he gave no indication of the direction in which this protection was to be sought. In reading that work it struck the writer that the requisite protection could not be obtained from armour without exceeding the available limit of weight, that the only means of affording it was to enable the sailor to find shelter in water as the soldier does in earth, and that this was an additional reason for the use of automatic artillery.
But when Mr. Arnold Foster wrote, the need of protection, though urgent, was as nothing to what it has since become. In 1891 'high explosives' were known indeed, but their poisonous effect was not appreciated; subsequently, it was recognised that the fumes of modern shell were at least as dangerous as the splinters, and indeed more so, for shields and casemates might afford some protection against the splinters, but would be useless against the fumes. Here, again, it seemed that the notion of the sheltertrench was the only adequate solution of the problem.
Still the writer hesitated to publish his views without first knowing that an automatic heavy gun was not only desirable but possible, and he probably would have kept his ideas to himself if they had not almost by accident been communicated to a professional friend, under whose guidance they quickly assumed a very different and far more practical shape.
As soon as the design was fairly complete, there was no longer any reason for silence, and accordingly an
