nection of the axis in peritrochio, the lever and lantern wheel or pinion put together like one solid and compact body, they all revolve together with two distinct motions; viz. a rotatory one on their own axis, and a progressivę circumvolutionary one on the ring or circle, constantly acting upon and impelling the cog-wheel and spindle by their united powers and combined actions,

In witness whereof, &c.

Specification of the Patent granted to GEORGE SAvage, of Huddersfield, in the County of York, Watchmaker; for a new Method of regulating or equalizing the Force or Power of the Main-spring in Watches or other similar Machines for measuring Time.

Dated January 26, 1808.

With Engravings.

To all to whom these presents shall come, &c. NOW KNOW YE, that in compliance with the said proviso, I the said George Savage do hereby declare, that my said invention is accomplished by introducing a minor spring, to be wound up by the major or main spring at certain periods. The drawing in the margin of these presents exhibits a method of effecting the desired end: it shews that, as the wheels revolve, the pins placed in a small circle in the hour or centre wheel, as denoted by the dots at A, Fig. 1. (Plate XIII.) force up at certain periods, according to the number of pins inserted, a lever

1

which is connected with the detent or small pallet c, and thereby unlock the quarter-piece d, Fig. 2, which covers the minor-spring, and the lever b is brought again to its proper place by the pressure of the spring f. This

operation

operation causes the major or main-spring to wind up the minor-spring situated, according to this drawing, under the quarter-piece, and administers an equal impelling power to move the train of wheels: e exhibits a bridge, the same as is commonly used in watches, made to denote seconds of time, upon which the minor-springbarrel moves, In witness whereof, &c.

Mode of constructing the recoiling Scapement. Communicated by Mr. ROBERT KING, of Scarborough, In a Letter to the Editors,

Gentlemen,

With an Engraving.

HAVING AVING perused, in the seventh number of the Retrospect of Philosophical, &c. Discoveries" an account of the construction of the wheel and pallet_for Graham's dead-beat escapement, but which is erroneous in the opening between the pallets, no allowance being made for the motion of the pendulum; and as the Editors of that work, in their remarks on that account, express a wish that some person of established judgment and considerable practice, would in like manner lay down rules for the construction of the teeth and pallets of other most approved escapements, I have transmitted you the following:

To give a maintaining power to perpetuate the vibrat ing motion of pendulums in the most equable manneṛ possible, has always been one of the desiderata among mechanics. And the angle of the inclined plane on

which the teeth of the wheel acts, has not been agreed on by the several artists employed in the making of clocks and watches, or other time-pieces. The angle of sixty degrees has been generally used; but with recoiling pallets acting on a tooth during the recoil, the com→ plement of the angle will in that case be the angle of the inclined plane, which, being only thirty degrees, will not push back the maintaining power with equal ease during the action of the pendulum on the wheel, as it is pushed forwards for the maintaining of the motion; and consequently, every vibration of the pendulum is checked, and not suffered to have its full swing, and yet gives no correction thereby towards the isocronisme of the motion, or equalization of the times of the vibrations. Now, to give the pendulum the greatest freedom of swing with recoiling pallets, the face of the pallets ought to be at the angle of forty-five degrees to the tangent of the acting tooth or point in every part of its action; or at least to the point where the action begins. This angle was adopted by Mr. Henry Hindley of York, as may be seen by examining the scapements of clocks made by him. To draw this scapement, having drawn the wheel, and divided it into the number of teeth proposed, of the same size you intend in the work (which in the example here given is a wheel of thirty, that number being most common for a pendulum beating seconds) you next consider the number of teeth between the planes or faces of the pallets, which in this figure I have made seven and a half, as may be seen by counting the spaces which will be the distance between, the points a, and B, in the periphery of the wheel. (See Plate XIII. Fig. 4.) Then from the centre C draw the lines Cc, and C d through the points a, and B, which will be two radii

at

at the angle of 90° or 7 teeth and a half. Then to these lines draw the perpendiculars a D and B D-intersecting each other in the point D, which is to be the center of the verge; then through the point a, with the angle of forty-five degrees to the line a D, draw the line g ħ, which will be the line of the face of that pallet; then from the centre C draw the line Cƒ, at the distance of half a tooth from the point a, and with the distance af, set off from the point a, it will give the point of the pallet e; then draw the line D e, which will give the angle e Df, or the angle of impulsive swing, which is equal to the distance to be set off from the point B to b, which set off from B along the line C d, give the place in the face of the other pallet, where the tooth will strike it next after the escape: at this point b, the face of the pallet must be drawn at an angle of forty-five degrees from the line b D, so that the point of the pallet may be free from the circumference of the wheel; for, when the point of the tooth at a, shall have slided along the face of the pallet to e, where it escapes the pallet, then the point b will, by the elevation of the point e of the other pallet, be brought into the line of the circumference of the wheel, and meet the point of the tooth i, which will recoil upon it, until the force of the pendulum is lost, and it is again acted on by the pallet, and so proceed to the point of escapement, where it quits the pallet at the same distance from the verge. In these pallets the line of impulse is equal in each pallet, and the pendulum has the same power on the wheel in the recoil, as the wheel has on the pendulum by its forward action on the pallet; and the wheel acts with equal force on each pallet, as being at equal distance from the verge. And the pressure of the wheel against the pal

let