The use of special machines was applied to the manufacture of the smaller portions of the watch much earlier than to the production of the plates, as above described; indeed, it may be said that some of the earliest machines employed were semi-automatic. But so radical have been the changes and improvements insti tuted within the last 15, and especially during the last eight years, that methods previously used can hardly be considered as modern watchmaking.
One of the most interesting of those which are strictly modern is one of the turning of the delicate arbor on which is mounted the most swiftly moving wheel of the entire watch, the 'balance.* The fact that this wheel moves with such rapidity, together with the fact of its location at the farthest point from the reservoir of power, in the mainspring, makes it abso lutely essential that it should be subject to the least possible amount of running friction. Consequently the pivots of this 'balance staff* are exceedingly minute. It is, of course, an ab solute necessity that every portion of the staff, including the delicate pivots, should run abso lutely true. Exact truth was a difficult attain ment under the methods ordinarily used, al though great care was used to avoid any stress of distortion of the staff, by a very gradual reduction of the blank, so that 10 to 14 separate turnings were required to complete a staff. By the most approved modern methods and ma chines, all of the required turnings, including the pivots, are performed before the completed staff is severed from the rod of wire from which it is made. So complete is the machine, and so accurate and rapid is its operation, that a staff is delivered from the machine each 90 seconds.
The pinions of various sizes, which com pose so large a portion of the time train of the watch, are formed from the best quality of steel rods and, as this wire must of course have a diameter as great as the largest pan of the pinion, it follows that a large proportion of the material must be removed to bring the pinion to its finish dimensions. An early method consisted in first cutting the wire into blanks of sufficient length, then to insert those blanks in a suitable chuck in the running spindle of a lathe and with a suitable tool to remove the metal from the projecting end so as to bring it to an accurate point or centre. The blank was then reversed in the chuck and the opposite end pointed in like manner. In some instances a portion of the excess metal was removed in connection with the operation pointing.
Following this somewhat crude method of pointing and initial turning, came the inven tion and adoption of so-called 'roughing out machines,' in which a rod of wire is inserted in the machine which automatically cuts off, points and turns the blanks on both ends and delivers them uniform in dimensions and re quiring only the slight finish turnings and the cutting of the teeth when they will be ready for hardening and tempering.
A later form of machine has been devised, which takes a rod of wire at one end and de livers the pieces girlish turned* at the other end.
The finished blanks are then placed in a magazine or 'feeder' in a cutting machine which automatically takes them one at a time and forms the teeth of the required number and shape, the work of the attendant being to keep the magazine supplied with blanks, and to examine the pinions occasionally in order to detect any imperfections; so that a single per son can easily attend six cutting machines. For the greater convenience and comfort of the operator a chair is fitted to run on rollers on an iron track on the floor. These rolling chairs are also provided in connection with numerous other machines where a single person attends to a number of machines.
Within two or three years D. H. Church of the Waltham factory has created machines in which pinions are completely turned and the required teeth are cut. All the work being automatically performed, so that completed pinions are made, the last operation being to sever the finished pinion from the rod of wire from which it was formed. The exactness with which these machines operate makes pos sible the production of pinions which could not have been produced with tools and ma chines in previous use.
As each watch movement required from 30 to 50 screws it will be evident that for a daily product of more than 2,500 watches more than 100,000 screws would be required. By the earlier methods employed in American watch factories an active man could make from 800 to 1,200 screws per day so that more than 80 men would be required for the production of this large number. But within a few years screw-making machines have been devised which are capable of making from 4,000 to 10,000 screws each per day, and being entirely automatic in action, a single attendant is able to attend six or more machines.