The advent of the torpedo, however, has given the idea a new importance that cannot be overlooked.  The degree of that importance is at present beyond calculation.  There is at least no evidence that it would be very high in normal conditions and between ordinarily efficient fleets.  The comparative success of the opening Japanese attack on the Port Arthur squadron is the only case in point, and where only one case exists, it is necessary to use extreme caution in estimating its significance.  Before we can deduce anything of permanent value we must consider very carefully both its conditions and results.

To begin with, it was a new experience of a new class of weapon, and it by no means follows that the success of a new expedient will be repeated with anything like equal result.  It will not be irrelevant again to recall the case of fireships.  At the outset of the sailing era in 1588, this device prepared the way for a decisive success against a fleet in the open.  In the succeeding wars the new weapon found a prominent place in the organisation of sea-going fleets, but its success was never repeated.  Against ships in ill-defended harbours it did occasionally produce good results, and during the infancy of tactics its moral and even material effects in fleet actions were frequently demonstrated.  But as naval science developed and the limitations of the weapon were more accurately measured, it was able to achieve less and less, till in the eighteenth century it was regarded as almost negligible.  Even its moral effect was lost, and it ceased to be considered as a battle unit.

Now, if we examine closely the Port Arthur case, we shall find it pointing to the existence of certain inherent conditions not dissimilar from those which discredited fireships as a decisive factor in war.  In spite of the apparently formidable nature of a surprise attack by torpedo the indications from the one case in point are that these conditions make for greater power in the defence than in the attack.  The first condition relates to the difficulty of locating the objective accurately.  It is obvious that for this kind of operation the most precise intelligence is essential, and of all intelligence the most difficult to obtain in war is the distribution of an enemy’s fleet from day to day.  The Japanese had fairly certain information that the bulk of the Port Arthur squadron was lying in the outer anchorage, but it had been constantly moving, and there was a report that three battleships had just been detached from it.  The report was false, but the result was that of the five divisions of destroyers which the Japanese had available, two were diverted against Dalny, where no enemy was found.  Such uncertainty must always exist, and in no circumstances is it likely to be less than where, as in the Japanese case, the attack is made before declaration, and while the ordinary channels of intelligence are still open.