to be unable to disentangle itself for action as a purely naval unit in time to discharge its naval function. In other words, it must always be able to act in the same way as a free field army covering a siege.
Where the objective of the expedition is not the local naval base, the choice of a position for the covering squadron will turn mainly on the amount of support which the army is likely to require. If it cannot act by surprise, and serious military resistance is consequently to be expected, or where the coast defences are too strong for the transport squadron to overpower, then the scale will incline to a position close to the army, though the extent to which, under modern conditions, ships at sea can usefully perform the delicate operation of supporting an infantry attack with gun fire, except by enfilading the enemy's position, remains to be proved. A similar choice will be indicated where strong support of men and boats is required, as when a sufficiency of flat-boats and steam towage cannot be provided by the transports and their attendant squadron; or again where the locality is such that amphibious operations beyond the actual landing are likely to be called for, and the assistance of a large number of boats and seamen acting with the army is necessary to give it the amphibious tactical mobility which it would otherwise lack. Such cases occurred at Quebec in 1759, where Saunders took his covering battle-squadron right up the St. Lawrence, although its covering functions could have been discharged even better by a position several hundreds of miles away from the objective; and again in 1800 at Alexandria, where Lord Keith ran the extremest hazard to his covering functions
in order to undertake the supply of General Abercromby's army by inland waters and give him the mobility he required.
If, on the other hand, the transport squadron is able to furnish all the support necessary, the covering squadron will take station as close as possible to the enemy's naval base, and there it will operate according to the ordinary laws of blockade. If nothing is desired but to prevent interference, its guard will take the form of a close blockade. But if there be a subsidiary purpose of using the expedition as a means of forcing the enemy to sea, the open form will be employed; as, for instance, in Anson's case above cited, when he covered the St. Malo expedition not by closely blockading Brest, but by taking a position to the eastward at the Isle de Batz.
In the Japanese operations against Manchuria and the Kuantung Peninsula these old principles displayed themselves in undiminished vitality. In the surprise descents against Seoul and at Takusan the work of support was left entirely with the transport squadron, while Admiral Togo took up a covering position far away at Port Arthur. The two elements of the fleet were kept separate all through. But in the operations for the isolation and subsequent siege of Port Arthur they were so closely united as to appear frequently indistinguishable. Still, so far as the closeness of the landing place to the objective permitted, the two acted independently. For the actual landing of the Second Army the boats of the covering squadron were used, but it remained a live naval unit all through, and was never organically mingled with the transport squadron. Its operations throughout were, so far as modern conditions permit, on the lines of a close blockade.
To prevent interference was its paramount function, undisturbed, so far as we are able to judge, by any subsidiary purpose of bringing the enemy to decisive action.
All through the operations, however, there was a new influence which tended to confuse the precision of the old methods. Needless to say it was the torpedo and the mine. Their deflective pressure was curious and interesting. In our own operations against Sebastopol, to which the Port Arthur case is most closely comparable, the old rules still held good. On the traditional principle, dating from Drake's attack on San Domingo in 1585, a landing place was chosen which gave the mean between facility for a coup de main and freedom from opposition; that is, it was chosen at the nearest practicable point to the objective which was undefended by batteries and out of reach of the enemy's main army.
In the handling of the covering squadron Admiral Dundas, the Commander-in-Chief, gave it its dual function. After explaining the constitution of the transport squadron he says, "The remainder of my force ... will act as a covering squadron, and where practicable assist in the general disembarkation." With these two objects in mind he took a station near enough to the landing place to support the army with his guns if it were opposed, but still in sight of his cruisers before Sebastopol, and at such a distance that at the first sign of the Russians moving he would have time to get before the port and engage them before they could get well to sea; that is, he took a position as near to the army as was compatible with preventing interference, or, it may be said,
his position was as near to the enemy's base as was compatible with supporting the landing. From either aspect in fact the position was the same, and its choice presented no complexity owing mainly to the fact that for the first time steam simplified the factors of time and distance.
In the Japanese case the application of these principles was not so easy. In selecting the nearest undefended point for a landing, it was not only batteries, or even the army in Port Arthur, or the troops dispersed in the Liaotung Peninsula that had to be considered, but rather, as must always be the case in the future, mines and mobile torpedo defence. The point they chose was the nearest practicable bay that was unmined. It was not strictly out of mobile defence range, but it so happened that it lay behind islands which lent themselves to the creation of fixed defences, and thus it fulfilled all the recognised conditions. But in so far as the defences could be turned by the Russian fleet a covering squadron was necessary, and the difficulty of choosing a position for it was complicated by the fact that the objective of the combined operations was not merely Port Arthur itself, but also the squadron it contained. It was necessary, therefore, not only to hold off that squadron, but to prevent its escape. This indicated a close blockade. But for close blockade a position out of night torpedo range is necessary, and the nearest point where such a position could be secured was behind the defences that covered the disembarkation. Consequently, in spite of what the strategical conditions dictated, the covering squadron was more or less continuously forced back upon the army and its supporting force, even when the support of the battle-squadron was no longer required.