under the surface they may remain latent for still longer periods, and take effect when disturbed by digging. It will be understood at once that this is a question of great tactical importance. For instance, a bombardment with " high-persistence " shells of trenches shortly to be attacked would be an obvious mistake, since the attacking troops after capturing the trenches would not be able to occupy them. On the other hand, in a raid on trenches which it was not proposed to retain it would be correct to burst bombs of " high-persistence " gas in trenches and dug- outs. So also in using a gas barrage or in bombarding an area, high-persistence shells should not be used if it is intended shortly to attack over that ground. The use of vesicatory shells especi- ally will deny a given area to both sides, or " nullify " it. Gas shell is now classified not according to its physiological effect, but according to tactical use, viz. as persistent or non-persistent, the former being used for neutralization (e.g. mustard gas) and the latter for surprise destructive bombardment (e.g. phosgene).
The questions may be asked why use the irritant type of gases at all if lethals are available? and among the lethal gases, why not use only the most powerful, namely, those which pro- duce immediate death? In both cases the answer is found in the question of quantity. The specific lethal gases will only produce their effect in very high concentrations, which means that a large number of shells must be used simultaneously over a certain area. Shells of other types, though they may not kill at once, will produce casualties in very much lower concentrations. The question of the number of shells to be fired to produce a given effect is of great importance, not merely from the point of view of expense and the call on manufacturing resources, but still more in the field, as regards the number of guns required to fire the shells, the exertion of the guns' crews, and the question of transport. Again, an effective lachrymator will produce an atmosphere that cannot be endured in one-thousandth part of the concentration which the lethal shell would require for its purpose. This is a matter of great importance, especially in neutralizing enemy batteries. Lachrymators rather went out of fashion towards the end of the war, not only because the great munition efforts of both sides had produced an enormous quan- tity of lethal shell, but still more because the neutralizing effects and harassing effects which were their raison d'etre could be obtained better by mustard gas.
Vesicating shells, which are of high persistence and whose effect is often delayed, are specially useful against targets behind the front line. Although a trench or strong point which it is intended physically to occupy cannot be subjected to mustard- gas bombardments, the use of this substance in combination with an attack round the flanks proved very valuable in reducing defences which could not have been carried either by assault or by explosive bombardments. The possibilities also of mustard- gas barrages in defence are very great. They should be used against communications, depots, railway stations, and especially staff offices, telephone exchanges and everything that affects the enemy's organization. An entirely odourless vesicator and one which does not produce a smoke easily recognizable will be par- ticularly effective in this way.
As a general rule both lethal and irritant shells should be used in scientific alternation. With lethals, of course, the object is to catch the enemy unprotected by his masks, and in order to get a good effect a large number of shells must be fired at once. Such effects may be specially aimed at when large numbers of troops are known to be concentrated in certain places, previous to an assault. But it is also an important object to force the enemy to wear masks as long as possible, not only to fatigue him, but to exhaust the protective powers of the mask. This can be effected with irritants, and after some hours of bombardment with these, fresh bursts of lethal shells may be tried.
Methods of Employment. The study of the characteristics of gas-clouds is very complicated. The cloud may consist of true gas, or minute drops of liquid, or infinitesimally divided solid particles. The last are known as " particulate clouds," and in their behaviour resemble a colloid vapour. Their action has to be studied physically and electrically as well as chemically.
The production of cylinder clouds, of course, is simple. The critical temperature of the gas employed must be above normal temperature. The liquefied gas is filled into a cylinder with a nozzle on the principle of a soda-water syphon. The cylinder is placed in position in the trench, and the nozzle is provided with a short length of pipe, which is placed on the ground in front of the trench, and ensures that the gas on issuing is well clear of it. The valve being opened, the liquefied gas is discharged with some force, and as its evaporation causes a fall of temperature a heavy cloud is formed which travels with the wind. The necessary density of cloud is obtained by opening simultaneously a suffi- cient number of cylinders per unit of length of trench, other cylinders being held in reserve to continue the discharge for the time considered necessary. At first the cylinders were placed in groups against the front wall of the trench. This method had the disadvantage that a cylinder might be burst at an inopportune time by an enemy shell, and later the Germans placed their cylinders under the floor of the trench, protected by sandbags, etc., while the British placed theirs in chambers excavated at some depth below the parapet.
The earliest cloud discharges lasted only twenty or thirty minutes, or at most an hour, the necessary concentration being calculated at ten tons of the chemical per km. of front attacked per hour. By the end of 1916 the French were using 100 tons per km. per hour, and the emission was continued for three or four hours. In the course of 1918 the British Special Brigade was using 200 to 250 tons of gas per km. per hour, and keeping up the cloud for eight, ten or even fourteen hours. The transport of gas cylinders up to the front trenches was naturally extremely laborious. It was very difficult to avoid attracting the enemy's attention to the carrying and emplacement of them, and there was always the risk of cylinders being burst by the enemy's shell. The results that were achieved under such condi- tions testify in the highest degree to the devotion and courage of the troops employed. The enormous discharges of 1918 were effected by loading the cylinders on trolleys and running them up to the front trenches on light railways just before they were to be used. The nozzles were opened by an electric device.
On the other hand the use of gas in shells presented all sorts of difficulties from the outset. It was first necessary to find gases that could withstand the shock of discharge from the gun and the effect of bursting the shell. Some of the most lethal gases could not be utilized because they were chemically unstable, and were liable to become decomposed into their constituent elements by shock. The cyanide compounds were to some extent of this nature, so that the Germans never used them. The French discovered a stable cyanide compound in which they had consi- derable faith, and the British used them to a certain extent.
Some stable lethal gases were found in due course, in addition to the irritants, and a whole series of problems presented them- selves. The first question was the possible effect of the gas on the material of the shell. Some gases, such as phosgene, had no action on steel or cast iron. Others required containers of lead or porce- lain; the French had a very effective method of blowing a glass lining into shell. The methods of sealing the shells and of filling and closing them offered merely technical difficulties, though these were considerable. The most suitable means of opening the shell was the next question, which had to be first considered, and then practically tested. The first gas shell used by the Germans contained a large proportion of high explosive; it is not known whether the idea was to follow the Hague Declaration in that the sole purpose of the shell should not be to spread deleterious gases, or to have a " double-purpose " shell, destructive and toxic. The result, however, was to produce an inferior explosive shell, while most of the gas was dissipated by the explosion.
The British efforts were directed at first to getting the maxi- mum amount of gas into a shell and releasing it with as little disturbance as possible. The gas-cloud would issue in the form of an oblate spheroid which travelled down the wind, gradually enlarging and being diluted by the air. Without wind the gas would remain on the surface, settling down in trenches or de- pressions of the ground. For opening with the least amount of
