The present type of headstock with the running mandrel, in unison with which the work revolves, is comparatively modern. All the old lathes were dead-centre lathes. All the native lathes used today in India, Persia, China, Egypt, North Africa, and in some other countries, are of the same type. The work is pivoted between two fixed point centres, and is rotated in alternate directions by means of a cord wrapped round the work, so that actual cutting only takes place during half the time. The turner mostly sits on the ground, and is equally adept in guiding the tools with his toes as with his hands; and he is generally an itinerant, setting down the stakes between which the point centres are carried adjacent to his customer's residence. Except when an attendant lad moves the cord, the turner has but one hand and one foot free for the manipulation of the tools. In Europe, the standing posture generally adopted by workmen caused the lathe to be raised up, and a spring pole overhead, which was operated from a treadle, gave the reciprocating movement to the cord and to the work. Such a lathe was illustrated in works published in 1568 and 1569. It occurs in Moxon, 1703, while the archery bow used for the same purpose occurs in Plumier, 1701 and 1749, and in Bergeron, 1792.
The reciprocating lathe was not superseded rapidly. Even after the introduction of the driving-wheel, many continued to prefer the old pole, or spring bow. Schopper, 1568, seems to have been the first to illustrate the wheel-driven lathe. De Caus, 1624, Moxon, 1677, Plumier, 1701, and others describe it; Plumier and others show the wheel and bow in combination on the same lathe. The driving-wheel was sometimes placed overhead, sometimes underneath, sometimes in a separate framing. The very old treadle-driven grindstone, revolved by the workman's foot, would suggest a similar application to the lathe. The early wheels were of wood, very clumsily made, insufficient in weight, and they ran with much friction in bearings. Even well into the present century many lathes were still provided with both the bow and the wheel, the latter being placed overhead. The wheel was thus used in the dead-centre lathe, driving directly on to the work, or, when that was of small diameter, on to a pulley encircling the work.
The dead-centre lathe was also employed for chucked work, one of the centres being unused, and the other being fitted with a chuck which was revolved on a spindle by means of a cord or band. These, however, have no representatives in modern engineers' lathes. The old millwrights made use of dead-centre lathes for metal-turning, but the driving was done through a wheel and pinion for increase of power, and the germ of back-gear is probably to be found in these. But the dead-centre lathe still survives in the watchmaker's turns and in the double-axle turning lathes; in the first because the running mandrel would not produce sufficiently accurate results, in the second because it is a very convenient arrangement. The occupation of a portion of the work being revolved between dead-centres by the cord of the pole, bow, or wheel, was a distinct drawback, which was removed by placing the driving-cord on a separate mandrel, which in turn drove the work. A mandrel lathe occurs as early as 1568 in Schopper. It is probable that the first idea of the mandrel arose from the practice of moving one of the dead-centres out of the way when it was necessary to do boring and hollow work generally. In such a case a collar was substituted to support one end of the work while being bored or cupped. From this device to the use of a permanent mandrel, having the front end available for the attachment of articles to be bored or turned, was a natural transition. Iron mandrels followed those of wood, and a screwed nose replaced the cement or nails used on the wooden mandrels. Yet, still many of the early iron mandrels were driven by the pole or the archery bow, with a reciprocating movement.
These early mandrels went through many transitions, being used also for cutting screws with a traverse motion, and being mounted in various ways that appear very odd at the present time, but the study of which lies outside our immediate subject. In them, however, we can trace the beginnings of much present practice. Although they ran at first in wood bearings, after some time bearings of lead or tin were cast round the necks into vee'd recesses in the wood of the headstock; for it must be remembered that all the early headstocks were of wood. These bearings were divided, being cast in two operations. The American white-metal bearings have therefore a counterpart in the more crude practice of a century or more ago. At that time this practice afforded the only way in which the alignment of the running centre and of the fixed centre could be insured, since the machines and methods available to-day were absolutely unknown at that period.
Gradually the use of metal began to take the place of that of wood; but 100 years since there were few headstocks made wholly of iron, excepting those for heavy engineers' work. The heads and pulleys were of wood, and the mandrels and their bearings only of metal. For the front bearing, a bored metal plate was often screwed to the front of the wooden headstock, and the back-centre screw was pinched when set against the tail of the mandrel, by dividing the hinder upright of the headstock, and pinching it round the tail screw by means of a bolt.
In some wood-turners' shops to-day one sees the back poppet-head made wholly of timber, and clamped in place on the bed by means of a stout wooden wedge. The wedge passes through a tongue formed as a prolongation of the poppet, passing down between the bearers of the wooden bed, and it bears against the under faces of the shears, or bed bars. Most of the early lathe-heads, both fast and loose, were held in some such fashion, and when the running mandrel was employed, the fast head was still made of two separate blocks of wood wedged down to the bed at the proper distance apart.
Naturally the truth of such lathes was very imperfect, because of the effect of temperature on the wood of which the beds and headstocks were composed. The first important departure in this respect occurred when Holtzapffel, in 1704, began to cast the heads in brass, and afterwards in iron, and substituted mandrels with conical necks for those with parallel necks. In a work published in Paris in 1671, there occur conical bearings for the front of the lathe mandrel; tool past, and pinching screw. Screws were used by Holtzapffel for clamping, displacing the use of wedges. Some very heavy lathes were constructed at about the same time, both by Maudslay, and at Soho Works, and the slide-rest in its essential elements was invented at this period. The influence of these events on the growth of the modern lathe is traceable in many ways.
It appears, therefore, that in all the early lathes, the methods of fitting of the heads were based upon the idea of two elements set and fixed rigidly between and upon the bed shears. The methods were crude, but the intention evident. There was no provision thought of, of means for adjustment of centres, neither was it necessary for work running between dead-centres, and turned wholly by hand. With the introduction of the slide-rest, and the cutting also of long screws in the lathe, the alignment of the heads, and the resulting possibility of need for minute adjustment to insure or preserve perfect alignment became absolutely necessary. In the provisions for effecting this lie much of the differences in the construction of the heads of English and American lathes—differences which arise from the radically opposite views taken in each country in regard to the methods of arriving at the same results. With the heavy demands made on self-acting lathes, and the invention of new types of machine tools unknown a hundred years ago, stronger and stiffer designs were evolved, and parts proportioned to the work they had to do.
- English & American Lathes by Joseph Horner, Feb 1900 pages 2-4