Salvage is a highly technical subject, but a few remarks at this stage are necessary if the reader is to appreciate the extent to which considerations of salvage affected the problem under discussion.

Salvage operations must vary according to the circumstances of each particular case. The size of the vessel, the damage which she has sustained, the manner in which she is resting on the bottom of the sea, the nature of the ground, the tides, the depth of water, the degree of exposure to rough seas, the proximity of shelter for salvage craft, and the distance from the land are all factors of importance, but they by no means exhaust the list.

One of our main purposes in considering salvage operations was that of ascertaining the chief obstacles to salvage, so that we could provide the enemy with as many of those identical obstacles as lay in our power.

Another important object, concerning the immediate problem at Zeebrugge, was that of deciding the best type and size of vessel to be used in addition to the question of what particular damage each vessel should receive, and how she should be fitted to defy attempts at removal.

There are three principal methods of removing a sunken ship. First, bodily removal with the aid of some lifting agent. Second, dispersion by explosive means. Third, piecemeal removal by cutting away.

Regarding the first-mentioned method, a small vessel can be lifted by passing hawsers beneath her and securing the ends to salvage craft on the surface overhead. The hawsers being hauled taut at low tide, the vessel will lift off the bottom when the rise of tide lifts the salvage craft, and can then be transported bodily elsewhere. Larger vessels can be lifted by the use of compressed air, or by pumping out the vessel after closing all holes under water. Provided the ship is upright the compressed air method can leave out of account the damage sustained below the vessel's normal waterline, but the remainder of the hull must be rendered airtight. Air can then be pumped into the hull until the vessel is lifted, and she can be towed away as required.

This method has been used successfully when removing large vessels, but the practicability of rendering them airtight chiefly depends on the damage which they have sustained. The pumping-out method, comparatively speaking, is the most simple one to adopt, provided that the damage to the hull is small. The damaged portion must be repaired by divers unless the more elaborate method of building a coffer-dam—i.e., a sort of dock—around the ship, is pursued. Divers cannot work in a strong tidal current or in rough weather. The repair of holes under water is rendered extremely difficult, if not actually impossible, when the bottom of the ship is badly holed with the ship resting on the damaged portion. The ship must be made watertight, or nearly so, below the surface of the sea before she can be lifted. The word "watertight" is qualified here because, as a matter of accuracy, the ship can be pumped out and lifted, provided that the pumps can eject water at a greater rate than the latter is flowing in. Before passing on to consider the next method it may be as well to remark that special difficulty is experienced when moving sand—i.e., silt—has access to the holes in the ship.

Dispersion by means of explosive charges may, under certain circumstances, be a simple operation, but, on the other hand, there are certain conditions which put this method outside the pale of choice. For instance, in the case of a ship sunk in a narrow channel where much silt is experienced, the explosive method is almost worse than useless. For every explosion in a given section of a vessel will tend to shatter that portion into several pieces. Each piece falls to the bottom and forms a new obstruction. Silt then enormously aggravates the situation, for the sand will collect against the obstruction until it becomes a miniature sandbank. Such shoals are then difficult to remove. A bucket-dredger—i.e., a vessel fitted with an endless chain of buckets for scooping up the bottom—will break her buckets as soon as they encounter the steel kernel of the shoal. On the other hand, a suction dredger—i.e., a vessel designed to suck up sand off the sea-bottom—cannot raise solid material. Neither type of dredger can remove the cause of the shoal; any removal of sand under such conditions is merely temporary; the sand will recommence building up the shoal as soon as the dredger ceases work. Dredging against such obstacles is of little more use than dredging against rocks.

There remains the third method, namely, piece-meal removal by means of "cutting away." Cutting away can be accomplished, in the ordinary course of events, by means of acetylene gas cutters or by pneumatic tools. Acetylene gas will cut through steel with little more effort than a knife cutting through india-rubber. But acetylene gas cannot be used under water and cannot cut through large thicknesses of cement. Pneumatic tools provide a very laborious and tedious means of cutting large quantities of steel. Work under water entails the use of divers. Thus, the removal of a ship by the piecemeal process is an exceedingly prolonged undertaking, especially as each piece must be lifted out when cut away; for reasons already stated the pieces must on no account be allowed to fall to the sea-bottom.

From the foregoing remarks we arrive at the following conclusions. The blockships should be too large to lift off the bottom by the hawser method. They should be extensively damaged and sunk in such a manner that they would rest on the damaged portion of the hull. They should be fitted to counter "cutting away" tactics, and should be sunk in positions where silt would render impracticable the explosive method of dispersion; the damage should be so situated as to give the silting sand access to the hull through the holes in the latter.