The explosion of an H. E. shell is really a series of explosions. The process of the burst is about as follows: The firing pin strikes the percussion primer, which explodes the detonator. The detonator is filled with some easily detonated substance, such as fulminate of mercury. The concussion of this explosion sets off the charge held within the long tube which extends down the middle of the shell and which is known as the booster. The booster charge is a substance easily exploded, such as tetryl or trinitroaniline (T. N. A.). The explosion of the booster jars off the main charge of the shell, T. N. T. or amatol. This system of detonator, booster, and main charge gives control of the explosives within the shell, safety in handling the shell, and complete explosion when the shell bursts. Without the action of the booster charge on the main charge of the shell, the latter would be only partially burned when the shell exploded, and part of the main charge would thus waste itself in the open air.

The shell used by our Army before the war had been largely of the base-fuse type. Interchangeability of ammunition with the French required that we adopt shell of the nose-fuse type. The boosters and adapters that went with this type were unfamiliar to our industry.

The adapter is the metallic device that holds the booster and fuse and fastens them in the shell. The adapter, therefore, is a broad ring, screw-threaded both outside and inside. The inside diameter is uniform, so as to allow the same size of booster and fuse to be screwed into shell of different sizes. The outside diameters of the adapters vary with the sizes of the shell they are made to fit, the rings thus being thicker or thinner as the case may require. Fuses of several sorts are employed by the modern artillerist; and with shell equipped with adapters, any fuse may be inserted in the field right at the gun.

Unexpectedly the manufacture of boosters and adapters proved to be much more difficult than it appeared to be at the start, and the shortage of these devices was a limiting factor in the American production of shell.

On May 1, 1917, drawings and specifications were sent to the principal manufacturers of ammunition and ammunition components inviting bids on 3-inch ammunition. These bids were opened on May 15, 1917, and after full discussion with the Council of National Defense orders were placed for 9,000,000 rounds of 3-inch shell and shrapnel ammunition. The bids for shell and shrapnel ammunition for all the other calibers of guns and howitzers we had on hand then were about to be asked, when the French mission to this country arrived; and the sending out of proposals was deferred, while discussion ensued as to changing our 3-inch and 6-inch artillery to 75-millimeter and 155-millimeter calibers, so as to make our ammunition interchangeable with that of the French. This decision was made June 5, 1917.

There then took place much discussion and consideration of the French ammunition. The French had several distinct types of shell, ranging from the very thin walled high capacity kind to the thicker walled types. The French specifications were radically different from our own or those of the British. The steel shell in the French practice was subjected to a drastic heat treatment, which did not seem necessary to us for the thicker walled types of shell.

The French fusing system also was entirely different from that used by our service. French fuses were carried separately, and the adapter and the booster casing were screwed permanently into the shell.

Our decision to adopt French types of ammunition made it necessary to rearrange all our plans, and to obtain drawings of the shell, boosters, adapters, and fuses from France. This caused much negotiating, and a considerable amount of time was consumed in getting the necessary specifications and drawings here.

As a result of recommendations from French officials against production in this country during 1917 of the so-called "obus allongé" and the semisteel type of shell, no attempt was made to produce these for the 155-millimeter guns and howitzers during the first year of the war, but as a result of new recommendations and investigations of our officers in France in the spring of 1918 both of these types of shell were put into quantity production here. When the armistice was signed they were being turned out in such quantities that it appeared that there was sure to be an ample supply on hand in the early spring of 1919.

Radical differences of manufacture existed between the French and British in the matter of specifications and methods of production. Large quantities of British ammunition had been made in this country, and we had adopted the British 8-inch howitzer, so that it appeared we should use British practice in the manufacture of shell. Manufacturers claimed that great delay would result in the production of shell here if the heat treatment and hydraulic tests were insisted upon as the French specifications called for, and investigation proved this to be essentially true, as no facilities for heat treating and hydraulic testing existed.

The upshot of the entire matter was that it was decided to use French dimensions and shell for the 75-millimeter and 155-millimeter calibers so as to obtain uniformity of ballistics, but to permit American metallurgical practice to obtain in the manufacture. Shells made under these specifications were tested by the French commission in France. The verdict on these shell can be summarized in this quotation from their report:

Our designs for shrapnel and time fuses had been proven to be entirely satisfactory, and they were continued as they were. In fact it was generally agreed that ours was the best time fuse used on the allied side during the war. That our decision in the matter of continuing production of shrapnel and time fuses was warranted, is borne out by the fact that we obtained early deliveries in sufficient quantities to meet requirements.

In the use of the adapters and boosters, which introduced an entirely new component to our service in shell making, we had had no experience, and subsequently met with great difficulties due to this lack of experience. Delays were encountered because in this part of shell manufacture it was generally necessary to await information from France whenever difficulties were encountered, or to conduct experiments before we could proceed.

When we began receiving our bids for 3-inch gun ammunition there were comparatively few factories in the United States that were able to turn out complete rounds of ammunition. There were many factories, however, capable of turning out one or more of the shell components. It was necessary to place orders for complete rounds of ammunition with those factories that could furnish them, and have the remaining components manufactured separately, and to provide assembling plants. To get as many factories as possible on a production basis in anticipation of the future large orders for ammunition that must necessarily follow with extension of operations by our field forces, orders for our initial quantities of ammunition were distributed as widely as possible.

To prevent confusion and loss of time because of the scramble for steel forgings and other raw materials it was decided that the Government would purchase all raw materials as well as furnish components for ammunition.

How successful we were in getting into quantity production on ammunition after the numerous and large obstacles in the early months of the war can be indicated best by the fact that of the 11,616,156 high-explosive shell for 75-millimeter guns machined up to November 1, no less than 2,893,367 passed inspection in October; while of the 7,345,366 adapters and boosters for 75-millimeter guns that had been machined up to the 1st of November, 2,758,397 passed inspection in October.

The figures for the 4.7-inch and 155-millimeter guns and howitzers follow:

Ammonium picrate or explosive D upon which this country had depended almost entirely up to the time of our entry into the war was forced into the shell under hydraulic pressure. The adoption of the point-fused shell and an explosive for shell filling new to this country, namely, amatol, made necessary the provision of new methods for shell loading and the expansion of plant facilities for these new methods capable of loading the vast and tremendous numbers of shell required in modern warfare. As a result of reports, following investigations by our officers of methods used abroad, various new shell-loading plants were built in the United States.

The names, location, and output of the shell-loading plants in our country are as follows:

It was found necessary in the early stages of the war to fill all shell with T. N. T., regardless of cost, until there could be built the required and properly equipped plants for the mixing and loading of amatol.

Two methods for loading T. N. T. were adopted. The one most largely used, however, was the casting method by which the chemical was brought to a molten condition in a steam jacketed kettle and poured into the shell. To do this two operations were usual. First, the shell was filled approximately two-thirds full with the molten material, and then as soon as a crust was formed this was broken through and the second filling took place. This process was necessary to prevent the formation of cavities in the filling charge. Such cavities cause breakdowns, resulting almost invariably in incomplete or entire failure of detonation.

The ammonium nitrate first produced in this country during the war was of such a character that proper densities could not be obtained when mixed with T. N. T. to form amatol. This difficulty was overcome after much investigation, and proper methods were outlined for the ammonium nitrate manufacturers, with the result that Grade III ammonium nitrate was produced as a sharp, hard crystal at a setting point of not less than 290° F. This was found to be perfectly satisfactory.

The so-called 50-50 amatol, composed of 50 parts ammonium nitrate and 50 parts T. N. T., is loaded into shell by a casting method similar to that used in loading T. N. T. alone.

The so-called 80-20 amatol, composed of 80 parts ammonium nitrate and 20 parts T. N. T., was originally loaded cold, by hand, and then followed up with mechanical pressing. As a substitute for this method, which is accompanied by a certain element of danger, the use of hot 80-20 amatol, was resorted to in England. This was tamped by hand to the proper density, it being more compressible than cold amatol.

As this is an exceedingly tedious method of operation it was entirely done away with in England, except for large shell, by the use of what is known as the horizontal extruding machine. With this machine the British were able to load 80-20 amatol with great success into the 75-millimeter shell and higher calibers up to 8 inches.

This machine took a mixture of T. N. T. and ammonium nitrate in a jacketed hopper, so that the temperature might be maintained, and the hopper fed it down through a funnel upon a screw that was placed against the shell by counterweights to give the proper density. One of these machines was imported here from England, but, as it was unsatisfactory from a construction standpoint, new and satisfactory machines were built on the same principles of construction in our own amatol loading plants.

Experimental work with these machines was carried on at the Government testing station Picatinny Arsenal, Dover, N. J., and the du Pont Experimental Station, Gibbstown, N. J., as well as experimental plant operations at the Morgan plant of the T. A. Gillespie Co., Parlin, N. J., and the Penniman plant of the du Pont Co., Penniman, Va. All difficulties of the operations were overcome so satisfactorily that the greater portion of the loaded shell was produced by this method.

The metal parts as received at the shell-filling plant are inspected and cleaned to remove all traces of foreign matter such as grit or grease before being sent to the loading room. After being loaded the shell are again inspected. At intervals a split shell is loaded and then taken apart and examined, so that any loading defects may be found quickly and conditions remedied, before any large quantities of shell are produced.

The cavity left in the amatol by the tube of the extruding machine is filled with molten T. N. T., and a cavity is produced in this T. N. T. into which the booster fits. This is necessary in order to provide for complete detonation. The booster cavity is produced either by the use of a former, which upon removal leaves a cavity of the proper size, or by plunging the booster into the shell filling before this is cooled, or by drilling out a cavity for the booster after the filling has been thoroughly cooled.

A large number of rounds of ammunition of all calibers had also to be loaded with a flashless compound that was inserted in the propelling charges, so that the discharge of the guns would not betray their positions to the enemy at night, while a smoke compound was inserted in a large quantity of shell so that each missile of this character might be located after firing to determine the accuracy of the shot.

Coordination of manufacture of metallic parts so as to cause the proper quantities of shell, fuse, and boosters to be produced without leaving any incomplete rounds that would have to be held awaiting other components caused the greatest difficulty.

The magnitude of the task of providing the necessary shell components in the tremendous quantities required can be better appreciated by a realization of the fact that the various parts of each component must be made to fit each other properly and perfectly. Gauging had to be resorted to frequently in the process of manufacture to make certain that there was perfect interchangeability of parts of each component to prevent any waste of time in selecting parts to fit each other.

The complete components, too, must themselves be made with equal care and scrupulous attention to make certain that they fit properly. Thus, the booster had to be made in such a fashion and with such precision and accuracy that it would fit perfectly into the shell as well as into the booster cavity in the shell filling into which it is screwed and also at the same time accommodate the fuse which screws into the booster.

This extreme accuracy made necessary a large number of gauges, which had to be designed at the same time as, and in coordination with, the design of the component. For example, in a complete round of artillery ammunition, 80 dimensions must be gauged. To standardize the gauges used for these 80 dimensions, 180 master gauges are required, while the actual number of different gauges used during the various stages of manufacture of a complete round is over 500.

Government inspectors required over 200 gauges in their work of inspecting and gauging the finished components for the shell, so in all about 800 gauges were used in the process of manufacturing a complete round of artillery ammunition, to insure interchangeability of parts, proper fit for the projectile in the gun, and perfect functioning of the various parts.

All fixed ammunition was assembled at the shell-filling plants, making it necessary to install at these points storage capacity and equipment to handle the propellant powder as well as to fill the high-explosive shell. Boosters and fuses were loaded at separate plants and shipped to the shell-filling assembly places to be packed for shipment with the shell for transportation overseas.

The cost of a loaded 75-millimeter shell with the fuse and propellant charge ready to be fired is about $11. Such a shell contains a little over 1½ pounds of high explosive, which costs $1. The loading and assembling of the complete round costs $4.

A loaded 155-millimeter shell complete with fuse costs about $30, exclusive of the propellant charge of powder, which is loaded separately. A shell of this caliber holds about 14¼ pounds of high explosive, which costs $10, while the loading and assembling costs $4.

The 75-millimeter and 155-millimeter shell were used in the greatest quantities on the European battle fields, and at the time of the signing of the armistice our American loading plants were concentrating on filling ammunition for guns of these two calibers.

The nature of the work carried on at these shell-loading plants, of course, made the danger of a disaster ever present. Prior to our entry into the war an explosion at the Canadian Car & Foundry Co.'s plant, Kingsland, N. J., resulted in the entire destruction of the plant with large loss of life.

In October, 1918, the Morgan plant of the T. A. Gillespie Co., South Amboy, N. J., was wiped out by an explosion in which about 100 employees lost their lives. Plans for rebuilding this plant, had progressed far when the armistice was signed. In the fall of 1917, 40 people were killed in an explosion at the Eddystone Loading Plant, Eddystone, Pa.

For the successful carrying out of our program for the production of vast quantities of explosives and propellants, as well as shell loading, the women of America must be given credit, on account of the highly important part they took in this phase of helping to win the war. Fully 50 per cent of the number of employees in our explosive plants were women, who braved the dangers connected with this line of work, to which they had been, of course, entirely unaccustomed, but whose perils were not unknown to them.

In connection with the production of shell themselves, the American Ordnance Department adopted certain changes of design which were not only radically different from what we had known before the war but were interesting for the way in which they were brought about and for the results they accomplished.

The modern shell as we knew it before the war was simply a metal cylinder cut off squarely at the base and roundly blunted at the nose. The shell is zoned with a so-called rotating ring, a circular band of copper which by engaging the rifling channels of the gun gives to the shell the whirl that keeps it from tumbling over and over and thus holds it accurately on its course in flight.

In the proof-firing of the 6-inch seacoast guns it was discovered that their fire was none too accurate; and the American ordnance engineers began studying the shell to see if the fault lay there. One of these experts was Maj. F. R. Moulton, who before accepting a commission in the Army had been professor of astronomy at the University of Chicago. Maj. Moulton began a study of the 6-inch shell; and soon it was discovered that the mathematics which could chart the orbits of comets could also deal with the flight of projectiles, calculate the influences of air resistance and gravitation, and eventually work out new, scientific contours for offsetting these influences as much as possible.

Maj. Moulton first dealt with the inaccuracy of our 6-inch shell. He discovered the cause in the rotating band. Although but a slight portion of this band was upraised above the surface of the shell's circumference, yet the enormous force exerted upon the projectile to start it from the gun actually caused the cold copper to "flow" backward. The result was that when the shell emerged from the muzzle of the gun it bore around its sides an entirely unsuspected and undesirable flange. This flange not only shortened the range of the shell by offering resistance to the air, but it was seldom uniform all the way around, a condition giving rise to the idiosyncracies of our 6-inch shell as they were fired at the target.

The remedy for this was a redesigned rotating band, making it somewhat thicker in front. The "flow" of the copper could thus be accommodated without causing any detrimental distortion to the projectile. When this improvement was made the 6-inch shell became as accurate as any.

But Maj. Moulton was to make an even greater contribution to the 6-inch shell. This shell, like those of our other types, was square ended at the base. Maj. Moulton in his new design tapered in the sides somewhat, making the shell "boat ended." He elongated the nose, bringing it out to a much sharper point. The result was the first American "streamline" design for a shell. Shell of this new model were built experimentally and tested. The 6-inch gun could fire its old shell 17,000 yards, while the streamline shell went 4,000 or 5,000 yards farther—2 or 3 miles added to the range of an already powerful weapon by the application of brains and mathematics.

The French were experimenting with streamline shell. We adopted the French streamline 75-millimeter shell and put it into production, calling it our Mark IV shell. Our regular 75-millimeter shell, known as the Mark I 1900 shell, had a maximum range of 9,000 yards. The Mark IV shell proved to have a maximum range of 12,130 yards, giving an increase in range of well over a mile. America up to April 3, 1919, turned out about 524,000 of these streamline shell.

The French also built shell of semisteel, steel to which iron was added. It was claimed that these shell, by bursting into fine fragments upon exploding, were more effective against troops than all-steel shell, because the fragments of the latter were larger. We adopted this shell also and produced it experimentally. In contour it was a compromise between the old cylindrical shell and the extreme streamline type and was easier to make than the latter.

The following table lists the name of each manufacturer of the various types and sizes of shell for big guns and states the quantity turned out by each: