The conical pivot shown is the usual form in the finer grades of American watches; and as this form of pivot combines strength with a minimum tendency to attract the oil from the jewel hole, it is to be highly recommended. The back-taper T should never be neglected for reasons previously (47, 6) given. The proportions that should exist between the diameter of the pivot and the length of its bearing surface, as well as the shape of the end of the pivot, cannot be discussed here, as the scope of this work will not permit; but it should be borne in mind that the smaller the pivots, consistent with strength, the less the fluid friction will be. The sides of the pivots should be straight and parallel for a minute distance from their bearing surfaces; while the form of the rest of the pivot should be a gradually increasing curve, terminating at the point where the back-taper begins. The proper proportion of the diameter of the pivot to the diameter of the jewel hole varies according to conditions; but it has been previously (37) shown in a general way what this should be. ~55. The Escapements~ should be constructed in such a way that a maximum durability of oil may be secured. The acting surfaces of the teeth of the scape wheels should be made as small as possible consistent with durability (43, 8); while enough metal should be left _near_ the acting surfaces to be sufficient to retain the oil and prevent its attraction to the web of the wheel. The teeth of chronometer scape wheels should not be oiled, as it is liable to seriously alter the rate. When the oil becomes viscous by oxidation or by cold it would produce too much variation of fluid friction and so diminish the effort (25) of the mechanism. Some watchmakers oil the fork of the lever in anchor escapements _very_ slightly, by applying oil and then using pith to remove any surplus, while others never oil the fork. The writer has frequently observed ferric oxide or "rust" on the roller, fork, and on the plate or potance; but whether this was the result of not oiling or of oil having been applied which afterward become gummed, or evaporated, it would be interesting to know. ~56. The Curb Pins~ sometimes produce the ferric oxide mentioned by their action on the hairspring. This has been remedied by the same method as used in the fork just referred to, and if a _very minute_ quantity of oil can be applied--such a minute quantity that if the whole spring were equally covered by a coating of oil equally _thin_, such film being _so_ thin that it would have _no_ tendency to cause the coils to adhere, or to cause small particles of matter to adhere--then it may be that this method deserves notice.