very few mills without them; that printers and book- 

 binders had generally adopted them for pamphlet 

 trimming. 157 



It was growing towards night, and as I had an 

 engagement for the evening, and moreover, began 

 to feel hungry, I again essayed to go, when Mr. 

 Donkin asked if I had ever given a thought to the 

 possibility of making an absolutely perfect screw. 

 Mr. Saxton had told me this had been a hobby with 

 Mr. Donkin for many years, and on the previous 

 Sunday I had gone with Saxton to the Royal Observ- 

 atory at Greenwich, where we had seen the great 

 mural wheel that had been turned at Donkin's 

 works, and graduated by Troughton with one of 

 Donkin's screws. 



I replied that when a boy I had been much inter- 

 ested by my grandfather's ideas and efforts in that 

 direction; that to generate a screw of a certain pitch, 

 instead of setting a sharp knife edge at an angle of 

 the axis of the spindle to be cut, as I had been shown 

 the day previous as Maudslay's design, my grand- 

 father had drawn hard iron wire by light drafts to 

 insure accuracy; that this he had wound like a spiral 

 spring on an accurately-turned iron mandrel, with 

 a fixed collar at one end and a loose one at the other, 

 driven up by a screw nut, to force and hold the coils 

 when wound in close contact. On this he had cast 

 a fusible metal nut. I had found this in grand- 

 father's shop with a number of wooden rods wound 

 in the same manner with wires of various sizes; these 

 latter he had explained as having been used in 

 testing and correcting his wire gauge by the numbers 

 of coils to the inch. The iron one, he said, was an 

 effort to produce an accurate screw for his friend, 

 David Rittenhouse. He did not say for what use 

 but I infer it had something to do with Rittenhouse's 

 astronomical work. I hastily described old John 

 White's expedient for cutting press screws before 

 the time of engine lathes that I described in my 

 second paper. 158 



Mr. Donkin had evidently got on his hobby; he in- 

 sisted on my stopping to see a couple of his screws for 

 graduating. They were of steel, about i% inches 

 diameter and about i % inches long, as near as I can 



157 The paper-cutting machine was patented by Joseph 

 Woodhouse, of New York, but not until May 30, 1835. There 

 is no restored drawing. See John W. Maxson, Jr., "Coleman 

 Sellers, Machine Maker to America's First Mechanized Paper 

 Mills," The Paper Maker (1961), vol. 30, no. I, pp. 13-27. 



169 Chapter 5, above. 



recollect; one of 25, the other of 50 threads to the inch. 

 This was firmly placed over the 25-thread one, so that 

 50 threads calipered them. In front a movable mi- 

 croscope, with a finely-graduated micrometer eye- 

 piece; by sliding the microscope and turning the 

 screws, the least possible variation could be detected. 

 It was growing too dark to see this to advantage. 

 Mr. Donkin spoke enthusiastically as to what Mr. 

 Maudslay had done towards establishing standard 

 screws; 159 but as to absolutely correct ones, he said: 

 "We may have a screw with a deep or long nut on it, 

 that works smooth and easy, and with the most deli- 

 cate handling no imperfection can be discovered, yet, 

 on using the screw for a dividing engine, errors would 

 soon be apparent." 



He asked if I had ever thought of or seen any device 

 for testing irregularities or gain or loss of one thread 

 over another? 



None but a very simple one of Isaiah Lukens, town 

 clockmaker of Philadelphia — a little sliding carriage 

 parallel with the screw to be tested. This carriage 

 had a fixed point in the groove or channel of the 

 screw. By turning the screw it would be made to 

 advance in either direction as the screw was turned. 

 In this carriage in the plane of the axis of the screw 

 was an adjustable stud or center that could be set at 

 any number of threads desired from the carrying 

 point. On this stud was a lever or index hand, as 

 Lukens called it, the short end fitting into a channel 

 of the screw, the long end resting on a graduated arc, 

 adjusting the fulcrum stud so that the index hand 

 pointed to o. On turning the screw to advance the 

 carriage by the fixed point, the least variation would 

 be shown by the point of the index lever. There was 

 also an arrangement to measure short inequalities by 

 using the obliquity of the thread by either raising or 

 depressing the lever on the center pin or fulcrum. 

 I did not known what mode had been taken to cor- 

 rect errors when found, but I thought Mr. Saxton, 

 who had served his apprenticeship with Lukens, would 

 be able to give all information. 



He had never spoken to Saxton on the subject, but 

 he should certainly take the first opportunity of doing 

 so. He did not then say what he had done, but only 

 remarked that it was wonderful that minds over 3,000 

 miles apart should in any degree travel the same roads. 

 He threw the covers off of some of his apparatus, no 



159 That Maudslay and Donkin were friends is suggested by 

 their joint British patent (no. 2948, July 24, 1806) for an epi- 

 cyclic gear train, applied to a hoisting winch. 



121 



