where in the air, and free from all contact with any part of the apparatus, it attains an higher elevation without a possibility of friction or wear.

I have described my double capstan as it is to be used in large vessels, where messengers are necessary, from the great size of the cables; but it is obvious that it is equally applicable to smaller vessels, as their cables can be managed with it in the same manner as is directed for the messenger. The same principle may also be easily applied in windlasses, by having a small horizontal barrel placed parallel to the body of the windlass, and having both fitted with rings, in the same way as the capstan already described. The proper place for the small horizontal barrel is forward, just before the windlass, and as much below its level as circumstances will admit; it should be furnished with catch-palls as well as the windlass.

Besides the advantages already stated, my proposed improvement to the capstan has others of considerable utility. Its construction is so very simple, that it is no more liable to derangement or injury than the capstan itself. Its cost can be but small, and every part of it can be made by a common ship carpenter, and be repaired by him at sea if damaged by shot. It will take up but little room, only that of a half-barrel cask; and it is of a nature so analogous to that kind of machinery to which sailors are accustomed, that it can be readily understood and managed by them.

In order to render the description of my double capstan more clear, I annex a sketch of it, as fitted up inR the manner proposed.

VOL. XIII.-SECOND SERIES.

E e

Re

Reference to the Engraving of Mr. BOSWELL's improved Capstan, to prevent the Necessity of Surging. (Plate VIII. Fig. 4.)

A, represents the larger or common capstan used on board ships.

B, another capstan of less dimensions, placed in a similar manner.

C, the coils of the messenger passing alternately round the large and small capstans, but with their direction reversed on the different barrels, so that they may cross each other in the interval between them.

DDDD, projecting rings round each capstan or barrel, so fixed on the two barrels, that those on one barrel should be exactly opposite the middle of the intervals between those on the other barrel.

On the Construction of the Teeth of Escapements.
With an Engraving.

Communicated by Mr. ROBERT KING, of Scarborough, in a Letter to the Editors.

GENTLEMEN,

IN order that the principle for the escapement of pallets, as given in my last paper, inserted in your 71st number, may be clearly understood, and more easily applied to wheels of different numbers of teeth, and to pendulums of different lengths, a farther illustration may be made, in observing, that as Mr. Ferguson, in his Select Exercises, proposes a clock of a very simple construction, but seems apprehensive that it would be difficult to make an escapement for his wheel of 90 teeth

with proper motion; it may not be improper to shew how that escapement may be effected. We are to observe, that the impulsive angle or half-swing of the pendulum of the clock is nearly half a tooth; and so by taking fewer teeth, or shortening the extent of the fangs or inclined planes so as to be nearer the centre of the verge, the vibrations may be increased at pleasure. For as the proposed swing or vibration from the centre of the swing in the pendulum is to half a tooth the impulsive angle on the pallets, so is the length of the pendulum to the distance of the point of percussion on the plain of the pallet to its distance from its centre or the verge of the pallets; so that you must make the pallets to escape such a number of teeth as will give the proposed vibration. To effect this, the following shews the method at large. Ferguson's clock having a swingwheel of two inches and a quarter diameter, the circumference will be .7.0686 inches; and being to have 90 teeth, each tooth will be .07854 of an inch, and the half tooth,03927 of an inch. And the pendulum being designed to beat seconds, will in this part of the world be nearly 39,2 inches in length. Now, suppose I should wish to apply it with four inches vibration, it would then vibrate two inches on each side of the centre; then the proportion would stand as two inches the vibration on one side of the perpendicular is to the length of the whole pendulum, or 39,2 so is ,03927 the space taken up by half a tooth, or the length of the line impulse, on the face of the pallet (which will be near enough, it allowing a small drop of the difference between the slope and its perpendicular for inequalities of workmanship) to 744692 of an inch, or half the opening of the pallets, or the distance from the point of impulse to the centre