octant by Thomas Godfrey and John Hadley in the early 1730s, and the .... Randall C. Brooks, The Precision Screw and Astronomical, Nautical and Surveying.
The Brightly Circular Dividing Engine Randall C. Brooks In 1992 the National Museum of Science and Technology in Ottawa acquired a circular dividing engine that had been made by Charles Henry Brightly in Philadelphia in 1890. Following Brightly's death in 1897, the engine was sold to Ulmer & Hoff, an instrument firm in Cleveland. The Thomas Pocklington Instrument Company, which purchased the engine in 1940 and moved it to Toronto, used it extensively during World War II and then intermittently through the mid-1960s. In 1992, when the engine came to the Museum, I had a chance to interview Thomas Pocklington's son William. This interview offered a fascinating glimpse into a world which is seldom recorded in print, and one which, because of recent advances in machining technology, has virtually disappeared.1
Overall view of the Brightly dividing engine. Main components are the dividing wheel with its axis below table top (dark shaft); division cutting head (center); driving wheel and ratchet indexing mechanism (right). The driving pulley (edge on right) was driven by a belt from above. The engine stands about 1 m high, and the dividing wheel is approximately 1.3 m in diameter. The engine is made of brass, bronze, cast iron and steel, with an assembled weight of some 400 kg.
75
The introduction of circular dividing engines in the late eighteenth century revolutionized the manufacture of navigation and surveying instruments. However, because they were difficult and expensive to build, these sophisticated machine tools remained a rarity—only about 50 are known to have been made prior to 1900~and they were invariably prized and pampered.2 The best dividing engines could draw scale division lines to within 1/40,000* inch, if they were operated by skilled and careful craftsmen who took account of even such factors as body heat. They remained the most precise machines in use until the development, in the 1880s, of the engines that ruled the diffraction gratings needed for spectroscopic work.3 The need for dividing engines was stimulated by the invention of the octant by Thomas Godfrey and John Hadley in the early 1730s, and the sextant at mid-century. The first dividing engine was completed around 1738 by ThemaVHindley, a clock and instrument maker in York, England. In France, the due de Chaulnes designed and made dividing engines in the 1760s, also with some success. One of these engines, believed to have been completed in 1765, is preserved in the Museo e Instito de Scienza, Florence. Jesse Ramsden's first circular dividing engine, which he completed in 1775 and which won a substantial prize from the Bureau of Longitude, was mechanically superior to those made by Chaulnes, but inferior in concept. It is now in the collections of the National Museum of American History.4 Edward Troughton, Ramsden's successor as the leading instrument maker in England, adopted the concept used by Chaulnes. His circular dividing engine of 1805 remained in operation for several decades with subsequent improvements including a degree of automation. It is now in the Science Museum, London. Charles Henry Brightly Brightly was born in 1817 in Bungay, in Suffolk County, England and, with his parents and other family members, immigrated to the U.S. in 1831. If he served an apprenticeship in England, as seems likely, the details are unknown.5 Brightly was in Philadelphia in the years 1844-48, working as a machinist, and again in 1867, working as an instrument maker. There are, however, no clues to his whereabouts during the intervening 19 years. In 1870 Brightly went into partnership with Charles H. Heller (1839-1912), an instrument maker who had spent the last five years as a partner in Wm. J. Young & Co. Heller & Brightly were immediately successful, selling engineering instruments around the world and capturing coveted prizes.6 Although we cannot clearly identify the various contributions of the two partners or their several employees, it appears that Brightly worked on their large circular dividing engine, which had been built by M. 76
Kaiser in Stuttgart, making it automatic. Brightly may also have made the firm's linear dividing engine and their smaller circular dividing engine.7 Brightly withdrew from the firm on July 13, 1889, went into business on his own, and issued advertisements that read: CHARLES H. BRIGHTLY, Late of Heller & Brightly, Manufacturer of Mathematical, Engineering, and Surveying Instruments. Germantown Junction, P.R.R., 16th St. Station, P.&R.R.R. Philadelphia, Pa."8 The earliest account of his dividing engine comes from the 1904 catalog of J.C. Ulmer Co. (successors to Ulmer & Hoff): We have our own automatic graduating machine built by Mr. Brightly, late of Heller & Brightly, in the year 1890; this engine he built for his own use and he spared neither time nor expense to make it as nearly perfect as possible; it embodies all the improvements that his many years of practical experience and constant use of this delicate class of instrument could suggest to him.9
Closeup of the worm gear and a section of the scale on the wheel. The scale above the 0" mark was laid down during fabrication to check the progress of forming the thread and the primary scale scribed on the silver inset. Two other test scales can be seen on the silver insert below the primary scale. Each tooth of the gear on the wheel equals %" with 1440 teeth around the 360" of the wheel.
77
The most critical elements of a dividing engine, and the most difficult to fabricate, are the worm gear which drives the dividing wheel, and the wheel itself, which which has carefully cut teeth on its edge. Elimination of periodic and non-periodic errors is the primary challenge. The circular plate of the instrument to be divided is mounted concentrically on the engine's wheel. When the wheel has rotated to the correct position, a cutting knife drops onto the plate making a scoring mark for the scale. Motion was originally provided by a foot treadle, then by weights and pulleys; by the end of the nineteenth century, electric motors provided power to complete the thousands of motions required to complete the scale of the instrument.
Cutting head as seen from the front. Cutting knife is on left near the bottom. Note the cams on the shaft left of center that controlled the rasing and lowering of the knife.
In our interview, William Pocklington stated that, when optically tested, the Brightly engine was found to have an error of 4-5 seconds of arc (approximately l/260,000Ul of a degree). This was somewhat more than the 2-3 seconds that Ulmer had claimed when they sold the engine, but claims for dividing engines were often exaggerated, especially before optical testing methods became the norm. To ensure optimum performance, the engine was kept in a temperature controlled area that was separate from the main shop, and protected at night with a plywood case. 78
Pocklington further explained that one revolution of the dividing engine took approximately one hour and gave a resolution of 30 seconds of arc. That's 4,320 strokes! If the graduations were on silver, one revolution sufficed. If, however, the graduations were on German silver, which is "hard as the dickens," it was necessary to go around the circle two or three times to get the grooves deep enough, "unless you want to put on an awful lot of weight on the scribe." Pocklington also mentioned the problem of getting a dead clean edge, or preventing the "bell mouth" that could occur when the scribe fell off the edge of the circle. The biggest challenge in running the engine was setting up the work, making sure the work piece was concentric with center of the engine, and that the drive plate was true. According to Pocklington, "these are just little things that don't show up in the book—too many things are presumed and when you get down to the actual work, you find out you should have done that or I should have checked this." Brightly had equipped the dividing engine with a special centering device with which, according to Ulmer, "a variation of 0.00002 of an inch is readily determined." Pocklington, however, mentioned that they had found the centering device to be "terribly difficult to use," and so had replaced it with a dial micrometer setup reading to I/IO.OOO*1 inch.
Ratchet that controlled the rotation of the main dividing wheel. The drive pulley turned until the ratchet moved one notch stopping the wheel while the cutting knife dropped down to score a division mark, then it resumed to the next notch of the ratchet.
79
1. Ron Pocklington lent the engine to the National Museum of Science and Technology in 1992, and his sister, Mary Clough, donated it to the Museum in 1995. 2. This tally, of course, neglects the many smaller linear and circular dividing engines made by such firms as Perreaux in Paris and SIP in Geneva, and which were designed primarily for instruction in student laboratories. 3. Randall C. Brooks, The Precision Screw and Astronomical, Nautical and Surveying Instruments of the 17" -19" c (PhD thesis, Univ. of Leicester, 1989), see chap. 4. For information on British dividing engines, see also John Brooks, "Dividing Engines," Annals of Science, 44, 110, 1991 and Alan Chapman, Dividing the Circle (New York and Toronto, 1990). See also J. A. Bennett, The Divided Circle (Oxford, 1987), p. 130133, and Jean Randier, Marine Navigation Instruments (London, 1977), p. 93ff. 4. Ramsden's circular dividing engine was purchased from Ramsden's heirs by Knox & Shain, mathematical instruments makers in Philadelphia, presumably for use in their shop. Henry Morton, president of the Stevens Institute of Technology, recognized the important role that the engine had played in the history of technology, bought the engine from Knox & Shain sometime around 1880, and deposited it at the Smithsonian in 1890; Morton's heirs converted this deposit to a donation in 1902. This story can be traced in the records of Smithsonian accession #40282. 5. Genealogical details provided by Brightly's descendants, in files at National Museum of American History. Brightly is not listed in Gloria Clifton, Directory of British Scientific Instrument Makers: 1550-1851 ( London, 1995). 6. Charles E. Smart, The Makers of Surveying Instruments in America Since 1700 (Troy, NY , 1962), pp. 82-83. R.C. Miller, "The Heller & Brightly Records," Rittenhouse 4 (1990): 43-55. 7. The larger engine is pictured in R.C. Miller, "George Kegelman and Kegelman Brothers, Mathematical and Optical Instrument Makers," Rittenhouse 5 (1991): 56-58. 8. Advertisements in Engineering News (1891-1895). The date of Brightly's withdrawal from Heller & Brightly is from Illustrated Philadelphia (Philadelphia, 1889), p. 132. 9. Quoted in Smart, op. cit., p. 161.
80
