It is interesting to compare the attitude of our militarists to defence against gas with their attitude before the war to a possible German invasion. The fear of the latter, although the naval experts always stated that it was impossible on any serious scale, had been so impressed on the military mind by the propaganda of the National Service League and its like before the war that, from 1914 to 1918 hundreds of thousands of troops were quite unnecessarily kept in England. There however, this very fundamental difference between a defence against invasion and a defence against gas. The one would increase the importance of the professional soldier: the other would not. One does not need to be a very profound psychologist to see in this fact one reason why the military authorities dropped anti-gas training, and why I, being a biochemist and therefore a person of the type who would become important if gas war returned, am advocating its extension. As to which of us is justified, I would suggest that it is more likely to-day that poisonous gas will be used against British soldiers or civilians in future wars than it was in 1912 that Britain would be invaded by the Germans. We have seen that a case can be made out for gas as a weapon on humanitarian grounds, based on the very small proportion of killed to casualties from gas in the war, and especially during its last year. Against this may be urged the probability that future research will produce other gases or smokes which, as weapons, will be as cruel as, or more cruel than, the chlorine and phosgene used in 1915 and 1916. The answer to this is quite simple. First, as regards gases or vapours. Only a limited number of chemical substances are appreciably volatile, and of their vapours only a small proportion are poisonous. Now every chemical substance has a definite molecular weight. Those with a small molecular weight, _i.e._, whose molecules are relatively light, are on the whole the most volatile, _i.e._, go most easily into vapour. Now the large majority of the possible volatile chemical substances of small molecular weight, and therefore relatively simple chemical composition, are already known. Mustard gas, for example, was discovered and its properties described in 1886. There are probably substances of high molecular weight whose dense vapours are even more poisonous than mustard gas. But the charcoal of our respirators has the property of adsorbing heavy molecules of vapour quite independently of their chemical composition. It is, therefore, somewhat unlikely, though not, of course, impossible, that any very poisonous vapour will ever be found which will go through a mask impermeable to mustard gas or chlorine. It is, to my mind, far more probable that skin irritants may be discovered which are even more unpleasant than mustard gas.