the statements made to the effect that it was he who was really in command on the day of the engagement, Admiral Ting deferring to his suggestions.  I am in no position to affirm whether this is really the truth or not, but if it be indeed the fact, it cannot be held to be astonishing that disaster should have overtaken a fleet manoeuvred by a soldier!  I recollect that Mr. Purvis also informed me that the boilers of two or three of the vessels (instancing the destroyed Chao-Yung) were worn-out and unfit for service.  Laxity of discipline, too, seems to have resulted in disobedience or disregard of orders.  As an instance of this, it is alleged that instructions telegraphed from the conning-tower of the flagship were varied or suppressed by the officer at the telegraph, and that a subsequent comparison of notes with the engineer afforded proof of this.

I was forcibly struck by the comparatively unimportant part played in this action by that “dark horse” of modern naval warfare, the dreaded and much-discussed torpedo.  Both squadrons had several torpedo-boats present, though, as I have shown, those on the Chinese side did not enter the action until it had been proceeding more than an hour.  The Japanese allege that they did not use the torpedo at all during the action, and however this may be, there is nothing to show that the weapon made on either side a single effective hit.  I drew the impression from what I saw, that it would be apt to be ineffectual as used by one ship against another, an antagonist in the evolutions of the combat, as the prospect of hitting, unless the ships were very close together, would be small.  The specially-built boat, running close in, and making sure of the mark, would of course be dangerous, although the storm of shot from the quick-firing guns ought even in that case to be a tolerably adequate protection.  The torpedo undoubtedly was not given a fair chance at the battle of Yalu, but the result seems to indicate that its terrors have been overrated, that artillery must still be reckoned the backbone of naval warfare.  Probably the torpedo will turn out to be most effective in surprise attacks on ships and fleets at anchor.  The experience of Wei-hai-wei seems to point to this.

CHAPTER III

It was dark long before we got back to the bay where we had anchored the Columbia, and we might have found it impossible to make out her whereabouts if Webster had not hoisted lights to guide us.  When again aboard we got up steam and stood out to sea.  We should have run for the Yellow Sea at once but for the presence of the Chinese agent, whom we had had no opportunity of transferring from the Columbia.  A motion to throw him overboard was negatived, and we resolved to hold on for Port Arthur, where we could get rid of him without going much out of our way.  Besides, we felt curious to see if any further encounter would take place between the hostile squadrons.  Such, however, was not fated to be the case.  The Japanese allege that they intended to renew the attack in the morning, and tried with that view to hold a course parallel with that of the retreating Chinese, but lost them during the night.