No. VI. in the foregoing table requires some explanation. To heel a ship over to a certain angle a certain amount of _work_ must be done, and in the scientific sense _work_ is done only when something is moved through a space against a resistance. When the weight of a ton is raised 1 ft. high, one foot ton of work is said to be done; if 2 tons were raised 1 ft., or 1 ton were raised 2 ft., then two foot-tons of work would be done, and so on. The same would be the case if a pressure equal to those weights were applied so as to move a thing in any direction through the same distances. It should be carefully noticed that the foot-ton is quite a different unit in this case from what it is as the moment of a couple. If we heel a ship over by applying a pressure on the masts, it is plain that the pressure must act through a certain space, and the same heel could be caused either by means of a smaller pressure or a greater, according as we apply it higher up or lower down; but the space through which it must act would vary, so that the product of the pressure and space would, however, be always the same. No. VI. shows the amount of work that would have to be done in order completely to upset each of the vessels when already steadily heeled over to 14°. The amounts in the two cases are so different that we can easily understand how a squall which would not endanger the _Monarch_ might throw the _Captain_ over. A squall suddenly springing up would do more than heel a vessel over to the angle at which it is able to maintain it: it would swing it beyond that position by reason of the work done on the sails as they are moving over with the vessel, and the latter would come to a steady angle of heel only after a series of oscillations. Squalls, again, which, although suddenly springing up in this manner, could not heel the ship over beyond the angle where the stability vanishes, might yet do so if they were intermittent and should happen to coincide in time with the oscillations of the ship—just as a series of very small impulses, coinciding with the time of the vibrations of a heavy pendulum, may accumulate so as to increase the range of vibration to any extent. It is believed that in the case of the _Captain_ the pressure of the wind on the underside of the hurricane assisted in upsetting the vessel. This, however, could only have exerted a very small effect compared to that produced by the sails. The instability of the _Captain_ does not appear to have been discovered by such calculations as were made before the vessel went to sea. It was observed, however, that the ship when afloat was 1 ft. 6 in. deeper in the water than she should have been—in other words, the freeboard, or side of the ship out of the water, instead of being 8 ft. high as intended, was only 6 ft. 6 in., and such a difference would have a great effect on the stability. [Illustration: FIG. 76.—_H.M.S. Glatton._ ]