ships employed in a time of war under the Queen's flag. No amount of resistance which an armour plate could show, would give the satisfaction which I received from the excellent behaviour of the simple combination of fuel and thin loose iron sheets. Coal armour and torpedoes together have given to the fast merchant shipping of England a significance in warfare wholly new." In connection with this point we may remark that in one of the best papers read at the meetings, and in which Mr. White of the Admiralty describes the structural arrangements of H.M.S. Iris, it was shown that that vessel has her bunkers so arranged as to completely surround the engine and boiler rooms, although they are of an aggregate length of 138 feet out of the total length of the vessel, which is 300 feet. The Iris is an unarmoured warvessel and enjoys the distinction of being the fastest sea-going vessel in the world. We must, however, demur to the high estimate put upon coal armour, for the simple reason that one cannot be certain that it will be in the ship when wanted. If in the wars of the future our ships are to choose just their own time and place for fighting, and can comfortably and quietly, from a tender in waiting, fill up all their bunkers before going into action, the coals may afford a valuable defence against the enemy's fire, but if the bunkers are half empty the case will be altered very much for the worse. We have further in the paper a brief account of the more recent experiments on steel armour, which we summarised in our April number. The future of the compound iron and steel plates will, we suppose, be largely dependent upon the results of the trials of them so made in connection with the Inflexible's turret plates. The conditions laid down with regard to them are, "that a piece about 6 feet long, cut off the bent plate of the turret, and secured to backing, shall neither be perforated nor cracked through by the shot which would perforate a similar piece of iron; and that if the edges of the test piece are supported by a frame no one of three such shots planted at intervals of two feet shall get through." It is also stated that the Italian Government intend shortly to test, with one of their large guns, some armour plates made wholly of steel, simply cast, and produced at the iron works at Terre Noire and Creusot in France. If the results of these trials are at all

D

satisfactory, the price of armour will be much reduced, will be in

fact but that of large castings.

Perhaps most of our readers may

think with us that this will be a doubtful advantage.

Sir Spencer Robinson's paper consists chiefly of a detailed history of the armouring of war-ships, and an elaborate resume of experiments on armour plates from as far back as 1863 to 1877. As might be expected from the prominent part the writer has taken in the construction of war-ships, there is much of considerable interest; there is so much, however, altogether, that we cannot attempt to give even an abstract of it, but must confine ourselves to stating the conclusions to which he appears to have come. These are that sandwich armour will prove to be a failure, and that the most effective defence for ships would be provided by a compound backing after a plan invented and patented by a Mr. Hughes, and known as the Millwall system. An experimental target was made upon this plan, and tried at Shoeburyness about ten years ago. The arrangement of the backing is thus described:"Behind the armour plates are a series of hollow stringers seven inches deep, running horizontally, their outer faces in contact with the inner part of the armour plates. The inner part of the stringer has two flanges rivetted to a skin formed of two thicknesses of -inch iron. Behind this there is a vertical hollow stringer of the same description, the hollow part turned outwards. Both stringers where hollow are filled with wood. At the back of these vertical stringers or frames there is a covering plate or inner skin of 1-inch iron."

It is hardly necessary to remark that a ship constructed on this plan would be a still more expensive article than our present very dear ironclads are, unless it would result in a very material decrease in the weight, and consequently the size, &c., of the ship. The experiments on the Millwall shield appear, so far as we can see, to have been very inconclusive, and it would seem to be desirable that advantage should be taken of the coming trials of steel armour to test the value of Mr. Hughes's, or some other perhaps less expensive but equally efficient system of compound backing. In the early days of armour plating a number of targets were tried in which there was no backing, and all sorts of substances were

tried as backing. The Warrior had a wood backing four times the thickness of her armour plates, and it may be said that the Warrior target beat all its competitors as regards resistance to shot. It was conclusively proved that much more effective resistance to shot was obtained with some kind of cushion behind the armour than with an unbacked plate, but it was soon seen that the Warrior had more backing than was really necessary, and in the Minotaur and other vessels only nine inches of wood backing was used and a five and a-half inch substituted for a four and a-half plate, the total weight per square foot of surface remaining the same. When the question of the desirability of having a mixed backing of wood and strips of iron was raised, it was said in answer that the best thing is to put as much iron in one solid plate as possible rather than increase the weight of backing. The resistance to penetration varied as the square of the thickness, and consequently a seven inch armour plate, for instance, was worth double as much as a five inch plate. Now we have got to very much thicker armour, and it is found that the effective resistance of very thick plates does not follow the same law, but varies more nearly as the thickness than as the square of the thickness. It may be then that more iron may be put into the backing with as much advantage as if put in the solid plate. The only way to be certain is to have some experimental tests.

Sir Spencer Robinson's objection to the so-called sandwich system of armour, in which, as explained in our former article, two armour plates are used with a layer of teak between as well as behind them, appears to be chiefly that a shell exploding in the wood backing between the two plates will blow the outer one bodily off the target, and it is even said that a guncotton shell has been designed for the special purpose of doing so. The advantage gained by having two thin rather than one thick plate is that, in the present stage of armour plate manufacture, much wider plates can be obtained when they are of the lesser thickness, and thus the great evil of numerous joints in the armour is avoided.

THE NATURE AND MOTIONS OF ATMOSPHERICAL DEPRESSIONS.-PART II.

N turning from the theoretical view to treat of the weather conditions experienced over these islands, it would have been desirable to have illustrated the subject by means of the "Daily Weather Charts," but as that would be impracticable within the compass of an article, I must content myself by endeavouring to make my meaning clear without their aid.

It must be apparent to everyone who has given the least attention to our argument, that the depressions with which we are acquainted are quite different in their general character from those formed by the sun and moon within the tropics; and this will be further evident from a variety of considerations. The depressions in the vicinity of the equator have east winds both to the north and south of them, whilst the cyclones in our neighbourhood have always east winds on their northern and west winds on their southern sides. In the calm belt or region of equatorial depressions cyclones or revolving storms are unknown, and the reason of this will readily appear by means of the experiment of the rotating hemispheres. If the two outer hemispheres, A and B,* be made to rotate, the one with, the other against watch hands-and we know that these are the conditions of the circulation north and south of the equator-the central hemisphere C will remain at rest. We thus see that a cyclone circulation is impossible under such circumstances, and that a revolving storm can only happen between two rotating waves which are revolving in the same manner— either both with or both against watch hands. In one word the equatorial depressions produce the circulation, whereas the cyclones are the result of it. The winds do not circulate round the equatorial depressions; a true depression, without any progressive movement, would only give rise to waves or oscillations. If, however, it should possess a progressive movement, it would produce two separate systems of circulation, one on each side of the disturbing cause, and revolving in contrary directions.

*See Nautical Magazine, January, 1879, page 13.

If we form a miniature whirlpool in a cup of tea, it is quite clear that the depression or hollow noticed in the middle is the result of the circulation, for the faster the liquid is made to move the deeper the hollow becomes. The circulation really produces two effects whilst the centre is depressed the sides are elevated. I am inclined to the opinion that whirlpools are made up of a number of rotating waves as shown in Figures 1, 2, and 3.

Figure 1 represents the falling sides of the waves or motion away from the centre; Figure 2 the rising sides of the waves or motion towards the centre; Figure 3 shows the whole of the waves, the broken line showing the falling side and the continuous line the rising side of the waves. The large arrows show the direction in which the waves are revolving.

In each section shown in Figure 3 there is therefore a rising and falling motion of the water-the falling motion, I believe, giving