desire to serve the Government was

evident from the start. The B. F. Goodrich Co. had been the only

producers of the rubber parts of the first 25,000 masks. In this

original contract it had gained valuable technical and cost knowledge;

but in order that the Government might not be limited to one source of

supply for such parts, the Goodrich Co. voluntarily imparted to the

Goodyear Tire & Rubber Co. and to the United States Rubber Co. the

information that would enable them to bid intelligently for portions

of the work. This was a distinct departure from the usual practice in

competitive industry.

All during the fall of 1917 and early winter of 1917-18 the development

of the mask continued, the Government experts working hand in hand with

private contractors. Because of the newness of this sort of manufacture

and because of the wide variety of unusual articles required, entailing

in some instances the actual creation of hitherto unknown commodities,

the Government at all times was required to act as the procurer of raw

materials for the masks. In this period of development America designed

her own typical mask--a gradual evolution, but one which, though based

on the British design, arrived at a perfection which had been unknown

in warfare before.

The triplex glass used in the eyepieces was a patented commodity

produced only in one small factory in Philadelphia. It was necessary to

expand the facilities for the production of this necessary material.

Meanwhile some of the men engaged in the work had improved the eyepiece

by providing it with an aluminum mounting. But this very improvement

brought embarrassment to the work, since the Akron rubber contracts

had provided for eyepieces inserted in the fabric itself, and to apply

the aluminum frame brought about a radical change in the manufacturing

methods at the rubber factories.

There were also many other problems that had to be solved before our

authorities were satisfied to go ahead in quantity production. There

was the matter of rubberizing the face-piece fabric, for instance. Two

methods of rubberizing cloth were in use. The first method was to roll

out a thin sheet of rubber and then press it into the cloth fabric by

running the whole thing under heavy rollers. This was known as the

calender method. The other method, called the spreader method, was

more intricate. In this process the sailcloth, tightly stretched, was

carried around a roller. Above the roller a few thousandths of an inch

was a knife blade extending from edge to edge. The rubber compound in

liquid form was then fed upon the roller in such manner that a thin

film of it pressed under the knife blade and upon the cloth on the

roller. The rubberizing method finally adopted was a combination of

the calender and spreader methods. The rubber was applied green to

the cloth. The curing process thereafter was highly important. If the

curing process were too short, the rubber would be sticky and would

pull off the sailcloth too easily. If the rubber were over-cured, it

would crack and split.

Nothing short of absolute perfection in every part would do, since the

slightest imperfection anywhere was likely to cost a man his life.

Consequently we installed at the various producing plants not only 100

per cent inspection, but we constructed laboratories for putting the

materials through the most elaborate and exhaustive sorts of control

tests, and then reinspected the parts at the assembly plants, both

before and after the assembly.

[Illustration:

CHEMICAL DEVELOPMENT DEPARTMENT OF LONG ISLAND LABORATORY, GAS

DEFENSE DIVISION, SHOWING INTERMITTENT FLOW CANISTER TESTING

MACHINE.]

[Illustration:

PHOTO TAKEN AT PHILADELPHIA CONTROL LABORATORY OF GAS DEFENSE

DIVISION, SHOWING APPARATUS USED IN EXPERIMENTAL WORK ON THE

EFFECT OF RESISTANCE TO INHALATION AND EXHALATION OF MASKS.]

[Illustration: SIDE VIEW OF GAS CHAMBER AT CHEMICAL DEVELOPMENT

LABORATORY, SHOWING SUBJECTS ON OUTSIDE BREATHING TESTS.]

[Illustration: ANOTHER VIEW OF GAS CHAMBER AT CHEMICAL DEVELOPMENT

LABORATORY, SHOWING SUBJECTS INSIDE.]

All the rubber used was continually sampled and analyzed in the

laboratories. The tensile strengths of all fabrics were determined by

standard destructive tests. We also tested the adhesion of the rubber

coating by standard chemical methods and worked out flexibility tests

for the breathing tube.

After all of the factory inspection and material-control tests, the

masks themselves were sampled and worn in highly toxic atmospheres. In

this work thousands of our masks were worn by the officers and men of

the Gas Defense Division in concentrated atmospheres of the most deadly

gases. For such work we constructed testing rooms whose atmosphere

could be completely exhausted and changed in 90 seconds. The efficiency

of canisters was tested either by the lungs of the inspectors or by

mechanical breathing into telltale solutions.

The story of the carbon (charcoal) which went into the American

canister is one of the most interesting phases of the whole

undertaking. Investigations carried on by the research staff of the

National Carbon Co., aided by a clue from the University of Chicago,

led to the selection of coconut shell as a raw material. Any carbon

absorbs a definite number of times its weight of gas. Therefore the

densest carbons will be most efficient, volume for volume, as gas

absorbers in a given space. Coconut shells and other nut shells were

found to be the most compact form in which carbon exists in nature

in commercially practicable quantities, being considerably superior

in this respect to anthracite coal and to such woods as ironwood and

mahogany. Another essential for charcoal used in the canisters was that

it must be so hard that it would not crumble easily and produce dust

that would clog up the air passages and prevent easy breathing through

the canister. Coconut shell fulfilled both of these conditions better

than any other known material.

Further study by the National Carbon Co., backed up by wonderful

large-scale development work, paid for and carried out by the National

Electric Lamp Association under the direction of their Mr. F. N. Dorsey

(who later became Col. Dorsey and chief of the Development Division of

the Chemical Warfare Service), gave us the details of a new process

for treating the charcoal to make it absorptive. After the original

burning of the nut shells, or other carbon materials, the resulting

carbon was given a second highly specialized heat treatment, and this

activated it until it had a powerful affinity for gas. Such carbon,

made from nutshell material, would absorb 150 times its own volume of

chlorpicrin, one of the most deadly of the war gases, the action being

approximately instantaneous.

It must not be supposed, however, that investigation of carbons

stopped with these experiments. In the search for the ideal carbon

we experimented with almost every hard vegetable substance known.

Literally, hundreds of kinds of carbon were tested. Next to coconut

shells, the fruit pits, several common varieties of nuts abundant

in the United States, and several tropical nuts, were found to make

the best carbon. Pecan nuts, and all woods ranging in hardness from

ironwood down to ordinary pine and fir, were found to be in the second

class of efficiency. Among other substances tested were almonds,

Arabian acorns, grape seeds, Brazil-nut husks, balsa, osage oranges,

Chinese velvet beans, synthetic carbons, cocoa bean shells, coffee

grounds, flint corn, corn cobs, cottonseed husks, peanut shells and

oil shale. While many of these substances might have been used in an

emergency, none of them would produce carbon as efficient, volume for

volume, as that of the coconut shells and other hard nuts.

Some idea of the scale of the American mask production may be seen in

our requirements for coconut shells. In our survey of raw materials we

included the entire coconut resources of the world. Such figures were

relatively easy to obtain because the copra, or dried coconut meat,

industry is an important one, particularly in southern Asia and the

South Sea Islands of the Pacific. Ceylon was the greatest single source

of coconuts, 2,300,000,000 nuts being gathered there annually. British

India was next with 1,500,000,000 nuts. Our own Philippine Islands were

third, with an annual production of 900,000,000 nuts. Then followed in

order the Dutch East Indies, British Malaya, French Indo-China, Siam,

and the Pacific archipelagos, the total production of the Orient being

7,450,200,000 nuts annually. This was a supply that would provide 4,000

tons of coconut shells every day. The total production of coconuts

in Central America, the West Indies and the Caribbean coast of South

America amounted to 131,000,000 nuts annually, equal to a supply of 75

tons of shells daily.

When we first began to build masks our demands for carboniferous

material ranged from 40 to 50 tons a day of raw material; but by the

end of the war, due to vastly increased mask requirements, we were in

need of a supply of 400 tons of coconut shells per day. This demand

would absorb the entire coconut production of the tropical Americas

five times over. It was equal to one-tenth of the total coconut

production of the Orient. Since transportation from the oriental

countries was out of the question on the scale demanded by our mask

program, it was evident that we were likely to be seriously embarrassed

by the lack of raw materials; and, indeed, at no time before September,

1918, did we have on hand a reserve supply of shells and other charcoal

materials that would last for more than a few days, though at no time

after the start was the actual output of masks retarded by lack of

these materials.

[Illustration: AIRPLANE PICTURE OF CARBON PLANT AND CANTONMENT OF GAS

DEFENSE DIVISION AT ASTORIA, LONG ISLAND.]

[Illustration: GENERAL VIEW OF CARBON PLANT NO. 3 ON LEFT AT GAS

DEFENSE DIVISION PLANT, ASTORIA, L. I.

Storage bins are in central background, with administration building

and carbon plant No. 2 in the right foreground.]

[Illustration: CARBON PLANT NO. 2 AT ASTORIA, L. I., SHOWING ALSO

OFFICE AND LABORATORY.]

In building up our supply of coconut shells we naturally turned first

to the resources in the United States. America normally consumes fresh

coconuts at a rate sufficient to supply about 50 tons of shells daily.

The war restrictions on the use of sugar had the effect of cutting

down the consumption of coconuts, used largely in candy and cakes,

and consequently one of our efforts was to increase by widespread

propaganda the use of coconut. The "Eat-More-Coconut" campaign more

than doubled the American consumption of coconut in a brief space of

time; and the 50 tons of shells daily, which had been the original

supply, grew in volume until in October, 1918, with the help of

importations of shell, we averaged about 150 tons per day exclusive of

the Orient.

The first heating of coconut shells to make charcoal reduces their

weight 75 per cent. Therefore it was evident that we could most

economically ship our oriental supply in the form of charcoal produced

on the other side of the Pacific Ocean. For this purpose, in August, we

established under the direction of an officer of the Chemical Warfare

Service a charcoal plant in the Philippine Islands. From this plant

agents were sent to Ceylon, India, Siam, and other oriental countries

to purchase enormous supplies of nutshells. This work was only gaining

momentum when the armistice was declared. As it was, the Philippine

charcoal plant actually shipped over 300 tons of coconut shell carbon

to the United States and had 1,000 tons on hand ready for shipment on

November 11.

The method adopted in the Philippines was to burn the shells in long,

shallow trenches. As soon as the smoke had disappeared and the flames

came clear and lambent through the incandescent mass, the bed of

coals was smothered by means of galvanized-iron lids thrown over the

trenches. It is interesting to note that the coolies hired by the

Chemical Warfare Service in the Philippines would not work at charcoal

burning more than a few hours each day, because they declared that the

heat from the pits would give them tuberculosis and other lung troubles.

Meanwhile agents and officers of the Gas Defense Division were

searching the tropical regions of Central and South America for other

nuts valuable for this purpose. The best of these was found to be the

cohune or corozo nut. These nuts are the fruit of the Manaca palm tree.

They grow in clusters, like bananas or dates, one to four clusters to a

tree, each cluster yielding from 60 to 75 pounds of nuts. Cohune nuts

grow principally on the west coast of Central America in low, swampy

regions from Mexico to Panama, but are also found along the Caribbean

coast. Before the war created a demand for cohune nuts none of them

had ever been imported commercially in this country, although it is

understood that France had a prewar commercial use for them.

The chief virtue of the cohune nut from our point of view was its

extreme thickness of shell, the kernel of this large nut, which is 3

inches or more in length and nearly 2 in diameter, being relatively

small. We were importing cohune nuts at the rate of 4,000 tons per

month at the time of the armistice. A disadvantage in the use of cohune

nuts was that their husks contained a considerable amount of acid

which rotted the jute bags and also caused the heaps of nuts to heat

in storage. The fire department at the Chemical Warfare Service nut

storehouse at Astoria, N. Y., was kept busy putting out spontaneous

blazes in the storage piles of cohune nuts. We also sent agents to West

Africa and there arranged for the shipment of some hundred tons of palm

nuts a month.

A third source of tropical material was in the ivory nuts used in

considerable quantities in this country by the makers of buttons. In

the button factories in this country there is considerable waste of

this nut material, amounting to 400 or 500 tons a month, this waste

including the nut dust which was useless to us and had to be screened

out. The price of ivory-nut waste was high, because of the use of this

material in the manufacture of lactic acid. Nevertheless, we used a

considerable quantity of it.

Another great branch of activity in securing carbon supplies was

undertaken in this country. In the search for fruit pits and for

domestic nuts it was found that the quantity of apricot pits, peach

pits, cherry pits (largely from the canning industry), and walnut

shells on the Pacific coast amounted to 23,600 tons annually. We

arranged for the whole Pacific coast supply of these commodities and

converted a part of a San Francisco plant of the Pacific Gas & Electric

Co. into a plant for the preliminary carbonization of 100 tons a day of

these materials.

The next step was to turn to the consumers of the country and ask them

to save their peach and apricot stones, their prune, plum, and olive

pits, their date seeds, cherry pits, butternut shells, Brazil nut

shells, and their walnut and hickory nut shells. The work of securing

these and advertising the Government's need to the public was turned

over to the American Red Cross. There was some question at the start as

to whether the charter of the Red Cross would permit it to undertake

such a war activity; but, since it was determined that this was purely

a defensive operation, the legal forces of the Red Cross decided that

the organization could go into a campaign of this kind.

[Illustration:

BAREFOOTED NEGROES IN SPANISH HONDURAS SHOVELING COROZO NUTS

INTO BASKETS TO BE LOADED INTO BOATS FOR SHIPMENT TO GAS

DEFENSE DIVISION.]

[Illustration: 5,000 TONS OF PEACH PITS PILED UP AT SAN FRANCISCO.

This is enough to produce 1,600 tons of carbon for use in gas mask

canisters.]

[Illustration:

NUT SHELLS STORED ON DOCK OF EAST RIVER WHARF, ASTORIA, L. I.,

AFTER BEING UNLOADED FROM BARGES. SHELL-CRACKING TOWER ALSO

SHOWN.]

[Illustration:

1,200 TONS OF APRICOT PITS AT SAN FRANCISCO READY TO BE REDUCED

TO CARBON FOR GAS MASKS.]

[Illustration: TROOPS IN TRENCH AT LONG ISLAND CITY READY FOR A GAS

ATTACK.]

[Illustration: TROOPS WEARING GAS MASKS CHARGING IN OPEN ORDER AT LONG

ISLAND CITY.]

[Illustration: "GAS!" TROOPS HASTILY DON THEIR MASKS AT THE ALARM.]

[Illustration:

TROOPS WEARING GAS MASKS CUTTING BARBED-WIRE ENTANGLEMENTS IN

TESTS AT LONG ISLAND CITY.]

"Help us to give him the best gas mask." That was the slogan which

was carried on the posters, catching the attention of almost every

person in the United States. More than 1,000,000 pieces of literature

were distributed. The Red Cross established 163 collection points,

and collection barrels appeared on the streets of practically every

community in the United States. The Junior Red Cross, the Food

Administration, and the Department of Agriculture gave valuable

assistance. The Boy Scouts organized nut gathering parties. The

governor of Massachusetts proclaimed November 9, 1918, to be gas mask

day for the collection of carbon material, and 28 other States fixed

gas mask days in November. Two reels of motion pictures were shown

through the country. Journalists aided the campaign in newspapers and

magazines. Frederic J. Haskin sent out a valuable article which was

published in many of the important newspapers of the United States. One

Oklahoma town took a day off en masse and gathered a whole carload of

nuts.

This campaign started September 13, 1918, but was abruptly cut short

on the 11th of November. Thus it is impossible to give exactly the

result of it, since many of the scheduled shipments of nuts and fruit

pits were canceled and found their way into fuel bins. However, at one

time there were on the rails, en route to the carbon plant at Astoria,

100 carloads of materials supplied by the patriotism of the American

people. It was estimated that some 4,000 tons were collected in this

brief period, exclusive of the material from the California canning

industry.

The procurement of the nuts, however, was but the first step in the

production of carbon for use in our mask canisters, for after charcoal

is first burned its pores are still filled with various impurities

which may be summed up by the word "tar." When the charcoal was given

a second heating, under careful temperature regulation, this tar was

burned out, with the result that the charcoal itself became much more

active in its absorption of gas. In fact, properly activated charcoal

is more than absorptive--it is catalytic in its action toward the

gaseous poisons used in the war, not only absorbing them but hastening

their breakdown (digestion) into less injurious substances.

The activating of charcoal offered at the start considerably more

of a problem than the question of making the charcoal itself,

since activating had never before been conducted on a commercial

scale. Two months of experimentation showed us that the best

distillation of shells and pits for charcoal was that conducted in

illuminating-gas-making retorts. The activation thereafter had to be

done in special equipment permitting of fine control of temperature.

The Government eventually spent more than $1,000,000 in a charcoal

activating plant, providing for America the best protection known to

science against the poisons which Germany had introduced into warfare.

The cement granules, which also had to go into the canisters, supplied

another problem. We originally used a special soda-lime for this

material, but only obtained a satisfactory product after Maj. H. W.

Dudley, R.E., came to America as our British advisor and brought to

us the British granule formula. The basis of this cement was lime, to

absorb gases of an acid nature. Portland cement was used, to give

hardness and prevent disintegration and the formation of dust in

the canister. Then infusorial earth was added, to make the compound

porous in texture. A little sodium hydroxide was put in, to increase

the alkalinity of the mixture. Finally there was an infusion of

sodium permanganate, which is a powerful oxidizing agent. This latter

chemical was added as a precaution against arsine. Arsine and arsenical

compounds were difficult to use in warfare, but the Germans had

introduced them to some extent, justifying us in adding this protection.

In making the granules the sodium permanganate solution was mixed with

the cement. The mixture was roughed out into slabs, allowed to set for

three days, dried, ground up, screened to the proper size, and packed

in drums for future use.

As has been noted, the charcoal and cement were packed in the canister

in alternate layers. The cement had the virtue of working while the

carbon slept--that is, the carbon was active when there were gases

present to be absorbed, but the cement kept on thereafter, digesting

the gases which had been absorbed by the charcoal. The cement was not

quick in action, but it had a remarkable capacity for consuming some

poisons.

To return to the chronological development of manufacturing facilities,

after we had placed the contracts for the first 1,000,000 masks in

the early fall of 1917, we began looking around for facilities for

producing carbon and cement in the quantities which we should need in

the near future. We found at Astoria, the district near Hell Gate at

the junction of the East River and Long Island Sound in New York, the

large gas works of the Astoria Light, Heat & Power Co. perhaps the

largest illuminating-gas plant in the world. This was a subsidiary of

the Consolidated Gas Co. of New York, which concern readily agreed

to turn over to the Government some of its retorts and to permit the

construction of a Government-operated plant on its grounds. We might

have been seriously delayed in the production of gas masks except for

the extraordinary and continuing efforts of Mr. W. Cullen Morris, Chief

Construction engineer of the Consolidated Gas Co., and Mr. Addicks, its

vice president. It was due to Mr. Morris that a $150,000 granule plant

was constructed, heavy complicated equipment installed, and operations

started in the short space of 30 days.

Let us now go back to the history of actual mask production. At the

start it was estimated that when the Hero Manufacturing Co. had reached

full capacity it could assemble and turn out 6,000 masks a day. The

fuel shortage and the railroad congestion of the late fall and early

winter of 1917-18 hampered our supplying the Hero Manufacturing Co.

with parts, until the mask production, averaging 2,430 a day as it

had in November, dwindled to 1,500 a day in December. The Goodyear

Co. at Akron had meanwhile established its Akron-Boston motor track

line. This was put at the service of the Gas Defense Division, hauling

various supplies from both Akron and Boston to the assembling plant at

Philadelphia. Sometimes in the mountains of Pennsylvania the trucks

would be blocked in snow and the patriotic citizens of the community

would get out with shovels and work until the supplies again started on

their way.

[Illustration: Slabbing the doughlike mixture of carbon and spreading

it on screen-bottomed trays at carbon plant No. 2.]

[Illustration: Chemical laboratory showing apparatus for testing the

absorbent power of the carbon and gran powder.]

[Illustration: Retort house. Discharging machine drawing out the nearer

half and pushing out the far half of the hot carbon.]

[Illustration: Carbon plant No. 3, showing treater room and west side

batteries of Dorsite treaters.

FOUR VIEWS OF CARBON PLANT AT ASTORIA, L. I.]

[Illustration: Zigzagging and combing work on masks.]

[Illustration: Doping the masks.]

[Illustration: Assembly department.]

[Illustration: Final inspection department.

FOUR INTERIOR VIEWS OF GAS MASK PLANT.]

All of the masks produced in the fall of 1917 were still regarded as

experimental and not yet up to the standard of masks which we were

willing to supply for actual service at the front. Consequently, not

one of them was exported, but the entire 1917 production, after the

first order of 25,000, was sent only to the training camps in this

country. By January 8, 1918, we were producing masks which we were

willing to put into actual service, and on that date the manufacture of

masks for export was started.

In January we exported 54,000 masks, which was 16,000 less than the

schedule which we had set for ourselves. But by February 20 we had

wiped out this deficit with a little over, for our schedule by that

date called for the production of 141,000 gas masks, and we had

produced 142,000.

Late in the fall of 1917 the requirements of the Expeditionary Forces

were reanalyzed in the light of information gathered abroad and in

accordance with the new military program. Requirements were multiplied

almost fourfold. Let us see how these requirements were met, and what

difficulties were solved in the course of the effort.

Experience had already shown that for many reasons the Government

needed its own mask factory, where improvements could be adopted as

soon as made and where inspections and the storage of parts could be

more centralized than in private plants.

With the necessary expansion then confronting us, any other policy

would have meant making face pieces in half a dozen or more private

plants, all starting at once with organizations untrained for this

work. This would have been fatal, for even with the Goodyear and

Goodrich companies manufacturing face pieces in Akron and the Kenyon

Manufacturing Co. making them in Brooklyn, we found it most difficult

to maintain uniform standards in all the plants. As new points came up,

it was constantly necessary to interchange inspection personnel and to

send men from one plant to another to teach manufacturing wrinkles.

Such practices consumed more personnel than we could train in the time

available. Moreover, it was impossible under the conditions that we

were then facing to build up more than barely adequate supplies of gas

mask parts and such raw materials as special fabrics. To have operated

many more face-piece plants would have meant to divide these stocks

of fabrics, elastic, tape, etc., still further. To have kept each of

these plants properly stocked, under the existing traffic conditions,

would have been impossible. A big central gas-defense plant was the

only solution of our difficulties.

The order approving the establishment of the gas-defense plant was

signed by Secretary Baker on November 20, 1917. The officers of the

Gas Defense Division found in Long Island City, not far from the new

chemical plant at Astoria, a group of modern concrete factory buildings

which had been put up in this newly developed section by several

different concerns, among them the Ford Motor Co., the Goodyear Tire

& Rubber Co., and the National Casket Co. One of these buildings,

known as the Stewart Building, was taken over by the Government and

modern machinery was installed. Mr. R. R. Richardson, of Chicago, was

appointed plant manager with a salary of $1 per year. He quickly set

to work organizing the factory and its staff. On January 9, 1918,

the first few factory operators were hired. Five days later the

executive offices at the plant were ready for occupancy. The plant

grew apace. One by one the other buildings were absorbed and added to

the establishment--first the Goodyear Tire & Rubber building, then

the National Casket building. Next a long storage building was built

between the Stewart and the Goodyear buildings. Runways were built

which connected up the various buildings, and, finally, in July, the

Ford Motor building was taken over and connected up to complete the

group.

Thus by the summer of 1918 we occupied five large buildings, with a

total of over 1,000,000 square feet, or 20 acres, of floor space,

connected up to make the gas-defense plant. Of the 12,000 employees in

this plant, 8,600 were women. Endeavors were made as far as possible

to hire those who had near relatives with the American Expeditionary

Forces. The degree of care required in the manufacture of masks was

beyond anything known in normal industry, and we rightly believed

that this personal interest in the work would bring about greater

care in manufacture and inspection. Since the factory was working at

top speed a great deal of attention was paid to welfare work. Women

employees were given 12-minute rest periods both in the morning and

the afternoon, and completely furnished attractive rest and recreation

rooms were set apart for women in the factory.

[Illustration: GAS-DEFENSE PLANT AT LONG ISLAND CITY.]

[Illustration: EMPLOYEES OF GAS-DEFENSE PLANT LONG ISLAND CITY,

NOVEMBER 11, 1918.]

The plant was unique in more than one respect. At the very start it

attempted the supposedly impossible, for it combined in its staff

and in its working organization civilian and military personnel. The

manager was a civilian, the assistant manager was Lieut. Col. Coonley.

Below them on the next tier of the organization were Army officers

in charge of several departments and civilians in charge of others.

Throughout the plant were certain groups of women workers or inspectors

in charge of civilians were others; in charge of sergeants or even

privates. The arrangement worked out well and the whole organization

pulled together as one team, without reference to civilian or military

status. Again, at the start there was laid down a policy of inspection

at every single stage of manufacture. The incoming parts, though

already inspected at their source, were reinspected and retested. After

every operation in the manufacture of the face piece there came an

inspection by specially trained women set apart from the operators.

Then again, there was a special control inspection. After the face

piece was finished, and when assembly was complete, the entire mask

went to a final inspection where it was looked over by several trained

women, who worked in dark closets and inspected the face pieces over

a bright light to make sure that no pin pricks had been made, either

maliciously or otherwise. Furthermore, wherever there was an inspector

there was a system of checking his or her accuracy, for 5 per cent of

every inspector's work was periodically selected at random and checked

over by other inspectors.

Hand in hand with this went many of the latest developments of factory

operation. The best machinery was employed, conveyors were used

wherever possible, and, when changes in the size of the operation

or the design of the mask made it advisable, the factory was at

once rearranged in order that the flow might always be orderly and

continuous.

From all of this the reader might judge that the operation, lasting,

as it did, for only a little more than eight months, was a costly one.

Such, however, was not the case, for a well-ordered and accurate cost

system, kept from the very start in accordance with the best practices

of factory accounting, showed that after charging in all equipment

changes and overhead, the plant made complete masks which cost the

Government about 50 cents per mask less than it cost to get complete

masks by purchasing parts and assembling them under private contracts.

Along with this manufacturing development went the building up of

an elaborate procurement force charged with the responsibility of

providing parts to be assembled at the gas-defense plant and the Hero

Manufacturing Co. This section faced a hard and intricate task, but

though there were instances where a shortage of parts temporarily

slowed down production, these were remarkably few. Many were the

difficulties of buying new parts; many of the parts were the product

of elaborate die work; die makers in the country were overworked.

Specifications had to be written, checked, and approved, and a field

inspection first had to be organized and trained so that the product

from all the different plants could be relied upon as satisfactory for

the assembling plants. But this problem was still further complicated

by ever-recurring changes in design, made necessary as improvement

followed improvement. Officers had to be trained in a day and then sent

out to train inspection corps in manufacturing plants in many parts of

the country. Inspection and procurement detachments were maintained in

most of the eastern industrial centers. There were over 100 enlisted

men and 9 officers in Akron, 30 enlisted men with 6 officers in Boston,

and men and officers in over 60 cities. Here again the civilian and

Army officer worked hand in hand; for Mr. Robert Skemp, a volunteer

civilian from Pittsburgh, was in charge of this procurement, reporting

to Lieut. Col. Besse and directing an organization made up almost

entirely of officers and enlisted men.

The March output of masks was 183,000; that of April, 363,000; May,

slightly less than this figure; that of June, 504,000; that of July,

671,000. In all, between January 1 and November 11, 1918, we built more

than 5,000,000 gas masks.

In February, 1918, shortly before the German drive commenced, we

received requisitions for sample lots of oiled mittens and oiled union

suits as protection against mustard gas and also for chloride of lime

to neutralize poison-impregnated earth. In their March drive the

Germans used gas in much more protracted concentrations than before.

Originally the masks had been worn only during the sporadic gas alarms,

and then only for a brief period at a time.

The double-protection mask which we had been building had been

admirable in its day, but it was no longer adapted to the sort of use

to which it was evident it must now be put. In long-continued wear the

mouthpiece would irritate the gums and lips of the soldier, and the

face-piece band would cause excruciating headaches after a few hours.

It had now become frequently necessary for men to wear their masks for

eight hours at a stretch. The word discomfort is a weak description of

the feelings of a man wearing one of our masks for that period.

Our authorities in France decreed for a single-protection mask and more

comfort, even at the expense of a little safety. The result of these

new conditions together with the establishment of closer relationship

with our Expeditionary Forces, through a visit of Col. Dewey to France,

was the determination to build masks in this country which should give

the protection of the masks which we had been turning out and at the

same time be comparatively comfortable. There had been brought out in

France a single-protection mask, that is, a mask in which the inlet

tube entered directly into the space between the mask and the face,

with the orifices so arranged that the fresh air was drawn across the

eyepieces. This was known as the Tissot mask. The principle of the

Tissot was correct as far as comfort was concerned, since it did away

with the irritating mouthpiece, but the chief danger in this mask arose

from the fact that it was made of thin, pure gum rubber. We took the

Tissot and endeavored to produce a mask of this type which should be

gas-tight and yet rugged. In this work we experimented on hundreds of

subjects to determine face and head sizes and shapes. It is interesting

to note in this connection that the size of a man's face has nothing to

do with the size of his head, as large heads with small faces and small

heads with large faces occur not infrequently.

We made two developments of the mask without mouthpiece or nose clip.

Both were ready for field tests in August, 1918. The one produced in

Akron and assembled at the Philadelphia contracting plant was known as

the Akron Tissot, or Type A-T.

At the start of operations in Long Island City Mr. Waldemar Kops, of

New York, a manufacturer of corsets, came to the Government, asking

an opportunity to do his part in the war. He was assigned to the

gas-defense plant, and later, with the commission of major, took

charge of the gas-defense Long Island laboratories. Maj. Kops had no

experience with gas masks until he came to the gas-defense plant, but

his experiments soon led to an improvement in the design of the Tissot

mask. It was called the Type KT mask--the Kops-Tissot. Only a few

hundred thousand were produced, though the latest model was scheduled

for enormous production beginning in December, 1918. It possessed much

of the protective efficiency of the old uncomfortable mask, the cut of

the face piece insured a gas-tight connection with the head, it was

relatively comfortable, and it was durable.

The call of the allies in the spring of 1918 for American troops in

as great numbers as the ships could carry them to France resulted in

still further increases in our mask requirements. At the height of the

drive we were making over 40,000 masks a day. Approximately 35,000

employees were engaged in the manufacture of various gas-mask parts.

Our carbon requirements were expanding at a rate that would have needed

400 tons of raw materials a day by December, 1918. We built 336,919 KT

masks and approximately 200,000 A-T masks. In exact figures the total

production of masks of all types was 5,692,499. Of these 3,666,683 were

built at the gas-defense plant and 2,025,816 were assembled by the Hero

Manufacturing Co. In addition, we furnished 3,189,357 extra filled

canisters for the replacement of those used up by 40 hours of field

service.

Hand in hand with this procurement and manufacturing achievement went

the development of the technical section of the Gas Defense Division.

This was known as the Long Island laboratories, manned by a personnel

of several hundred men and officers. Here in its laboratories were

solved the knotty problems that bridged the gap between experimental

work and production. Many new designs were worked out, only to be

rejected when tested. Here there were workrooms that could make sample

lots of 1,000 masks, and here were located the chemical laboratories

and the gas chambers in which the product of the gas-defense plant was

tested daily by control chemical analysis and by actual breathing and

wearing tests.

When the fighting stopped our military scientists and others