We remarked in a previous chapter that the lifting planes were sometimes on the wheel and sometimes on the anchor. In another chapter we pointed out clearly that the run on the locking surface of the pallets had an important bearing on the freedom of the escapement and hence on the rate of the dead beat escapement. In considering the cylinder escapement, so common in carriage clocks, we shall find that the lift is almost entirely on the curved planes of the escape wheel, and that the locking planes are greatly extended, so that they form the outer and inner surfaces of the cylinder walls. Thus we have here a form of the dead beat escapement, which embraces but one tooth of the escape wheel and is adapted to operate a balance instead of a pendulum. Therefore the points for us to consider are as before, the amount of lift, lock, drop and run, and the shapes of our escape wheel teeth to secure the least friction, as our locking surfaces (the run) being so greatly extended this matter becomes important.

Action of the Escapement.—Fig. 52 is a plan of the cylinder escapement, in which the point of a tooth of the escape wheel is pressing against the outside of the shell of the cylinder. As the cylinder, on which the balance is mounted, is moved around in the direction of the arrow, the wedge-shaped tooth of the escape wheel pushes into the cylinder, thereby giving it impulse. The tooth cannot escape at the other side of the cylinder, for the shell of the cylinder at this point is rather more than half a circle; but its point locks against the inner side of the shell and runs there till the balance completes its vibration and returns, when the tooth which was inside the cylinder escapes, giving an impulse as it does so, and the point of the succeeding tooth is caught on the outside of the shell. The teeth rise on stalks from the body of the escape wheel, and the cylinder is cut away just below the acting part of the exit side, leaving for support of the balance only one-fourth of a circle, in order to allow as much vibration as possible. This will be seen very plainly on examining Fig. 53, which is an elevation of the cylinder to an enlarged scale.

Proportion of the Escapement.—The escape wheel has fifteen teeth, formed to give impulse to the cylinder during from 20° to 40° of its vibration each way. Lower angles are as a rule used with large than with small-sized escapements; but to secure the best result either extreme must be avoided. In the escapement with very slight inclines to the wheel teeth, the first part of the tooth does no work, as the tooth drops on to the lip of the cylinder some distance up the plane. On the other hand, a very steep tooth is almost sure to set in action as the oil thickens. The diameter of the cylinder, its thickness and the length of the wheel teeth are all co-related. The size of the cylinder with relation to the wheel also varies somewhat with the angle of impulse, a very high angle requiring a slightly larger cylinder than a low one. If a cylinder of average thickness is desired for an escapement with medium impulse, its external diameter may be made equal to the extreme diameter of the escape wheel multiplied by 0.115.

Then to set out the escapement, if a lift of say 30° be decided on, a circle on which the points of the teeth will fall is drawn within one representing the extreme diameter of the escape wheel, at a distance from it equal to 30° of the circumference of the cylinder. Midway between these two circles the cylinder is planted (see Fig. 54). If the point of one tooth is shown resting on the cylinder, a space of half a degree should be allowed for freedom between the opposite side of the cylinder and the heel of the next tooth. From the heel of one tooth to the heel of the next equal 24° of the circumference of the wheel, 360÷15=24°, and from the point of one tooth to the point of the next also equals 24° so that the teeth may now be drawn. They are extended within the innermost dotted circle to give them a little more strength, and their tips are rounded a little, having the points of the impulse planes on the inner or basing circle. The backs of the teeth diverge from a radial line from 12° to 30°, in order to give the cylinder clearance, a high angled tooth requiring to be cut back more than a low one. A curve whose radius is about two-thirds that of the wheel is suitable for rounding the impulse planes of the teeth. The internal diameter of the cylinder should be such as to allow a little freedom for the tooth. The rule in fitting cylinders is to have equal clearance inside and outside, so as to equalize the drop. The acting part of the shell of the cylinder (where the lips are placed) should be a trifle less than seven-twelfths of a whole circle, with the entering and exit lips which are really the pallets, rounded as shown in the enlarged plan, Fig. 55, the entering lip or pallet rounded both ways and the exit pallet rounded from the inside only. This rounding of the lips of the cylinder adds a little to the impulse beyond what would be given by the angle on the wheel teeth alone. The diameter of the escape wheel is usually half that of the balance, rather under than over.

Size of Cylinder Pivot.—To establish the size of the pivot with relation to its hole is apparently an easy thing to do correctly, but to an inexperienced workman it is not so. The side shake in cylinder pivot holes should be greater than that for ordinary train holes; one-sixth is the amount prescribed by Saunier; the size of the pivot relatively to the cylinder about one-eighth the diameter of the body of the cylinder. It is very necessary that this amount of side shake should be correctly recognized; if less than the amount stated, the escapement, though performing well while the oil is fresh, fails to do so when it commences to thicken.

When the balance spring is at rest, the balance should have to be moved an equal amount each way before a tooth escapes. By gently pressing against the fourth wheel with a peg this may be tried. There is generally a dot on the balance and three dots on the plate to assist in estimating the amount of lift. When the balance spring is at rest, the dot on the balance should be opposite to the center dot on the plate. The escapement will then be in beat, that is, provided the dots are properly placed, which should be tested. Turn the balance from its point of rest till a tooth just drops, and note the position of the dot on the balance with reference to one of the outer dots on the plate. Turn the balance in the opposite direction till a tooth drops again, and if the dot on the balance is then in the same position with reference to the other outer dot, the escapement will be in beat. The two outer dots should mark the extent of the lifting, and the dot on the balance would then be coincident with them as the teeth dropped when tried in this way; but the dots may be a little too wide or too close, and it will therefore be sufficient if the dot on the balance bears the same relative position to them as just explained; but if it is found that the lift is unequal from the point of rest, the balance spring collet must be shifted in the direction of the least lift till the lift is equal. A new mark should then be made on the balance opposite to the central dot on the plate.

When the balance is at rest, the banking pin in the balance should be opposite to the banking stud in the cock, so as to give equal vibration on both sides. This is important for the following reason. The banking pin allows nearly a turn of vibration and the shell of the cylinder is but little over half a turn, so that as the outside of the shell gets round towards the center of the escape wheel, the point of a tooth may escape and jam the cylinder unless the vibration is pretty equally divided. When the banking is properly adjusted, bring the balance round till the banking pin is against the stud; there should then be perceptible shake between the cylinder and the plane of the escape wheel. Try this with the banking pin, first against one and then against the other side of the stud. If there is no shake, the wheel may be freed by taking a little off the edge of the passage of the cylinder where it fouls the wheel, by means of a sapphire file, or a larger banking pin may be substituted at the judgment of the operator. See that the banking pin and stud are perfectly dry and clean before leaving them: a sticky banking often stops a clock when nearly run down. Cylinder timepieces, after going for a few months, sometimes increase their vibration so much as to persistently bank. To meet this fault a weaker mainspring may be used, or a larger balance, or a wheel with a smaller angle of impulse. By far the quickest and best way is to very slightly lap the wheel by holding a piece of Arkansas stone against the teeth, afterwards polishing with boxwood and red stuff. So little taken off the wheel in this way as to be hardly perceptible will have great effect.

Sometimes the escape wheel has too much end shake. We must notice in the first place how the teeth are acting in the cylinder slot. Suppose, when the escape wheel is resting upon its bottom shoulder, the cylinder will ride upon the plane of the wheel, which will cause it to kick or give the wheel a trembling motion, then we know that the cylinder is too low for the wheel; therefore, we have not only to lower the escape top cock in order to correct the end shake, but we must also drive the bottom cylinder plug out a little in order to raise the cylinder sufficient to free it from the plane of the wheel. Now, if the end shake of the cylinder is correct previous to this, we shall now either have to raise the cock or drive the top plug in a little. But suppose the end shake of the escape pinion is excessive, and is, when the bottom shoulder is resting on the jewel, a little too low so that the bottom of the escape wheel runs foul of the cylinder shell; in this case we simply drive out the steady pins from the bottom escape wheel cock and file a piece off the cock, leaving it perfectly flat when we have enough off. We then insert the steady pins again, screw it down, and if the end shake is right, the escapement is mostly free and right also.

Now let us consider the frictions; there is the resistance of the pivots, which depends on their radius, on the weight of the balance, the balance spring, the collet, and the weight of the cylinder; these are called locking frictions. Then there are those of the planes, of the teeth of the wheel, of the lips of the cylinder. It is on these that the change and destruction of the cylinder are produced. To prevent this destruction, it is necessary to render the working parts of the cylinder very hard and well polished, as well as the teeth of the escape wheel.

The oil introduced in the cylinder is also a cause as in the dead beat. It may thicken; the dust proceeding from the impact of the escapement forms with the oil an amalgam which wears the cylinder. The firmness and constancy of the cylinder depend on the preservation and fluidity of the oil.

Then there are the accidental frictions; the too close opening of the cylinder, the play of the balance and of the wheel, with the thickening of the oil, changes the arc of vibration a good deal; the teeth of the wheel may not be sufficiently hollowed, so that the cylinder can revolve in the remaining space, for the oil with the dust forms a thickness which also changes the vibration. The drop should not be too great, for it is increased by the thickening of the oil and impedes the vibration.

Examination of Clocks.—In this particular escapement, when used for larger timepieces than watches, it is astonishing the variety of methods which are employed, yet the same results are expected. In examining such clocks we will first notice that the chariot, cock, etc., are so placed, many of them, that the last wheel in the train is a crown wheel, hence it is made to work at 90° with the escape wheel pinion which is set at right angles with the crown wheel pinion, and, as a matter of course, the cylinder is also set the same way. Now, this arrangement needs especial care, for it is quite natural that when the entire friction of the cylinder is only on the bottom part of the bottom pivot, the clock is sure to go faster than when the whole length of both pivots are more in contact with their jewel holes, which is always the case when the cylinder is parallel with all the pinions, instead of standing upon one pivot only. Now, although there must of necessity be a very great difference in timing the clock in the two different positions, yet we find no difference in the strength of mainspring or any part of the train, which is a mistake, for the result is simply this: the clock will gain time for the first few days after winding, and will then gradually go slower and slower until the mainspring is entirely exhausted. It is not very difficult to ascertain why it goes so fast after winding, for then the whole tension of the spring is on, and as there is not sufficient friction on the point of one pivot to counteract this, the banking pin is almost sure to knock, and will continue to knock for the first few days until a part of the spring’s pressure is exhausted. Now, in this case the knocking of the banking pin alone would cause the clock to gain time, even if the extra tension of the mainspring did not assist it to do so. Hence, on the whole, the result is anything but satisfactory, for such a clock can never be properly brought to time.

Having said this much about the fault (which is entirely through the want of a little forethought with the manufacturer), I will give as good a remedy as I can suggest to give the reader an idea of how these faults may be put to right, if he is willing to spend the time upon them. In the first place take out the cylinder and make the bottom pivot perfectly flat instead of leaving it with a round end, as they are mostly left, which only allows just one part of the pivot to be in contact with the endstone. By leaving this pivot flat on the bottom, there is more surface in contact; hence, in a sense, more friction.

In some cases the whole pivot left flat would not be sufficient to retard the mainspring’s force; then we must resort to other methods to effect a cure.

Well, our next method in order to try and get the clock to be a uniform timekeeper, is to change the mainspring for one well finished and not quite so strong as the original one. Perhaps some will say “why not do this before we go to the trouble of flattening the bottom pivot?” Just this; when a pivot is working only upon the bottom it is best to have a flat surface to work upon, as the balance is then oscillated with more uniformity, even when the mainspring is not exactly uniform in its pressure; therefore we do no harm—but good—by making the bottom pivot flat, and this alone will sometimes be sufficient to cure the fault of the banking knocking if nothing else.

To my mind, when such strong mainsprings are used as we generally see in this class of timepiece, neither of the jewel holes or pivots should be so small as they usually are. Fancy such small pivots as are mostly seen upon the escape wheel pinion being driven by such a strong mainspring. If we allow the clock to run down while the escape wheel is in place, we are very liable to find one or both pivots broken off before it gets run down. I think all such pivots ought to be sufficiently strong to stand the pressure of the mainspring through the train of wheels without coming to grief. But there is another reason why these pivots are liable to get broken off while letting the train run down: that is, the badly pitched depth we often find in the crown wheel and escape wheel pinion. We frequently find too much end shake to the crown wheel which, while resting one shoulder of the arbor against the plate puts the depth too deep, and on the other shoulder the depth is too shallow. Now, when the train is running rapidly this crown wheel is jumping about in the escape wheel pinion, so that the roughness of the running all helps to break off the escape wheel pivots. The best way to correct this depth is to notice how the screws fit in the cylinder plate—for these screws have to act as steady pins as well. If the holes where the screws go through are at all large, we then notice which would be the most convenient side to screw it securely in order to put a collet upon the shoulder of the crown wheel so that the depth will be right by making the end shake right with only fixing a collet to one shoulder. This depth, when correct, will also cause a more uniform pressure upon the escapement, and help to make the clock keep better time. We are supposing that this crown wheel is perfectly true, or it is not much use trying to correct the depth as mentioned above, for even if the end shake be ever so exact and the wheel teeth are out of true, we shall never get the depth to act as it ought, neither can the clock be depended upon for keeping going, regardless of keeping time. When this crown wheel is out of true it is best to rivet it true, not do as I have seen it done, placed in the lathe and topped true, and then the teeth rounded-up by hand. This method simply means a faulty depth after all, for in topping the teeth, those teeth which require the most topping will, when they are finished, be shorter from the top to the base than those which do not get topped so much; therefore, some of the teeth are longer than the others, while the shorter ones are thicker; for when the wheel was originally cut the teeth were all cut alike. These remarks will apply to several kinds of wheels; for whenever a wheel is topped to put it true, we may depend we are making a very faulty wheel of it unless we have a proper wheel cutting machine.

The crown wheel must not be too thick because we will find the tooth to act with the inner edge, and what is left outside only endangers touching the pinion leaf which is next to come into action. Make sure the escape pinion is not too large, which sometimes happens. If it is, it must be reduced in size, or better, put in a new one. The crown wheel holes must fit nicely and the end shake be well adjusted. Do not spare any trouble in making this depth as perfect as you are able, as most stoppages happen through the faults in this place. It would be advisable, when sure the depth is correct, to drill two steady pin holes through the escapement plateau into the edge of the plates. When steady pins are inserted this will always ensure the depth being right when put together.

In some of these clocks it is not only the crown wheel, but frequently the escape wheel has too much end shake. The former, as I have said, can be corrected by making a small collet that will just fit over pivot, fasten it on friction-tight, place the wheel in the lathe and turn the collet down until it is the same size as the other part of the arbor, then run off the end to the exact place for the end shake to be right. If it is properly done and a steel collet is used, it will not be detected that a collet has been put on. Now, when the escape wheel end shake is wrong we have to proceed differently under different circumstances for we must notice in the first place how the teeth are acting in the cylinder slot.

See that the cylinder and wheel are perfectly upright. Suppose, when the escape wheel is resting upon its bottom shoulder, the cylinder will ride upon the plane of the wheel, which will cause it to kick or give the wheel a trembling motion, then we know that the cylinder is too low for the wheel; therefore, we have not only to lower the escape top cock in order to correct the end shake, but we must also drive the bottom cylinder plug out a little in order to raise the cylinder sufficient to free it from the plane of the wheel. Now, if the end shake of the cylinder is correct previous to this, we shall either have to raise the cock or drive the top plug in a little. But suppose the end shake of the escape pinion is excessive, and is, when the bottom shoulder is resting on the jewel, a little too low so that the bottom of the escape wheel runs foul of the cylinder shell; in this case we simply drive out the steady pins from bottom escape wheel cock and file a piece off the cock, leaving it perfectly flat when we have got enough off. We then insert the steady pins again, screw it down, and, if the end shake is right, the escapement is mostly free and right also. It sometimes happens that the wheel is free of neither the top nor bottom plug, but should this be the case, sufficient clearance may be obtained by deepening the opening with a steel polisher and oilstone dust or with a sapphire file. A cylinder with too high an opening is bad, for the oil is drawn away from the teeth by the escape wheel.

If a cylinder pivot is bent, it may very readily be straightened by placing a bushing of a proper size over it.

These clocks are very good for the novice to exercise his skill in order to thoroughly understand the workings of the horizontal escapement. He is better able to see how the different parts act with each other than he is in the small watch. When the escape is correct he will find that the plane of the escape wheel will work just in the center of the small slot in the cylinder.

If he will notice how the teeth stand in the cylinder when the banking pin is held firmly upon the fixed banking pin, it will give him an idea of how this should be. At one side the lip of the cylinder is just about to touch the inside of the escape tooth, but the banking pin just prevents it from doing so, while on the other side the cylinder goes round just far enough to let the point of the next tooth just get on the edge of the slot, but it cannot get in owing to the intervention of the banking pin. If this is allowed to get in the slot just here, we then have what is called “a locking,” which is, in reality, an overturned banking. If the other side is so that the banking pin does not stop it soon enough, the edge of the slot knocks upon the inside of the teeth and causes a trembling of the escape wheel, and the clock left in this form will never keep very good time. We may easily remedy this by taking off the hair spring collet; holding the cylinder firmly in the plyers, and with the left hand turn the balance a little outwards; this will bring the banking pins in contact before the cylinder touches the inside of the wheel teeth, and all is right, providing we are careful in not doing it too much; if so, we shall find the banking knock—a fault which is quite as bad, if not worse, than the one we are trying to remedy. Those particulars are the most important of anything in connection with the cylinder escapement. Yet, as this kind of clock is now being made up at such a low price, these seeming little items are frequently overlooked; hence, when they get into the hands of the inexperienced, there is often more trouble with them than there need be if they knew where to look for some of the faults which I have been endeavoring to bring to light. There are several other things in connection with this particular clock, but we will not comment further just now, but take them up when we are considering the trains, etc.

In the meantime we will resume our study of the cylinder escapement with particular reference to badly worn or otherwise ill fitting escape wheels, as many times, the other points being right, the wheel and cylinder may be such as to give either too great or too small a balance vibration.

A poor motion can also be due to a rough or a badly polished cylinder, but such a cylinder we rarely find. That with a wrong shape of the cylinder lips the motion is not much lessened can be seen in quite ordinary movements, where the quality is certainly not of the best neither are the lips correctly formed, nevertheless they have rather an excessive motion. To cover up these defects in such movements the cylinder wheel teeth are purposely given the shape as shown at B in Fig. 56, and to give sufficient power a strong mainspring is inserted.

With an excessive balance vibration we can usually conclude that it is an intentional deception on the part of the manufacturer, while a poor motion can generally be ascribed to careless methods in making. The continued efforts in making improvements to quicken and cheapen manufacturing processes very frequently result in the introduction of defects which are only found by the experienced and practical watchmaker.

As to the causes which induce excessive balance vibrations? As this defect is generally found in the cheaper grades of cylinder escapements, having usually rather small, heavy, and often clumsy balances, those which have balances whose weight is probably less than they ought to be, need not here be further considered, and it only remains for us to look to the cylinder or the escape wheel for the causes which produce these excessive vibrations. It will be found that the cylinder is smaller in diameter than usually employed in such a size of clock; the escape wheel is naturally also smaller, and its teeth generally resemble B, Fig. 56, while A shows the correct shape of a tooth for a wheel of that diameter.

In using small cylinders we can give the escape wheel teeth a somewhat greater angle of inclination than generally used, but that the proper amount of incline is exceeded is proved by the fact that the balance vibrates more than two-thirds of a turn. It can also be readily seen that with a tooth like B a greater impulse must be imparted than one with an easy curve like A, and the impulse is still further increased as the working width of the tooth B (the lift) is greater, indicated by line b, while the same line in a correct width of tooth, as shown at a, is considerably shorter.

In addition to what has been said of these escapements, we also find them provided with very strong mainsprings to give the necessary power to a tooth like B with its steeply inclined lifting face or impulse angle.

To decrease the great amplitude of the balance vibrations many watchmakers simply replace the strong mainspring with a weaker one. But this procedure is not advantageous as the power of the escape wheel tooth is insufficient to start the balance going and this is due to two causes. First, the great angle of the escape wheel tooth, and secondly, the inertia of the balance. It is only by violently shaking such a clock that, we are enabled to start it going. And the owner soon becomes dissatisfied from its frequent stoppage due to setting of the hands and other causes so that he will be often obliged to shake it until it starts going once more.

For properly correcting these defects the best method to pursue is to replace the cylinder wheel with another one, whose teeth are of the shape as shown at Fig. 55 and without question a good workman will always replace the escape wheel if the clock is of fair quality. But if a low grade one, we would hardly be justified in going to the expense of putting in new wheels, as the low prices for which these clocks are sold preclude such an alteration. As we must improve the wheel some way to get a fair escapement action we can place it in a lathe and while turning, hold an oil stone slip against it, we can remove the point S, Fig. 56. After removing the point the tooth will now have the form as shown at tooth C, Fig. 57. We now take a thin and rather broad watch mainspring, bending a part straight and holding it in the line f f, and revolving the wheel in the direction as shown by the arrow, its action being indicated by figures 1 to 8; beginning at the point of the tooth at 1, at 2 it comes in contact with the whole of the lifting face, and from 3 to 8 only on the projecting corner which was left by the oil stone slip in removing the heel of the tooth. In this way all the teeth are acted upon until the corner is entirely removed. Of course oil stone dust and oil is first used upon the spring for grinding, after which the teeth are polished with diamantine. Care must be observed in using the spring so as not to get the end f too far into the tooth circle, as it would catch on the heel of the preceding tooth.

After the foregoing operation has been completed any feather edge remaining on the points of the teeth must be removed with a sapphire file and polished; we will now have a tooth as indicated by D, Fig. 57. This shape of tooth can hardly be said to be theoretically correct, nevertheless it is a close approximation of the proper form of tooth, which is shown by the dotted lines, and will then perform its functions much better than in its original condition.

Fig. 58 also shows how the spring must be moved from side to side—indicated by dotted lines—so that the lifting face will have a gentle curve instead of being flat; R represents the tooth.

After the wheel has been finished, as described, and again placed in the clock, it will be found that the balance makes only two-thirds of a turn, and as a result the movement can be easier brought to time and closely regulated.

In the above I have described the cause of excessive balance vibration, the method by which it can be corrected, and in what follows I shall endeavor to make clear the reasons for a diminished balance vibration or poor motion. It has been probably the experience of most watchmakers to repair small cylinders of a low grade, having a poor motion or no motion at all, and it would hardly be profitable to expend much time in repairing them. But considerable time is often wasted in improving the motion by polishing pivots and escape wheel teeth, possibly replacing the cap jewels, or even the hole jewels, increasing the escapement depth or making it shallower, examining the cylinder and finding nothing defective, and as a last effort putting in a stronger mainspring. But all in vain, the balance seems tired and with a slight pressure upon an arm of the center wheel it stops entirely.

In this case, as in a former one, in fact, it is necessary at all times to carefully examine the cylinder wheel. My reason for not considering the cylinder itself so much as the wheel is that the makers of them have made a considerable advance in their methods of manufacture, so we find the cylinders fairly well made and generally of the correct size. Even if the cylinder is incorrectly sized, either too large or small, it does not necessarily follow that the watch would have a bad motion, as I have frequently had old movements where the cylinder was incorrectly proportioned and yet the motion was often a good, satisfactory one. Generally speaking, the cylinder escapement is one which admits of the worst possible constructive proportions and treatment, as we have often examined such clocks when left for repairs, that, notwithstanding their being full of dirt, worn cylinder, broken jewel holes, etc., they have been running until one of the cylinder pivots has been completely worn away.

It only remains to look for the source of the trouble in the escape wheel. If we examine the wheel teeth carefully, we shall find them resembling those in Fig. 59, the dotted lines representing the correct shape of the teeth for a wheel of that diameter.

Why do we find wheels having such defective teeth? This is probably due to their rapid manufacture, as they very likely had the correct shape when first cut, but by careless grinding and polishing they were given improper forms, careless treatment being very evident at tooth F, which we find on examination has a feather edge at the point as well as at the heel of the tooth. If we grind these edges of the tooth with a ruby file, by placing it in the position as indicated by dotted lines h and h¹, and afterwards polishing the tooth point, we will find that the balance makes a better vibration. A wheel, having teeth like E, can still be used, but the balance will have a very poor motion, due to the fact that the impulse angle of the wheel tooth is too small; the impulse faces of the teeth having so small an angle, are nearly incapable of any action. With a tooth like G, if we should remove its bent point at the dotted line d, then the tooth would be too short, and as the inclination of the impulse face is incapable to produce a proper action, a new wheel must be used, having teeth as shown at Fig. 55.

The reasons why a tooth, having the shape as shown at F and G (Fig. 59), will cause a bad action of the escapement and also why in such cases with a greater force acting on the wheel, causes a stopping of the clock, I will endeavor to explain with the aid of the illustration Fig. 60. Here we clearly see the curved points of the teeth resting against the outer and inner walls of the cylinder while the escapement is in action.

Teeth H and H¹ represent the defective tooth, while K and K¹ shows a correctly formed tooth for a wheel of the same size, the correct depth and positions where the tooth strikes the inner and outer walls of the cylinder. It will be readily seen that the position of the tooth point upon the cylinder (at c) is most favorable in reducing the resistance to the least possible amount. But in the case of the teeth H and H¹ the condition is entirely different. We find that it was necessary to set the escapement very deeply in order that it could perform its functions at all, and, as a consequence, we have a false proportion; the effects being considerably increased by the worst possible position of the teeth H and H¹, where they touch the cylinder. While the cylinder c is turning in the direction shown by the arrows i i¹, the tooth does not affect the cylinder to any extent; but during the reverse movement of the cylinder, in the direction of o o¹, an excessive amount of engaging friction must take place. A close inspection of the drawing will enable us to see that there is a great tendency of the cylinder to drag the tooth along with it during each of these motions. It is evident that in such a case the friction will eventually become so great as to lock the escapement, and if greater pressure is applied by any means to teeth H and H¹, it is easily seen that this effect will take place much more rapidly. Replacing the escape wheel with one of correctly formed teeth and size is the best means at our disposal.