During the war the steam trawlers were commandeered by the Government for use as patrol boats and mine sweepers. It is estimated that 10 per cent of our steam trawlers and drifters and their crews were lost during the war. [Illustration: FIG. 17 A.—The otter trawl. B.—Attachment of board to net. OB. Otter board. B. Iron brackets. C. Chain to connect with warps. M. Metal strengthening pieces. M′. Iron shoe. HL. Head line. UW. Upper wing. LW. Lower wing. LL. Lacing connecting wings. GR. Ground rope. D. Balch of lower wing. SSS. Twine settings connecting balch to ground rope. A. Headline and lacing connected to board by shackle. B. Toe of ground rope connected to board by shackle. C.—Bosom of a bobbin foot-rope for use on rough ground. AB. Balch line on head of belly and connecting with bosom of wings. SS. Wire seizings connecting balch to small intermediate bobbins, 6″ diameter (EE). Large bobbins up to 24″ diameter (FF).] When steam trawling was first introduced it aroused general opposition, for there was not only the fear that their efficiency would lead to over-fishing in certain grounds, but it was said that the trawl, when dragged along the bottom, destroyed the eggs and killed the immature fish. The line fisherman found that steam trawling made it more difficult to catch demersal fish with baited hooks. He attributed this to the effect of over-fishing, but it is probable that contact with the otter trawls had made the fish rather more shy and, therefore, more difficult to catch by this method. It is unlikely that steam trawling will lead to serious over-fishing, except possibly amongst such sedentary fish as soles and plaice. It must be remembered that trawling is only commercially possible on comparatively smooth ground and down to depths of about 200 fathoms. Probably, therefore, the actual area trawled is only a small proportion of the total area that is inhabited by fish. It is possible, of course, that extensive and long continued trawling in a confined and relatively isolated area may scare the fish away; it is probable, however, that any area in which over-fishing appears to have produced temporary exhaustion will tend to recover automatically, since it would naturally be abandoned temporarily by the trawlers for more profitable fishing grounds. There is no doubt that trawling, unless the size of the mesh is carefully controlled, tends to remove large numbers of immature fish. Generally in ordinary beam trawling—cod, plaice, haddock, etc.—the mesh varies from 3 ins. diameter near the mouth of the net to about 1-1/4 ins. diameter at the cod end. If a much smaller mesh were used the resistance encountered by a full-sized net would be so great that it would be almost impossible to draw the net through the water. Smaller trawls of 1/2 in. mesh are used in shallow coastal waters for catching shrimps, small plaice and whiting. The size of mesh largely determines