The Contributions of Charles S. Peirce to Metrology Author(s): Victor F. Lenzen Reviewed work(s): Source: Proceedings of the American Philosophical Society, Vol. 109, No. 1 (Feb. 18, 1965), pp. 29-46 Published by: American Philosophical Society Stable URL: http://www.jstor.org/stable/985776 . Accessed: 25/09/2012 05:25

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

American Philosophical Society is collaborating with JSTOR to digitize, preserve and extend access to Proceedings of the American Philosophical Society.

http://www.jstor.org THE CONTRIBUTIONS OF CHARLES S. PEIRCE TO METROLOGY

VICTOR F. LENZEN * ProfessorEmeritus of Physics, Universityof California, Berkeley

I terminationsof gravitywith a reversiblependulum. (2) METROLOGY is the scienceof weightsand meas- The comparisonof the Low Moor Iron ures: of units of measurement,of standardsthat Yard No. 57 withthe BritishImperial Yards Nos. embodyfundamental units, and of the reduction 1 and 58, in 1883,at the StandardsOffice, London. of quantitativemeasures. The United States (3) Comparisonof theyard and meterfrom the Officeof Weights and Measures was a depart- lengthof the second's pendulumat Kew Observa- mentof the Coast and Geodetic Survey, initially tory,as determinedby CaptainHeaviside in inches authorizedby Congress in 1807 for the Survey and by Peirce in meters. Observationson the of the Coast of the United States, untilthe Office periods of oscillation,at high and low tempera- provided the nucleus of the National Bureau of tures, of Peirce Yard and Meter Invariable, Standards in 1901. Significantcontributions to ReversiblePendulums, for the comparisonof the the progressof Metrologywere made by Charles Yard and Meter. S. Peirce, initiallyan aid, then Assistant from (4) A pioneer determinationof the meter in July1, 1867,to December31, 1891,in the Survey. termsof a wavelengthof light,1877-1879. Charles Sanders Peirce (1839-1914) is best (5) Membershipin the AmericanMetrological knownas a philosopherwho foundedPragmatism, Society,service on its committees,and initiation and as a logicianwho contributedto the develop- of a resolutionat a meeting,December 30, 1884, ment of mathematicallogic. His Collected Pa- that a committeeof the society be appointedto pers, principallyon philosophyand logic, has address the Secretary of the Treasury, and to been publishedby the Harvard UniversityPress memorializeCongress, upon the need to establish (1931-1958). But he also contributedto linear an efficientbureau of weights, measures, and algebra, to the mathematicaltheory of map pro- physicalunits. jections,and to several fieldsof physicalscience. (6) Direction of the Officeof Weights and He observedsolar eclipses with the spectroscope Measures, fromOctober 1, 1884, to February22, or polariscope,discovered and measured spectral 1885. lines of the aurora borealis, investigatedphoto- (7) Testimonyto the Allison Commissionof metricallythe light from the stars and made a Congress, January 24, 1885, on behalf of a pioneer determinationof surfacesof equal star- strengthenedoffice of weightsand measures,one density,determined the intensityof gravitywith empoweredto grantcertificates for standards that a pendulumand developedthe theoryof this in- would be valid in courts. strument,calculated the ellipticityof the earth (8) Membershipon the Committeeon Weights, from variations of gravity,and contributedto Measures and Coinage of the National Academy spectroscopywhile employingit in the serviceof of Sciences. metrology. Peirce's principal contributionsto (9) Appointmentby the President of the Metrologywere: United States, in 1888, to the Assay Commission for United States Coinage. (1) The acquisitionfor the Coast Survey of a (10) Authorshipof definitionsof terms for line-meterfrom the German Imperial Standards weightsand measuresfor the CenturyDictionary. Office,to serve as a standardof lengthfor his de- Authorshipof revision of section on units in Mach's Science of Mechanics,a * The author gratefully acknowledges the inspiration translationfrom and referencesreceived fromProfessors Max H. Fisch and the German. Carolyn Eisele; also the courtesies extended by Dr. Lewis II V. Judson, Mr. Walter W. Weinstein, and Mr. L. Chis- holm of the National Bureau of Standards, Washington, The significanceof Peirce's work in Metrology D. C. can be understoodbest against a backgroundof

PROCEEDINGS OF THE AMERICAN PHILOSOPHICAL SOCIETY, VOL. 109, NO. 1, FEBRUARY, 1965 29 30 VICTOR F. LENZEN [PROC. AMER. PHIL. SOC. the historyof standards of weights and meas- Tralles was the deputyof the Helvetic Republic ures in Europe and Americaduring the nineteenth on this InternationalCommission which in 1799 century. completedthe creationof the Metric System. A Colonial America under British rule used platinumbar, constructedso that the distancebe- weightsand measuresprincipally of English ori- tweenthe end faces was one ten-millionthof the gin. The standards employed in the colonies distance from the North Pole to the Equator, were more or less authenticcopies of English became the metric standard of length and was standardsand had been broughtover fromEng- called theMetre des Archives. A similarplatinum land at varioustimes. bar became the Metre du Conservatoire. When On February10, 1807, PresidentThomas Jef- Hassler came to the United States in 1805, he fersonapproved an act of the Ninth Congress, brought with him a copy of the Metre des Second Session, that authorized and requested Archives,a bar of iron thatbecame known in the the Presidentto cause a surveyto be takenof the United Statesas the "CommitteeMeter." Hassler coasts of theUnited States.' AlbertGallatin, Sec- describedthis standardin a memoirpublished in retaryof the Treasury,on March 25, 1807, ad- the Transactionsof the American Philosophical dressed a circularletter to scientificmen, asking Society: for plans to carrythe surveyinto effect. Ferdi- An iron metrestandarded at Paris, in 1799,by the nand Rudolph Hassler repliedon April 3, 1807. Committeeof Weightsand Measures,composed of A commission,sitting in Philadelphia, recom- membersof the NationalInstitute and of deputies mendedHassler's plan, and by directionof Jeffer- fromother countries.5 Its breadthis 1.13 inches,its son he was so notifiedunder date of July 21, thickness0.36 inches,English measure. My friend, of by Robert Patterson,Director of the Mint Mr. J. G. Tralles,now memberof the Academy 1807, Sciencesof Berlin,was at thattime the deputy of the in Philadelphia.2 Execution of the survey was HelveticRepublic for this purpose;and as may be suspendedon accountof politicaldisturbances in seen in the accountof the operationsof this com- Europe and America,and actuallybegan in 1816. mittee,he was theforeign member directing the con- who had structionand comparisonof the measuresof length. Hassler was a native of Switzerland He had onemetre constructed for himself and one for come to the United States in 1805.3 In his native me, at the same timewith all the others,and sub- land, he had been a student of Johann Georg jectedin all respectsto the sameprocesses and com- Tralles, a Germanfrom Hamburg, who had been parisons. elected Professor of Mathematics,Physics and Hassler also brought to America a standard Chemistryat the Academyin Bern. Tralles and kilogram,also a gift from Tralles, and three Hassler in 1791 measureda base-linefor geodetic toises. The toise,slightly less thantwo metersin purposesin the neighborhoodof Aarsburg,a town length,had been a standardin France and, indeed, near Bern. Subsequently,Tralles participatedin furnisheda rival to the meter until the last the establishmentof the metricsystem in France. quarter of the nineteenthcentury.6 Hassler in L'Assemble'e Constituanteof France, In 1790 1806 or 1807 sold the standardmeter, kilogram, of acted to createa upon the initiative Talleyrand, and the three toises to John Vaughan, who on of weightsand measures.4 new and uniformsystem February 20, 1807, read before the American of various com- The deliberationsand operations Philosophical Society Hassler's description of led to the decision to mitteesand commissions these standards,and it was recorded "that he and in September,1798, an createa metricsystem, considersthe above standardsas in the offerof was assembledin Paris, internationalcommission the Societyat any futureday at the priceat which of the work done and to fix to take cognizance he paid forthem." 7 The instrumentswere loaned standards of the new system. authenticallythe to Hassler forcomparisons and applicationin his 1'Annals of the Congress of the United States, Ninth workas Superintendentof Weightsand Measures Congress,Second Session (Washington,1852), p. 1255. 2 F. R. Hassler,"Papers on Various Subjectsconnected 5Supra, note 2 (n. 2), p. 253. with the Survey of the Coast of the United States." 6 C. S. Peirce definedthe word toise in the Century of length Communicated 3d March, 1820. Trans. Amer. Philos. Dictionaryas follows: "toise. An old measure Soc., n.s., 2, 12 (1825): pp. 232420. in France, containing6 French feet, or 1.949 meters, 3 Florian Cajori, The Chequered Career of Ferdinand equivalentto 6.395 Englishfeet." Rudolph Hassler (Boston, 1929). 7 "Early Proceedingsof the AmericanPhilosophical 4G. Bigourdan, La Systeme Metrique des Poids et Society, 1744 to 1838," Proc. Amer. Philos. Soc. 22; Mesures (Paris, 1901), p. 4. (1884): p. 393. VOL. 109, NO. 1, 19651 CONTRIBUTIONS OF PEIRCE TO METROLOGY 31

and of the Coast Survey. Indeed, the Committee pointed Superintendentof the Coast Survey by Meter served as the standard of length for the PresidentJackson. Coast and Geodetic Survey until 1890. The interestand competenceof Hassler in the In preparationfor the surveyof the coast, Has- subject of weightsand measures,and the creation sler was sent to Europe in 1811 to secure the of a joint superintendencyof the Coast Surveyand necessary instruments.8A principal instrument of Weights and Measures, impresseda concern makerin London, Edward Troughton,undertook for standardsupon the Coast Survey throughout to make many of the requiredinstruments. We its history. In 1817 Hassler comparedthe Com- shall take notice only of the Troughtonscale, a mitteeMeter, which represented the Frenchstand- scale of 82 inches subdividedto tenthsof inches ard, withthe Troughtonscale, whichrepresented on a silver inlay in a brass bar about 86 inches the Englishstandard, at varioustemperatures. In long. In 1815 Hassler returnedto the United 1831 he again compared these standards. His States withthe Troughtonscale and otherinstru- final result,10the average of results obtained in ments for the Survey. On August 3, 1815, he 1817 and 1831, was stated in the formthat the was formallynotified by letterfrom Secretary of CommitteeMeter was 39.38091714 inches of the the TreasuryDallas of his appointmentby Presi- Troughtonscale, with meterand scale at 320 F. dent James Madison as Superintendentof the An Act of Congress,approved June 14, 1836, Survey of the Coast. Hassler thenbegan opera- directedthe Secretaryof the Treasury to cause tions to establishbaselines, but in 1818 he was a completeset of all the weights and measures relieved as Superintendentafter an act of Con- adopted as standards, for the use of custom- gress,approved on April 18 of thatyear, repealed houses, to be deliveredto the Governorof each so much of the Act of 1807 as authorizedthe State, "to the end that a uniformstandard of employmentof otherpersons in the executionof weightsand measuresmay be establishedthrough- said act, than the persons belongingto the army out the United States." 11 L. A. Fischer statesin and navy. his historicalaccount of United States standards, A resolutionof the United States Senate, May It is evidentfrom the reports of Mr. Hasslerthat he 29, 1830, directedthe Secretaryof the Treasury regardedthe EnglishYard as the real standardof to cause a comparisonto be made of the standards lengthof the UnitedStates and the Troughtonscale of weightand measureused in the severalcustom- merelyas a copy whose lengthshould be corrected houses in the United States. On November 2, if it was subsequentlyfound to differfrom the Eng- lishYard, and thisview was takenby thosewho sub- 1830, PresidentAndrew Jacksonappointed Has- sequentlyhad chargeof our standards.'2 sler Superintendentof Weights and Measures. The Secretaryof the Treasury,S. D. Ingham,on In fact, the brass scale of 82 inches had been March 3, 1831, reportedto the Presidentof the copied fromTroughton's own scale withoutveri- Senate thatthe comparison required by the resolu- ficationby comparisonwith the English standard. tion of May 29, 1830, commencedtinder the im- The Troughtonscale was held to be identical mediatesuperintendence of Hassler, and the Sec- with Bird's Standard Yard of 1760, which had retarymentioned the Troughtonscale as one of been testedand deemedidentical with the British the authenticunits adopted for the comparison. ParliamentaryStandard of 1758, by Sir George Hassler's reportswere dated January27, 1832, Shuckburghin 1798; and by Captain Kater in and June 29, 1832, and are recordedin House of 1821,on theoccasion of a determinationof weights Representatives,22nd Congress,1st Session,Doc- and measuresin England whichhad been ordered ument No. 299, "Comparison of Weights and by Parliamentin 1814. The new standardswere Measures of Lengthand Capacityreported to the executed in 1825 and were called the Imperial Senate of the United States by the TreasuryDe- Standards. On October 16, 1834, the Houses of partmentin 1832 and made by Ferd. Rod. Has- Parliamentburned and the Britishstandards were sler." 9 On August 9, 1832, Hassler was reap- destroyed. New imperial standards were made

8 Supra, n. 2. 10 Ibid., p. 72. 9 In ExecutiveDocuments 6. Printedby Orderof the 11 24th Congress,1st Session, Registerof Debates in House of Representativesat the First Session of the Congress12, 4 (Washington,1836), Appendix,p. xix. Twenty-SecondCongress. Begun and Held at the City 12 L. A. Fischer, "History of the Standard Weights of WashingtonDecember 7, 1831 (Washington,1832), -and Measures of the United States," Bulletin6of the DocumentNo. 299,pp. iv + 122,4 plates. Also supra n. 3. Bureau of Standards,1 3, (Washington,1905): p. 370. 32 VICTOR F. LENZEN [PROC. AMER. PHIL. SOC.

to replace the destroyedones, and in 1856 two Peirce was sent to Europe in April, 1875, by the copies of the ImperialYard were presentedto the United States Coast Survey,to acquire the Rep- United States throughG. B. Airy, Astronomer sold pendulumapparatus, to learn its application, Royal. In a letterof transmittal,dated December and to determinethe intensityof gravityat initial 21, 1855, Airy describedthe presentedstandards stationsin Europe.16 From the GermanImperial as Bronze Yard No. 11, specifiedas standardat Normaleichungsamt(Standards Office), Berlin, 61.790 F; and Low Moor Iron Yard No. 57, of which Professor Forster was the Director, standardat 62.580 F, and similarto No. 58 in the Peirce obtained,and broughtto Americain Aug- collectionof standardyards at the Board of Trade ust, 1876, a brass normalmeter No. 49, for use in London.13 in his researchon gravity. In 1876 the United III States Lake Surveyacquired from the Repsolds a steel meterscale, the Repsold 1876 Meter, as its Charles S. Peirce becameconcerned with stand- standard.17 Like Meter No. 49, the Repsold ards of lengthwhen he undertookto determine 1876 Meter was a line-meter,that is, the standard the absoluteintensity of gravitywith a reversible betweentwo lines en- pendulum. Peirce was put in charge of determi- of lengthwas the distance nationsof gravity for the Coast Surveyon Novem- gravedon a bar. In BulletinNo. 17 of the Coast ber 30, 1872.14 He firstconducted observations in and Geodetic Survey, dated October 11, 1889, 1873 at Massachusettsstations on the times of "The Relation betweenthe Metric Standards of swing of an invariablependulum for the deter- Length of the United States Coast Survey and minationof relative values of gravity. He re- the United States Lake Survey," C. A. Schott ceived authorization,however, to order fromthe and 0. H. Tittmannstate, firmof A. Repsold and Sons, Hamburg,an appa- for the It is well knownthat all resultsof the Coast and ratus with a Bessel reversiblependulum, GeodeticSurvey relating to lengthsare expressedin determinationof absolute values. Now the de- termsof the Committee Meter, and thatcertain metric terminationof absolute values of gravitywith a lengthsgiven by the Lake Surveyare expressedin reversiblependulum requires the measurementof termsof theRepsold meter of 1876; also thatspecial periodsof oscillationabout each of two knife-edges use has beenmade by theCoast and GeodeticSurvey of the BerlinMeter No. 49 in observationsfor the and also the measurementof lengths,such as the intensityof gravityat variousplaces at home and distance betweenthe knife-edges. Since a high abroad.'8 order of accuracy was sought,the measurement of lengths pertainingto a pendulum required A descriptionof Meter No. 49 was given by authenticstandards of length. The Committee Professor Forster in a letter to Peirce, dated Meter of the Coast Survey was an end-measure; March 29, 1878. As translatedfrom the German, Peirce broughtto the United States fromBerlin a he stated, line-meterfor which he claimedhigh accuracy.15 No. 49 is a brassbar witha shoulderat each end. in horizontal 13 American and British The surfacesof theseshoulders are a J. E. Hilgard, "Comparison of the axis of the bar. Let into Standard Yards," Report of the Superintendent of the planepassing through States Coast Survey for the fiscal year ending United 16 of Gravity at Initial 1880), Appendix No. 12, C. S. Peirce, "Measurements June 30, 1877 (Washington, and Europe,"Report of the Super- Further references to Reports of the Superin- Stationsin America p. 154. intendent,1876 (Washington,1879), AppendixNo. 15, will be writtenReport of the Superintendent,1877. tendent 202-337,410-416. Yard No. 11 and Yard No. 57, along with the Troughton pp. the Committee Meter, are exhibited in the 17 Lieut. Col. C. B. Comstock,Brevet Brigidier-Gen- scale and Triangulation Vault at the National Bureau of Standards, Washington, eral, U. S. A., "Reportupon the Primary D. C. A photograph of the exhibit is reproduced on a of the UnitedStates Lake Survey,"Professional Papers 24 (Washing- and Measures Standards of the United of the Corps of Engineers U. S. Army No. plate in "Weights to V. National Bureau of Stand- ton, 1882), Appendix I, "AdditionalData Relative States," by Lewis Judson, V 28, Miscellaneous Publication No. 247, issued October, MetreR 1876,"pp. 835-862; Appendix (February ards, by C. B. 1963. 1885), "Value of Metre R 1876,"pp. 923-925, 14 Carolyn Eisele, "Charles S. Peirce, Nineteenth Comstock. Century Man of Science," Scripta Mathematica 24 18 C. A. Schott and 0. H. Tittmann,"The Relation (1959): pp. 305-324; specific reference,p. 308. betweenthe Metric Standardsof Lengthof the United Lake Survey," 15 Carolyn Eisele, "The Charles S. Peirce-Simon New- StatesCoast Surveyand theUnited States comb Correspondence," Proc. Amer. Philos. Soc. 101 5 Coast and Geodetic Survey Bulletin No. 17 (Washington, specificallyp. 165. (1957): p. 429. 1890), pp. 165-173, VOL. 109, NO. 1, 19651 CONTRIBUTIONS OF PEIRCE TO METROLOGY 33

_jiLi.ji...00..0..g.i0j;j;jl-K"s~~~~~~~~~~~~~~~~~~...... 0,...... GO~~~OF MIfl$*? _ ~ ~ ~ ...... ~.~.~ A ...... ~

SRQN'. Z ......

4 .....~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.....I j

i_ ------

ctan th hitoi ComteeMtr,Bronz Yar No 11 an Irn YadN. 57 hTogtn sae is in,a

case6:#on-''the--''right.CourtesyoftheNationalBureaufSt .,., s.

FIG. 1. The StandardsVault at the NationalBureau of Standards,Washington, D. C. The case in the illustration conitainsthe historic CommitteeMeter, Bronze Yard No. 11, and Iron Yard No. 57. The Troughton scale is in a case on the right. Courtesyof the National Bureau of Standards.

the surface of each shoulder is a silver plate having StandardNo. 49 was carefullycompared at dif- threelines ruledon it.19 ferenttemperatures with standard meter No. 1605 19 Supra, n. 16, pp. 302-305. Professor F6rster wrote belongingto the German StandardsCommission. four lettersto Peirce concerningthis meter,letters dated: No. 1605 had been comparedrepeatedly with a November 15, 1877, March 29, 1878, June 24, 1878, and November 15, 1880. These lettersare in U. S. Bureau of platinum meter belonging to the Commission. Weights and Measuires 13 (1840-1889) Length, Weight. The lengthof the platinum meter was derivedfrom National Archives. The discovery of this meter among comparisonsmade in 1860 between it and the the many standards preserved at the National Bureau of M etre des Archivesand the M etre du Conserva- Standards, Washington, D. C., was made by Mr. Walter toire. Boththese latter of W. Weinstein, Records Officer,with the aid of a report meters,constituted plat- from Mr. Larry Chisholm of a record at the Office of inum,had been standardizedduring operations for Weights and Measures. the establishmentof the Metric System. The di- 34 VICTOR F. LENZEN [PROC. AMER. PHIL. SOC. rectcomparisons between the platinummeter and for 10 C, between30 and 180 C.21 It may be meterNo. 1605 were not as accurateas the direct notedthat Schott and Tittmann,in the previously comparisonsbetween the platinummeter and No. cited reportof 1889, obtainedfor the coefficient 49. Accordingly,Professor Forster gave the re- of expansion of No. 49 the value 18.795 X 10-, sults of the comparisonsbetween No. 1605 and with probable error 0.031 x 10-6, a value with No. 49 for futurereference, but determinedthe whichPeirce's value was declaredto be in "good lengthof No. 49 fromthat of the platinummeter. accord." He stated that the lengthof the platinummeter The comparisonof bars whichserved as stand- was not known with sufficientaccuracy, in view ards of lengthwas carried out on a comparator. of the differencein values obtained from com- A concise descriptionof this instrument,of the parisonsof the platinummeter with the Metre des formduring the period of Peirce's work,may be Archivesand the Metre du Conservatoire. Since given by quotingfrom the definitionof the term the directcomparison of the platinummeter with writtenby Peirce forthe CenturyDictionary. A the Metre des Archives was less completethan comparatoris its indirect comparison through the Metre du Conservatoire,Professor Forster gave as the an instrumentfor comparingthe lengthsof nearly equal bars,either from end to end or betweenlines lengthof No. 49 the value derivedfrom the indi- engravedon them. The usual opticalcomparator has rect comparison,namely, in millimeters, two microscopes,firmly attached to a bar or some- thingof the sort,with their focal planescoincident No. 49 = 999.9948 + 0.01869 t, and furnishedwith filarmicrometers, whose screws lie virtuallyin oneright line. Thereis also a carriage where t is the temperaturein degreesCentigrade, movingat rightangles to the screwsso as to bring a value for No. 49 subject to a much greaterun- firstone bar and thenanother under the microscope. certaintythan is indicatedby the probable error The microscopesare focusedupon the rulingsat of the comparisonsutilized, on account of the the ends of the standardbar, withthe cross-hairs imperfectionsof the Paris results. in theeyepieces accurately set upon the lines. The The Repsold pendulum apparatus included a small differenceis measuredby notingthe revolu- meter for measuring the distance between the tionsof the micrometerscrews necessary to bring knife-edges.In 1875 and 1877 Peirce made com- the cross-hairsover the lines. To comparean end parisonsin Berlinbetween the United States pen- measurewith a line measure,such as No. 49, con- dulum meter and the German pendulummeter. tact cylinderswere used and consideredpart of The mean of theseresults showed that the reading the line measure. of the U. S. pendulummeter exceeded that of Since the microscopesof comparatorswere set the Germanpendulum meter by 131.6 .t2O withscrew-micrometers, the evaluationof a revo- One of the problems in the comparison of lutionof a micrometerwas an essentialoperation. lengthswas the determinationof coefficientsof For this purpose, glass scales were used which expansion,since data were obtainedfor different had been ruled by a ruling machine,in Peirce's temperatures. Peirce, therefore,had to deter- timethe rulingmachine of L. M. Rutherfurd.A minethe coefficientsof expansionof the pendulum fundamentalscale was the United States glass and of the meter scales. The pendulum-meter,decimeter scale of centimeterswhich was upon a which was assumed to have the same coefficient piece of glass 0.29 cm. thickand roughlycut into as the pendulum,was comparedat differenttem- rectangularshape of 14.3 cm. by 5.3 cm. This peratureswith Meter No. 49. Peirce also had scale had been made by J. E. Hilgard, and the made a metermarked U. S. C. S.-C. S. P.-1878-A, differentcentimeters of the scale were compared for the purpose of determiningthe coefficientof togetherin 1872 by C. A. Schott.22 expansion of No. 49. Comparisons between Peirce's historicallysignificant memoir, "Meas- MeterNo. 49 and MeterA weremade at tempera- urementsof Gravityat Initial Stationsin America tures between30 and 180 C. The coefficientof and Europe," Appendix No. 15 to the Report of No. 49 was checkedby comparisonat varioustem- the Superintendentof the Coast Surveyfor 1875- peratureswith the platinummeter of the German 76, contains the records of many measurements Imperial Normaleichungsamt.The value of the coefficientof No. 49 was givenby Peirce as 18.83 21 Supra, n. 16, p. 272. The author is indebted to Mr. Weinstein for the discovery of the Peirce Meters A and B. 20Supra, n. 16, p. 305. 22SSupra,n. 16, p. 282. VOL. 109, NO. 1, 1965] CONTRIBUTIONS OF PEIRCE TO METROLOGY 35

withcomparators. In orderto illustratethe metic- 1 mm.of 681 teeth= 1.0002mm. ulous experimentalwork required to comparebars 1/10 mm.of 68 teeth= 99.87 micron. with respectto length,we shall describesome of 1/100mm. of 7 teeth= 10.28 micron. the operations.23 For the measurementsin the just citedreport, Peirce used glass centimeterNo. The U. S. Coast Survey pendulummeter was 1, which had been ruled upon the best ruling- similarto the Germanpendulum meter. At one machineof Rutherfurd.This centimeterwas on a end, the meterbegan withthree lines ruledat dis- piece of glass 0.20 cm. thick,3.5 cm. broad, and tances of tenthsof millimeters.At the otherend, 3.6 cm. long. Though used January 16, 1878, there was a scale of similarlines. In February, it was marked"January 18, 1878, No. 1, 18 329/ 1878, Peirce compareda part of this scale with a 360 1 rev." The lines upon the scale were about millimeterscale of tenths(No. 3) ruledon Ruther- 2 cm. long and one micronbroad. They were furd'smachine. He foundthe mean value of the filled with black lead and varnishedover. The divisionsof the glass scale to be 0.995 r (revolu- limitinglines of the centimeterwere each midway tions), whencehe assumed 1/10 millimeterto be betweentwo bundlesof 10 lines each, distantfrom equal to 0.996 r. From thisresult he foundvalues one another5/360 rev. of the micrometerscrew. of the spaces betweenthe lines on thebrass scale of The limitinglines were distant45/360 rev. from the pendulummeter. He also made a studyJanu- the nearestlines of the bundles. The extremities ary 21, 1878, of the value of the screwrevolutions of the limitinglines were marked by crosses of the upper microscopeof the Repsold vertical roughlycut by hand. The measures were made comparatorwith which he comparedthe lengthof over a longitudinalline, which was marked by the pendulum(distance betweenthe knife-edges) crosses at the two ends. with his pendulummeter which was suspended Peirce compared this glass centimeterNo. 1 alongside the pendulum. The value for one with the United States glass decimeterscale of revolutionbetween 1 r and 3 r was'found to be 0.1005 centimeters. On January 16, 1878, he super- mm. For the lower microscopehe ob- posed the glass centimeterface to face on centi- tainedthe value 0.1004 mm.for one revolutionof meter 5 to 6 on the upper glass decimeterscale the micrometerscrew. of centimeters. The differencewas measured Peirce's Meter No. 49 was the object of com- upon a Rutherfurdscrew micrometer. Centi- parisonby othermetrologists. Professor William meterNo. 1 was also comparedwith the U. S. A. Rogers, while a memberof the staffof the centimeter3 to 4. On January17, measureswere Harvard College Observatory,undertook a critical made on the same Rutherfurdscrew micrometer, study of certain standardsof length,which had using revolutions551 to 150k, of the U. S. 5 been compared eitherdirectly or indirectlywith centimetersfrom 0 to 5, the 5 centimetersfrom 5 the originalstandards, i.e., the Imperial Yard at to 10; and there were also made 5 measures of London and the Metre des Archivesat the Inter- centimeterNo. 1, so as to cover the same part of nationalBureau of Weightsand Measures.24 One the screw. The conclusion of this meticulous of the instrumentswith which he workedwas the study of the centimeterscale yielded the value Tresca Meter,a bar of copper withnarrow strips 1.0001 cm. for the length of the centimeter. of platinuminserted at each end. The defining Peirce stated that this centimeterwas equal to lines were traced on the platinumsurfaces at 2 6,809k teeth of Rutherfurd'smachine, and as it A.M., February6, 1880, by directcomparison with was 1/10,000 too long, he concludedthat 6,809 a meterdesignated No. 19 at the Conservatoire teethmade a centimeterat ordinarytemperatures, des Arts et Metiers,Paris, by G. Tresca, the son say about 180 C. Peirce statedthat from another of ProfessorTresca of the Conservatoire. Rogers investigationit had been ascertainedthat the ac- also had a bronze Yard and Meter, made of cidental error of position of a line ruled upon Rutherfurd'smachine was only ?0.01 micron; 24 WilliamA. Rogers: (1) "Communicationto F. A. P. the periodic error differedgreatly on different Barnard, December 26, 1883," Proc. Amer. Metrological Soc. 4 (New York, 1884): pp. 109-116. (2) Ibid. 5 plates,but never amountednearly to 0.1 micron. (New York, 1885) : pp. 33-48. (3) "Studiesin Metrol- He adoptedthe followingvalues: ogy," Proc. Amer. Acad. Arts and Sciences 18 (1882- 1883): pp. 287-398. (4) "Reportto the Prattand Whit- 23The operations are described in the cited memoir ney Co.," Standards of Length and their Practical Appli- (supra, n. 16) on pages 286-298. The instrumentem- cation,edited by George M. Bond, M. E. (Hartford, ployed was a filar micrometer. Conn.,1887), pp. 1-48. 36 VICTOR F. LENZEN [PROC. AMER. PHIL. SOC.

Baily's metal, of which the Imperial Yard was able errorin length+0.70 microns. Since Meter composed. This standardformed the basis of the No. 49 of Peirce was a secondarystandard, they unitswhich he undertookto obtainfor the gauges comparedit withthe CommitteeMeter and found of Pratt and Whitney, Hartford, Connecticut. that The defininglines of the Yard were obtainedindi- = rectlyfrom the ImperialYard throughcomparison No. 49 CommitteeMeter with Bronze Yard No. 11 of the Coast and - 14.18 ? 0.77 microns. Geodetic Survey, January and February, 1881. The Repsold 1876 Meter had been sent to Paris But duringthis period Rogers also made compari- and in January,1883, compared with the standard sons betweenthe Tresca Meterand Meter No. 49. at the InternationalBureau of Weightsand Meas- He foundthat ures. The result was that the Repsold 1876 to the provisionalmeter + 97.81 No. 49 + 11.1 = Metredes Archives. Meter was equal microns,at the temperatureof meltingice. In This resultwas comparedwith the values of No. consequence, the CommitteeMeter was deter- 49 given by ProfessorForster. We have pre- minedto be viouslyquoted F6rster'svalue forNo. 49 obtained 1 Meter (provisional)- 0.38 + 0.70 micron.25 throughcomparison of the platinummeter with the Metre du Conservatoire. The relation be- In a footnote,written after the National Proto- tween No. 49 and the Metre des Archives ob- typeMeter No. 27 had been broughtfrom France tained by this indirectmethod may be stated by in 1889, to serve as the United States standardof length,Schott and Tittmannstate that the value No. 49 + 5.2 ,u= Metredes Archives. assumed for the provisionalmetre, '2 = 1 + 6 u, Rogers, however,also used the relationobtained had been confirmed. In view of the assignment fromthe less completecomparison of the platinum of a new value to No. 49, differentfrom the value meter directly with the Metre des Archives, given preferenceby Forster, Schott reportedin namely, 1895 to F. R. Helmert, Director of the Royal Prussian GeodeticInstitute, Potsdam, and Direc- No. 49 + 21.4 , = Metredes Archives. tor of the Centralbureaufor the International GeodeticAssociation, that lengthsof the second's In statingthe above relationsthe temperaturewas pendulum,as determinedby Peirce withmeasure- takenas zero. The averagevalue forthe relation mentsbased upon No. 49, had been changed by betweenNo. 49 and the Metre des Archivesthen -9.4 microns.26It would appear fromthis result was thatPeirce would have been advised to have used No. 49 + 13.3 y = Metredes Archives, the average of the two values forNo. 49 given by F6rster,instead of the preferredvalue based on a result that did not differmarkedly from the an indirectcomparison, through the Metre du value obtainedby Rogers fromcomparison with Conservatoire,of the German platinum meter the Tresca Meter,namely, with the Metre des Archives. Neverthelesson with the 1876 No. 49 + 11.1 ju= Metredes Archives. comparingNo. 49 Repsold Meter, Peirce's claim that he broughtthe firstaccurate Schottand Tittmann,in a previouslycited pub- line-meterto the United States appears to be lication,also reporteda new value for Meter No. justified. 49. The principalpurpose of theirinvestigation Peirce also made a contributionto the creation was to compare the Repsold 1876 Meter, the of a standardof measurementfor micrometry.27 standard for the Lake Survey, under General 25 B. as withthe Com- Supra, n. 18, p. 173; also supra, n. 17, Appendices I C. Comstock Superintendent, and V. mittee Meter, the standard for the Coast and 26 Verhandlungender vom 25 September bis 12 October Geodetic Survey,in order that the resultsof the 1895 in Berlin abgehaltenenElften Allgemeinen Conferenz Lake Surveycould be tied to the transcontinentalder Internationale Erdmessung (Berlin, 1896), Beilage surveyfor the connectionof the surveyson the A. I., p. 167. 27 The documents for this topic are in Proc. Amer. Atlanticand PacificCoasts. They foundthe Rep- Soc. Microscopists, Sixth Annual Meeting, August 7-10, sold 1876 Meter to be 98.19 micronslonger than 1883, (Chicago, Illinois) 5: pp. 178-200. Courtesy of the CommitteeMeter, both at O C, witha prob- Max H. Fisch. VOL. 109, NO. 1, 19651 CONTRIBUTIONS OF PEIRCE TO METROLOGY 37

In 1879 a National Committeeon Micrometrywas Chicago meetingof August, 1883. The Society formedunder the chairmanshipof F. A. P. Bar- announcedrules for the controlof the standard nard. The secretaryof the committee,R. H. micrometer. Ward, reportedon "The Standard Micrometer, IV Scale A, 1882," to the American Society of Microscopists at the annual meeting, August, As we have seen, Peirce's first studies in 1883, in Chicago, Illinois. He reportedthat the Metrologywere of Meter No. 49 and its relation committeehad recognizedthe need to select or to the fundamentalmetric standard. He also par- preparea standardmicrometer, and ticipated,however, in the investigationof British standards. In order to place his contributionin aftermuch conference and correspondence,it was de- its historical setting,we shall provide further cidedto procure,from a sourcecapable of givingit background. originallyan officialcharacter, a new scale as nearly indestructibleas possibleand of carefullydetermined We have noted earlier that the Troughton value. The U. S. Bureauof Weightsand Measures, scale,the Bronze Yard No. 11, and the Low Moor throughthe kindnessof Prof.J. E. Hilgard,under- Iron Yard No. 57 representedthe English stand- took to prepareand authenticatesuch a standard, ard of lengthin . . . and a scale excellentlyruled on a platiniridium the Officeof Weightsand Meas- bar and verifiedwith great care by Prof.C. S. Peirce ures of theUnited States Coast Survey. Since the was placedat thedisposal of the committee in August, discussionnow will referto the yard,it will be of 1882. A subcommitteeon testingthis micrometerinterest to mentionsome steps in the creationof was appointedon whosebehalf Professor William A. the BritishImperial Yard. For this Rogerssubjected the plate to a prolongedand elab- purpose,it is orate studywhich was not completeduntil August, instructiveto quote fromthe definition of theword 1883. yardwritten by Peirce about 1890 forthe Century Dictionary. ProfessorHilgard's reportstated that the scale was divided into ten millimeters,each division Yard. The fundamentalunit of Englishlong meas- being marked by three lines distant from one ure. The prototypeof the Britishimperial yard (to whichthe United States and the measurementwas to Officeof Weightsand Meas- anotherten microns, ures conforms,though without express authority) be made fromthe mean positionsof one tripletof was legalizedin 1855. It is a bar madeof a kindof linesto thatof another. The firstmillimeter again bronzeor gunmetalknown as Baily's metal. It has was divided in the same manner into tenthsof a square sectionof 1 inch on the sides and is 38 inches millimeters.The firsttenth of a millimeterwas long. But at 1 inchfrom each end a well is drilledinto one of its surfacesso thatthe bottomis subdividedinto ten spaces of ten micronseach. in thecentral plane of thebar, and intothe bottom of There were thirteenof theselines at the beginning thewell is sunka goldplug, upon whose mat surface of the centimeter,the firsttenth of a millimeter is engravedone of the two defininglines. The beingmeasured from the mean of the firstthree to yardis definedas thedistance between these lines at 620 F., withthe understandingthat the bar is to be the mean of the eleventh,twelfth, and thirteenth. supportedin a particularmanner, and that the ProfessorHilgard gave the correctionsto the dif- thermometersare to be constructedaccording to cer- ferent divisions of the Scale A, 1882, at 70 tainrules. The linesare designedto be lookedat with degreesF. themicroscopes of a comparator;but they are notso freefrom blur thattheir In his studyof CentimeterA, ProfessorRogers middlescan be determined morenearly than to a millionthpart of the distance used as the originalunit the Tresca Meter and a betweenthem. This standardwas made afterthe bronze bar to which the Tresca Meter had been practicaldestruction of the previouslegal prototype, transferred. He also used the data concerning thatof 1760,in the burningof the Houses of Par- Peirce's Meter No. 49 and comparedthe bronze liament,October 16th, 1834, and was legalizedas a new prototypebecause its lengthagreed with what meterwith No. 49. ProfessorRogers concluded had been recognizedin 1819by the StandardsCom- that CentimeterA, definedby the middlelines of missionas thescientific standard yard-namely, with the terminalbands, requiredno sensible correc- a certainscale, or ratherwith Captain Kater's meas- tionat 62.0? F, and thatthe secondmillimeter was uresof that scale, known as Shuckburgh'sscale, having beenmade in 1794 one tenthpart of the standard unit and byTroughton for Sir GeorgeShuck- exactly burgh,who in his comparisonsof it firstintroduced requiredno correction. the comparatorwith micrometer microscopes. This The National Committeeon Micrometryac- scale was a copyof anotherwhich had beenmade for the cepted the plate and unanimouslytendered it to Royal Societyin 1742,from which the standard of 1760was copied. This was a bar havingupon one the American Society of Microscopists at the side twogold studs,each witha dot prickedupon it; 38 VICTOR F. LENZEN [PROC. AMER. PHIL. SOC. and it was used by bringingthe pointsof a beam- The iron CommitteeMeter, and the platinum compassinto these dots, which had thusbecome badly Arago Kilogramobtained in 1821 by AlbertGal- those of worn. Older standardsstill extant are latin,then U. S. Ministerto France, were avail- Queen Elizabethand of HenryVII. The latteris shorterthan the present yard by one thousandthpart able as standards. of its length,or about 1/27 of an inch. It is said Since the BritishImperial Yard as a standard thatthe yard was madeto be thelength of HenryI's of lengthfor the United States was representedby arm-doubtlessa fable,even if believedby thatmon- archhimself. the Troughtonscale and Yards Nos. 11 and 57, it was desirableto make intercomparisonsbetween Peirce's critical account of the yard was based them. This workof comparisonwas done by As- upon first-handobservation, as we shall have sistantJ. E. Hilgard, with the assistance of J. occasion to learn presently. Homer Lane, H. W. Blair, and 0. H. Tittmannof Hassler, who broughtthe CommitteeMeter and the Coast and GeodeticSurvey, between 1876 and the Troughtonscale to the United States, also 1879. Hilgard was at the time in charge of the determinedthe lengthof the CommitteeMeter in Officeof Weightsand Measures, and had been a terms of the Troughton scale and thereforein United States delegate to the meetingsof the termsof the Yard. He foundthat the Committee internationalcommittee which drafted the Metric Meter was 39.38091714 inches of the Troughton Convention,signed on May 20, 1875, providing scale, with meter and scale at 32? Fahrenheit. for the establishmentof an InternationalBureau Alexander Dallas Bache, Superintendentof the of Weights and Measures in France. In the Coast Surveyfrom 1843 to 1867, using Hassler's course of the intercomparisons,Yard No. 11 was coefficientof expansionas determinedfrom a brass sent to Ottawa, Ontario,in May, 1877, for com- wire,inferred from Hassler's resultthat the Com- parison with the copies of the British standards mitteeMeter was 39.368505inches of the Trough- there. Then, in July, 1878, Hilgard took Yard ton scale at 62? F, and the value 39.3685 inches No. 11 to England,and withthe help of Mr. H. J. was used in the Coast Surveyfrom 1851 to 1868. Chaney,Warden of Standards,Board of Trade, The CommitteeMeter, as we have alreadynoted, London, compared No. 11 with the British Im- was the standardof lengthfor the Survey from perial Standards,Yards No. 1 and No. 6. Hil- 1817 to 1890; this Meter was sent to France by gard's reports,dated July 8, 1879 and July 10, the United States Governmentfor comparison 1880, were publishedas appendicesto the Reports with the standard platinummeter of the Con- of the Superintendentof the Coast and Geodetic servatoireImperial des Artset Metiers. A report, Surveyfor 1876 and 1877,respectively.29 Hilgard dated March 5, 1864, had stated that the Metre foundthat Yard No. 11, initiallylonger than the du Conservatoirewas 0.00329 millimeterlonger British Imperial Standard by 0.000072 inch in than the Metre des Archives. A comparisonof 1853, had become shorterby 0.000088 inch, that theUnited States CommitteeMeter and the Metre is, No. 11 had shortened0.000160 inch, with re- du Conservatoire,on August 24, 1867, by the spect to the Britishstandard, from 1853 to 1878. American physicistF. A. P. Barnard and H. No. 11 was now equal to the British standard Tresca, yielded the result that the Committee at 62.250 F. Based upon comparison of the Meter was longer than that of the Conservatoire Troughtonscale with No. 11, Hilgard inferred by 0.00007 mm. The result of the comparison that at 62? F the Troughtonscale was too long was thatthe United States CommitteeMeter was by 0.00083 inch and standardat 59.62? F. Hil- 1.000 00336 metersat the temperatureof melting gard also inferredthat No. 11 had becomeshorter ice, forwhich comparison the Metre des Archives than No. 57 by 0.00025 inch in the twenty-five was the basic standard.28 years. The metric system was made legal for the It has been noted that Bache had derivedfrom United States by approvalof an Act of Congress, Hassler's value of the CommitteeMeter that it July 28, 1866. The meter was declared to be 39.37 inches and the kilogram 2.2046 pounds. 29 J. E.,, Hilgard, "Relation of Lawful standardsof Measure of the United States to those of Great Britain 28 F. A. P. Barnard and H. Tresca, "Report upon the and France," Report of the Superintendent,1876 (Wash- Comparison of an Iron Meter forwardedto France by the ington,1879), Appendix No. 22, pp. 402-406. J. E. Government of the United States of America," Report Hilgard,"Comparison of Americanand BritishStandard of the Superintendent, 1867 (Washington, 1869), Ap- Yards," Report of the Superintendent,1877 (Washing- pendix No. 7, pp. 134-137. ton,1880), AppendixNo. 12, pp. 148-181. VOL. 109, NO. 1. 1965] CONTRIBUTIONS OF PEIRCE TO METROLOGY 39 was 39.368505 inches of the Troughtonscale at been shortenedby molecularchanges in the alloy. 620 F. On thebasis of comparisonsof the British Bronze Yard No. 11, however,was the United Imperial Standards with several well-accredited States Standardfor the Yard from1856 to 1893; copies of the meter irrEngland, and with the aid Iron Yard No. 57 was a secondarystandard. of well-determinedcoefficients of expansion,Lieu- Hilgard in his reportof 1880 stated that Iron tenantA. R. Clarke at the Officeof the British Yard No. 57 had been sent to England for direct Ordnance Survey determinedthe meter to be comparisonwith the British imperial standards, 39.370432 inches, and after 1868 the value of and especiallywith its fellowof Low Moor iron, 39.3704 inches was used by the Coast Survey. Yard No. 58. On the occasionof a tripto Europe Hilgard pointedout that,if one appliedto Bache's in 1883, Charles S. Peirce carriedout such com- value of the Committee Meter the corrected parisonswith the assistanceof Mr. H. J. Chaney. Troughton scale, and also used the coefficient The Report of the Superintendentfor the year of expansionof bronzealloys of whichthe imperial ending June 30, 1884, states concerningPeirce: standardswere made,the CommitteeMeter would be 39.37054 inches, a result that differedlittle In Juneand July,1883, he carefullycompared the fromClarke's value. If one used forthe coefficient Coast and GeodeticSurvey Standard Yard No. 57 withthe Imperial Yard No. 1, and also withthe Iron of expansion of bronze the value determinedby Yard No. 58at theBritish Standards Office, London.31 Fizeau, one would obtain the value of 39.37023 inchesfor the meter. On July6, 1883, Peirce reportedto Hilgard, now The difficultiesin determiningthe relationbe- Superintendentof the Coast and Geodetic Sur- tween the meterand the yard were set forthby vey, that he and Mr. Chaney had made 24 com- Peirce in a reportto the Superintendentin 1884.30 parisons each of Yard No. 57 with the Imperial The main difficultieswere four: Standard No. 1, with temperaturesfirst below 620 F, and thenabove 620 F. Such 1. To obtain a length known to be equal to comparisons were on a meter. finished July3, butthen micrometers were compared, the values 2. To compare quantities practically incom- especially of the screw revo- mensurable. lutions. Comparisonswere also made between Yard No. 57 and its 3. To comparetwo bars, one of whichis stand- fellowNo. 58, fromJuly 6 ard at the freezingpoint while the otheris stand- to July 12, 1883. There were twelve sets by ard at 620 F. Peirce and six by Chaney. The bars were trans- 4. To compare a line-measurewith an end- posed and the temperatureswere between660 F measure. The imperialyards were line-measures, and 680 F. Peirce's resultswere concordant,but the Meter des Archiveswas an end-measure. differedmarkedly from those of Chaney,owing to the peculiarityin rulingof one line. In his report As we have seen, Hilgard foundthat Yard No. to Hilgard, Peirce stated that his comparisonsof 11, originalfellow of BritishImperial Yards No. 1 No. 57 and No. 58 and No. 6, had shortenedwith referenceto them in twenty-fiveyears by the amount0.000160 inch. are very concordantand make No. 57 almostfour Comparisonof No. 11 with No. 57 showed an divisionsof the micrometershorter. This accords withthe even greater shorteningof No. 11 with comparisonsof 1853. But the bisectionof respect one of the lines of No. 57 is a matterof possible to No. 57, so thatIron Yard No. 57 was inferred doubt,and in my opinion,the yard in the United to have becomelonger, relative to BritishImperial States oughtnot to rest upon comparisonsof this Yard No. 1, by 0.000108 inch, althoughit was bar. Eitherthe bar No. 11 shouldbe taken,or better, statedthat this might have been coveredby errors a newyard should be comparedwith No. 1 afterthe completionof the new comparator.32 of observation. Hilgard, however,expressed the opinion that the excess of No. 57 mightbe real. The resultsof the comparisonsof Yard No. 57 The bronzeof the imperialstandards, and also of with Imperial Yard No. 1, after Peirce had re- No. 11, was Baily's metal,consisting of 16 parts duced the observations,were reported to the of copper, 21 parts of tin, and 1 part of zinc; House of Commonsby the Board of Trade, Au- the British Yards No. 1 and No. 6 mighthave 31Ibid., p. 40. 30 Report of the Superintendent, 1884 (Washington, 32Assistants' Reports, U. S. Coast and Geodetic Sur- 1885), p. 81. vey,National Archives (Washington, D. C.). 40 VICTOR F. LENZEN [PROC. AMER. PHIL. SOC. gust 9, 1883, over the signatureof T. H. Farrer.33 be heavierthan the other,and in order to attain The reportstated: this asymmetrywith respectto weightwith the cylindricalform, the compensatingweights were Duringthe past year the StandardsDepartment has placed withinthe cylinder. The Peirce pendulums had the opportunityof assistingthe United States Governmentin a comparisonof theirstandard of were the firstpendulums thus constructed,al- lengthYard No. 57 withthe Standardsof thisoffice. thoughcredit for this innovationhas been given ProfessorC. S. Peirce,of the UnitedStates Coast to Defforgeswhose worksucceeded that of Peirce. and GeodeticSurvey, came to Londonfor thispur- The functionof the Peirce pendulumswas to carry posein Junelast, on behalfof ProfessorJ. E. Hilgard, whohas chargeof theBureau of Weightsand Meas- on researchinto the intensityof gravity. It was uresat Washington.A largenumber of comparisons intendedthat the invariabilityof length of the of thesemeasures was made withall possiblecare, pendulumwould make possiblerelative determina- and it was foundthat at 620 Fahr.,Yard No. 57 was tionsof gravityfrom times of swing. The rever- 0.000022inch longer than the Yard No. 1 deposited sibilityof the pendulum would serve to check at this office.The resultsof these comparisonsas calculatedby ProfessorPeirce, will be referredto in occasionallyon the invariableproperty. Experi- a printedmemorandum which will be separately mentationwith pendulums served Peirce forother drawnup. investigationsalso. The cylindricalform of the Peirce pendulums enabled him to test Stokes' It maybe added thatProfessor William A. Rogers theoryof motionof a cylinderin a viscousmedium in his criticalstudy of standardsof length,pre- fromthe rateof decreaseof the amplitudeof oscil- viously mentioned,reported that Peirce found In Yard No. 11 to be 0.000022 inch shorterthan the lation of a pendulum. the presentdiscussion, Imperial Yard No. 1, whereas Hilgard had ob- the importantfact about the four Peirce pendu- tained the value 0.000088 inch.34 Rogers then lums was thatNos. 1, 2, and 4 were meterpendu- used the average of the two results, namely, lums,that is, the distancebetween the knife-edges 0.000055 inch,for his own determinations.The was one meter,whereas No. 3 was a yard pen- resultofficially reported for Peirce applies, how- dulum. ever,to Yard No. 57, and we have not been able Peirce's views on the use of pendulums for to find evidence for Roger's conclusions with metrologicalpurposes were set forthin the Report respectto Yard No. 11. of the Superintendentfor the year ending June Prior to Peirce's comparisonsof standardsin 30, 1884. Peirce is quoted as follows: 1883, the reportsof Hilgard had raised questions concerningthe permanenceof lengthof the vari- The ratioof the meterto the yardis stilla matter of considerableuncertainty. Kater's value of the ous standardsof lengthin the United States and meter,39.3707 inches, is universallyregarded as too England. Possible molecular changes in alloys long. Clarke's39.3704 inches,though undoubtedly and the effectof variationsof temperaturehad morenearly correct, is not foundedupon an exami- raised doubts of the invariabilityof length of nationof sufficientlygood meters. The value given by ProfessorW. A. Rogers,39.37027 inches, is prob- standards. As we shall see later, Peirce made ablythe most correct, but may be in errorby one or a pioneer contributionto this problemwith the twothousands of an inch.... As theamount in doubt aid of spectroscopicmethods. is large as comparedwith the best measuresof the lengthof the second'spendulum, we may obtaina V value of themeter not unworthy of considerationby comparingthe value of thesecond's pendulum at Kew During the years 1881-1882 there were con- as determinedby CaptainHeavyside in incheswith at the Officeof the Coast and Geodetic the same value as determinedin termsof the meter structed by myself.This givesfor the meter39.3700 inches. Surveyfour invariable, reversible pendulums after In orderto obtaina bettercomparison, two reversible the originaldesign of Peirce.35 These pendulums pendulums,made on thesame pattern, one measuring were designatedPeirce Nos. 1, 2, 3, 4, and were a yardand theother a meterbetween the knife-edges, cylindricalin formwith hemispherical ends. The were swungsimultaneously, each near its standard temperatureat theCoast SurveyOffice. The pendu- reversibilityof a pendulumrequired that one end lums were also interchangedso as to determineat thesame time their coefficients of expansion.36 33 British Sessional Papers, House of Commons, 27 (1883), p. 854. 36Report of the Superintendent, 1884 (Washington, 34Supra, n. 24, No. 4, p. 40. 1885), p. 81. Letter y used in Heaviside's name is in- 35Report of the Superintendent, 1882 (Washington, correct according to the Great Trigonometrical Survey 1883), p. 32. of India 5 (Calcutta, 1879). VOL. 109, NO. 1, 1965] CONTRIBUTIONS OF PEIRCE TO METROLOGY 41

The recordsof the Coast Surveyshow thatPeirce relationshipas definedin 1893: 1 foot= 1200/3937 swung No. 2, a meterpendulum, and No. 3, the meter. yard pendulum,at highand low temperatures,for The U. S. Survey Foot will be used until geo- the purpose of comparing the meter with the detic surveynetworks are readjusted. yard,a methodof comparisonof whichhe claimed The Eleventh General (International) Confer- to be the originator. Such experimentswere re- ence on Weights and Measures adopted a new ported by the Superintendentin the annual Re- definitionof the meterby the conventionthat the port,37but no resultswere published. meter= 1 650 763.73 wave lengthsof the orange- The finaloutcome of decades of experimenta- red line whichresults from the 2P10 5d5 transition tion with metersand yards was the manufacture of krypton86. The InternationalCommittee of and intercomparisonof new prototypemeters by Weightsand Measures,has also approvedas sec- the InternationalBureau of Weights and Meas- ondarystandards 4 additionaltransitions in kryp- ures. In 1889,Assistant George Davidson, of the ton 86, 4 in mercury198, and 4 in cadmium114. Coast and GeodeticSurvey, brought from France National PrototypeMeter No. 27 and National VI PrototypeKilogram No. 20. The seals of the We have seen thatintercomparisons of standard packageswere brokenand the standardsexhibited yards indicatedthat some, if not all, standardsof at Executive Mansion, January2, 1890, in the lengthhad changedwith time. The problemwas presence of President Benjamin Harrison, the of concernto the European Geodetic Association Secretaryof State, the Secretaryof the Treasury, (Europaische Gradmessung),as well as to scien- SuperintendentMendenhall, and some invited tists in England and America. Peirce's work guests.38 Meter No. 21 and KilogrammeNo. 4 with standardsor scales of lengthled him to at- were receivedfrom the InternationalBureau of temptto detect secular changes of metallicbars Weights and Measures in July,1890. used as standards. His methodwas announced SuperintendentMendenhall, in Bulletin 17 of to the European GeodeticAssociation at the third the Coast and Geodetic Survey, dated April 5, meeting,October 1, 1877, of the PermanentCom- 1893, stated that "the Office of Weights and mission at Stuttgart. As translated from the Measures, with the approval of the Secretaryof Proceedingsof the Commission,the followingis the Treasury,will in the future,regard the Inter- recorded: nationalPrototype Metre and Kilogrammeas fun- damentalstandards, and the customaryunits, the To Mr. Peircea verysuitable means of determining the invariabilityof lengthsappears to consistin re- yard and pound, will be derived therefromin ferringthe unitof lengthto the lengthof the light accordancewith the Act of July28, 1866." wave. To measurethe relationbetween these inter- The National Bureau of Standards initially vals and the lengthof the lightwave will be easy, continuedthe ruling of 1893. In an addendum since the lengthsof the formeror theirrelations to to Circular No. 572, "Calibration of the Line the millimetercan be obtainedin makinggratings that in finenessand accuracyleave nothingto be Standardsof Lengthand MeasuringTapes at the desired.40 National Bureau of Standards," issued June 4, 1956, it is stated that Peirce's work on the spectrum-meter,although historicallyimportant, is representedonly by frag- a redefinitionof the yard becameeffective July 1, mentarypublications, and thereforeit is usefulto 1959. The new definitionof theyard is quote his statementin a letterto Simon New- 1 yard= 0.9144 meterexactly. comb,dated June 10, 1899. He said, In the United States thereis an exceptionat the My workof comparingthe lengthof a bar with requestof the United States Coast and Geodetic themean ruling of a gitter-plate,with a view of ob- Survey: taininga check on secular changes in the length of bars,has manymerits. My measurementsof the Any data expressedin feetderived from and pub- directionof theray were more accurate perhaps than lished as a resultof geodeticsurveys within the any measurementsof a largeangle evermade. The United States will continueto bear the following 39 National Bureau of Standards Technical News Bul- 37Ibid., p. 40. letin 44 (1960): p. 199; Physics Today 17 (1964): p. 101. 38 Report of the Superintendent, 1890 (Washington, 40 Verhandlungender vom 26 September bis 3 October 1891), AppendixNo. 18,pp. 735-758,specifically pp. 750- 1877 in Stuttgart Vereinigten Permanenten Commission 751. der Europiiischen Gradmessung (Berlin, 1878), p. 40. 42 VICTOR F. LENZEN [PROC. AMER. PHIL. SOC.

peculiarcomparator I inventedand used enabledme acteristicwas due to a differencein the sides of to buildup froma doublecentimeter with a probable the groovesruled in glass. The diamondploughing errorof less thana millionthpart, which was quitea feat. True Rowlandfound a break in my gitter throughthe surfaceraised a burr on the side of whichvitiated my value of thewave lengthof light. the furrow,and hence made the two sides of the But thatdoes not,I believe,affect my main purpose, cut of unequal length. To correctthis, the groove as longas theplate can be usedagain.41 was filledwith black-lead,the burr polished off, The referencesto Rutherfurd'sruling machine and the black-leadremoved. Plates were thereby in the presentand precedingsections make it ap- obtained which gave spectra of the utmostbril- propriateto interpolatea descriptionof this ma- lance with no differencein brightness. chine which was an importantcontribution to The speculummetal, on whichRutherfurd also exact science in America and even in Europe. ruledgratings, consisted of 70 per centcopper and The rulingmachine of L. M. Rutherfurd,of New 30 per cent tin. The advantageof this material York, constituteda notableadvance over that of was that it was softer,so that the diamond did the German Nobert, who preservedsecrecy re- not dig intoit as muchand did not have the same gardinghis own machine. A briefdescription of dispositionto jump. This made it possible to Rutherfurd'smachine, as illustratedin the article carry out much,finer divisions than otherwise "Spectrum,"by ProfessorA. M. Mayer, in the would have been possible. The gratingscould be American Cyclopedia,second edition (1876), is used only in reflection,but as Kayser has pointed as follows:The plate of glass or metalto be ruled out, this was no disadvantagein photographyof by the machinewas carriedon a framewhich was the ultraviolet,since absorptionof radiation by made to advance in short successive steps in a the gratingmaterial was therebyavoided.43 Ac- directionparallel to the axis of a screw with 48 cordingto determinationsby Kurlbaum,most of threadsto the inch. The head of this screw was the Rutherfurdgratings were 43.36148 mm. wide formedby a circularplate the rim of which was and held 29,521 furrows,that is, 680 per mm. notchedby 360 equally-spacedteeth. An oscil- Peirce publisheda briefprogress report in the lating lever pressed a pawl against one of the AmericanJournal of Science in 1879 and stated notches of the circular plate and rotated the thatthe deviationof a spectralline (Van der Wil- wheel a definitefraction of its circumference,then ligen's No. 16) had been measuredin 1877 and liftedthe pawl and retractedit for anotherfor- 1879 using a certaingrating (3401 lines to the ward motion on the circular plate. While the mm.).44 In a shortarticle in Nature, 1881, Peirce screw was advancingthe plate carryingthe glass gave for the closest-ruleddiffraction plates of or metal to be ruled, the diamond pointed tool Rurtherfurd,the mean widthof rulingas 68,078 was raised and carriedforward above the surface to 68,082 lines to the decimeterat 70? F.45 He of the glass or speculumplate. Afterthe screw stated that there was a solar line well suited to ceased its rotation,the diamondpoint gentlyfell precise observationand gave the wave lengthas upon the glass or metal and then retractedand 5624825, subject to corrections,and proposed it cut a line. The reciprocatingaction of the tool as a standard. Rowland in 1882 inventeda ruling was caused by the action of the verticalarm of engine by which improvedgratings were made. a leverwhich was attachedto the same shaftwhich He also inventedthe concave gratingand with it carriedthe lever. The machinewas operatedby determinedrelative wave lengthsof lines in the the reciprocatingaction of a rod on the lever solar spectrum. On February14, 1884, Rowland througha crank-pinof a drivingwheel revolved was awarded the Rumfordmedal by the Ameri- by a small turbine. can Academy of Arts and Sciences. In his re- Rutherfurdruled gratings on glass and on spec- sponse to the address of presentation,Rowland ulum metal. For the gratingson glass, it was stated,"My map of wave lengthsis based upon noticedthat the rightand leftspectra of the same ProfessorCharles S. Peirce's measurementof a order were of differentbrightness. Peirce, who 43H. Kayser, Handbuch der Spectroscopie 1 (Leipzig, workedwith Rutherfurd, reported in a paper read 1900): p. 403. to the National Academy of Sciences, April, 44 C. S. Peirce, "Note on the Progress of Experiments 1879,42 that examinationshowed that this char- for Comparing a Wave-Length with a Meter," Amer. Jour. Science, ser. 3, 18 (July, 1879): p. 51. 41Supra, n. 15, p. 429. 45 C. S. Peirce, "Width of Mr. Rutherfurd'sRulings," 42 Reported in Nature 20 (1879): p. 99. Nature 24 (1881): p. 262. VOL. 109, NO. 1, 1965] CONTRIBUTIONS OF PEIRCE TO METROLOGY 43 line in the green portion of the spectrum."46 At all eventsit is quitesure that of thewave-length Rowland reported subsequentlythat the wave determinationsmade up to 1880,those of Peirce,and of Angstromcorrected by Thalen,are by all oddsthe length of the solar line adopted by Peirce as best. Of the two,Peirce's is probablythe better by standard,and given by Peirce as 5624.825, was reasonof bettergratings.50 correctedby Rowland to be 5624.86, and finally corectedby Rowland and Bell for errorof ruling Rowland stated: "The determinationof Mr. C. and standardof lengthto 5624.66. The unit in S. Peirce of the U. S. Coast Survey is certainly which Rowland stated the foregoingresults was a veryaccurate one." 51 Rowland gave the wave the Angstromunit, equal to 10-10or 1/10,000,000,- lengthof the D1 line of sodium,as derivedfrom 000 meter. Peirce's standard,as 5896.27. Correctingfor the Records of the Coast and GeodeticSurvey indi- errorin the ruling -.07 gave the result 5896.20 cate thatPeirce prepareda reporton thespectrum- Angstromunits, a result that agreed with the meter,which was reviewedand praised by Pro- value obtained by Bell. The Rowland Primary fessor Benjamin Peirce, ConsultingGeometer to Standard for the Solar Spectrum was the D1 the Survey,47but whichwas not publishedon ac- line of sodium,based upon the work of five in- countof lack of space. Peirce's methodhas been vestigators,including Peirce.52 The individual -describedby Michelsonand Morleyas follows: values, with weightsassigned by Rowland, were as follows: The methodinvolved two distinct measurements, first that of the angulardisplacement of the image of a Investigators Wave Length Weight slit by a diffraction-grating,and second, that of the Angstrom,corrected by Thalen 5895.81 1 distancebetween the lines of a grating.48 Mullerand Kempf 5896.25 2 Kurlbaum 5895.90 2 Peirce gave his gratingsand decimeterscales Peirce 5896.20 5 foruse in Rowland's laboratoryin JohnsHopkins Bell 5896.20 10 University. Louis Bell, Fellow in Physics,stated in his paper, "The Absolute Wave-Length of The mean was 5896.156 Angstromunits in air Light," concerning Peirce, at 200 C and 760 mm. pressure. Thus the wave lengthof the D1 line of sodium,as derivedfrom No fullreport of his workhas as yetbeen published, the correctedwave lengthof a spectralline meas- thoughit is evidentlyvery careful, and has already ,consumedseveral years. Certainresults were com- ured by Peirce, was incorporatedin Rowland's municatedto ProfessorRowland of this University standardwhich was acceptedfor several decades. to serveas a standardof referencefor his greatmap Michelsonand Morleyacknowledged Peirce's pio- of the solar spectrumnow nearlycompleted; and it neer effortsby the statement,"The firstactual was to serveas a checkon theseresults and to fur- nisha valueof theabsolute wave-length as nearlyas attemptto make the wave lengthof sodium light possiblecommensurate in accuracywith the micro- a standardwas made by Peirce." 53 metricalobservations, that the experimentsdetailed A by-productof Peirce's workon the spectrum- in thepresent paper were undertaken.49 meter,and an importantcontribution for the use Bell stated that grating"H" with which a large of diffractiongratings for purposesof metrology, part of Peirce's work was done, showed as was was his firsttheoretical. explanation of "ghosts" suspected,a local error,equivalent to a correction in a spectrum. These ghosts,or falselines, which of one part in 55,000. Bell also statedthat Peirce are repetitionsof the principalspectrum, had been used glass decimetersIII and IV, standards of pointedout by Quincke in 1872. For the spec- lengthtested on indirectcomparison with Meter trum-meterit was essentialto know the influence No. 49, a standardof whichthe exact lengthwas of theseghosts upon the positionsof the truespec- in doubt. In another article in the American tral lines. Peirce developed a theory that the Journalof Science, Bell stated: ghostsare caused by irregularitiesin the grating 46 Proc. Amer. A cad. of Arts and Sciences 19 (1884): 50Louis Bell, "The AbsoluteWave-Length of Light," pp. 482-483. Amer. Jour. Science, ser. 3, 35 (1888): p. 270. 47 BenjaminPeirce is quotedin supra,n. 14, p. 316. 51 H. A. Rowland,"On a Table of StandardWave 48 A. A. Michelsonand E. W. Morley,"On a Method Lengths of the Spectral Lines," Mem. Amer. A cad. Arts of Makingthe Wave-Lengthof SodiumLight the Actual and Sciences 12 (Cambridge,1902) : p. 107. and PracticalStandard of Length,"Philosophical Maga- 52 H. A. Rowland,"A New Table of StandardWave- zinie, ser. 5, 24 (1887): p. 463. Lengths," Astronomy and Astrophysics, new ser., 12 49 Louis Bell, "The AbsoluteWave-Length of Light," (1893): pp. 321-347. Referenceto Peirce on pp. 321,340. Am4ier.Jouir. Science, ser. 3, 33 (1887): p. 167. 53 Supra,n. 48. 44 VICTOR F. LENZEN [PROC.AMER. PHIL. SOC. and confirmedit by measurementson Rutherfurd appear thatPeirce's theoryof ghostswas a factor gratings. His memoir,"On the Ghostsin Ruther- in the developmentof improved gratings by furd's Diffraction-Spectra,"appeared in the Rowland. AmericanJournal of Mathematics.54In the pre- VII viouslycited Handbuch, Kayser states,"the first theoreticalinvestigation was undertakenby Peirce, Charles Peirce's activityin Metrologyis also who testedand confirmedhis theoryfor a grating demonstratedby his activity in the American was organized on of Rutherfurd."55 Metrological Society, which In a paper, "On Concave Gratingsfor Optical December30, 1873, and whichpublished proceed- Amongthe com- Purposes," Rowland refersto Peirce's theoryof ings until1888 in fivevolumes. mitteesof the Society was one on units of force, ghosts as follows, and the Proceedingsfor May 19, 1875 recordthat Havingrecently completed a verysuccessful machine on February11, 1875, Charles S. Peirce had been forruling gratings, my attention was naturallycalled to the Committeeon Units of Force by to the effectof irregularityin the formand position added of thelines and theform of the surfaces on thedefini- the President,F. A. P. Barnard.58 The record tion of the grating. Mr. C. S. Peirce has recently of the meetingof the Society,December 29, 1877, shown,the AmericanJournal of Mathematics,that a at Columbia College, states,"Prof. C. S. Peirce periodicerror in theruling produces what have been referringto the actionpreviously taken relative to called in the spectrum.56 ghosts Units of Force, stated what he consideredto be Further,Rowland states, the proper form and characterof such units." "The Committeeon Units of Force and Energy Professor Peirce has measured some of Mr. Ruther- furd's gratingsand found that the space increased in reportedprogress; and Prof. C. L. Peirce was passing along the gratingand he also foundthat the added to that committee."59 foci of symmetricalspectra were different. Peirce also served on the Society's Committee Time. At a meeting,May 20, 1879, The conjunctionof these two resultsis set forth on Standard College,the Committee,consisting of in Rowland's paper. In his article "Screw" in at Columbia Abbe, Chairman,E. B. Elliott, H. A. the ninthedition of the EncyclopediaBritannica, Cleveland Newton and C. S. Peirce, submitteda reporton Rowland says, Standard Time. Under Section V, International In making gratingsfor optical purposes the periodic Standard of Time, the Committeeexpressed the error must be perfectlyeliminated, since the periodic opinionthat the meridian 180 degreesdistant from displacementof the lines only one millionthof an inch a convenientand practical from their mean position will produce "ghosts" in Greenwich offered the spectrum. Indeed, this is the most sensitive commonpoint of union among the nations. The method of detectingthe existence of this error, and relationsof the meridianto all parts of the world it is practicallyimpossible to mountthe most perfect were shown (Appendix V) at a glance by refer- of screws withoutintroducing it. ence to an accompanyingchart of the world on In his paper, "Gratingsin Theory and Practice," the QuincuncialProjections of C. S. Peirce. The Rowland presentsa table whichshows how a pri- proceedingsof the meeting state, "A qualified mary line weakens and the ghosts strengthenas assent to this reportis given by Messrs. H. A. the periodic error (in the grating) increases. Newton and Chas. S. Peirce." 60 "Thus one may obtain an estimateof the error Peirce was especiallyactive at the meetingof fromthe appearance of the ghost."57 It would the Society on December 30, 1884, at Columbia College. He gave 54 2 (1879): pp. 330-347. 55 Supra, n. 43,p. 447. a very interestingand detaileddescription of the 58 H. A. Rowland,"On Concave Gratingsfor Optical pendulumexperiments now being carried out by him, Purposes,"Amer. Jour.Science, ser. 3, 26 (1883) : pp. to determinethe forceof gravityat differentpoints 87-98. Throughthe courtesyof Dr. RobertP. Multhauf, in the UnitedStates and elsewhere.. . . In answer Head of the Departmentof Science and Technology, to Mr. Elliott,Mr. Peirce said thathe thoughtboth SmithsonianInstitution, the writerhas had the oppor- Clarke'sand Roger'smeters were too long.... By re- tunityto consulta valuablepaper on the presentsubject, preparedfor the Department,by Mr. Michael Aronson: 58 Proc. Amer. Metrological Soc. 1 (1880): p. 36. "C. S. Peirce and H. A. Rowland-Two Contributorsto This referencethrough the courtesyof ProfessorM. H. the Art of Ruling DiffractionGratings." Fisch. 57 H. A. Rowland,"Gratings in Theoryand Practice," 59 Ibid., p. 78. Astronontyand Astrophysics,n.s., 12 (1893): p. 144. 6OIbid., 2 (1880) :p. 29. VOL. 109, NO. 1,19651 CONTRIBUTIONS OF PEIRCE TO METROLOGY 45 quest of the President,Mr. Peirce gave an account of conferredwith the manufacturers of gaugesand ma- his-measureson old NewportMill or Tower. He said chinery,and also with electriciansand otherswith his measures pointed to the use of the old Norse referenceto the resolutionspassed by the United foot.... Prof. Peirce inquiredof Mr. Burchard (Di- StatesElectrical Conference concerning weights and rector of the Mint) whether or not any allowance measuresand thebest ways of meetingthe wantsof was made for the differentbuoyancy of brass (rela- thecountry.65 tive to gold) at the mintsin Carson and Denver, and those nearer the sea-level. Mr. Burchard said they During his tenure in charge of the Office of used the weights furnishedby the Officeof Weights Weights and Measures, Peirce gave testimony and Measures withoutcorrection.61 to a JointCommission of Congress,appointed un- At the time of this meeting at Columbia College, der the chairmanshipof SenatorWilliam B. Alli- Peirce was in charge of the Officeof Weights and son, "To Consider the Present Organizationof Measures of the Coast and Geodetic Survey. In the Signal Service, Geological Survey,Coast and this capacity he gave to the members of the Society Geodetic Survey,and the HydrographicOffice of an account of the uncertainties in the Troy and the Navy Department. . . ." The hearings were avoirdupois pounds, derived from England, as authorized by the Sundry Civil Act, approved used in the United States. With respect to the July7, 1884, and continuedby the Sundry Civil Troy pound of the Mint, Peirce said that the Act, approved March 3, 1885. Peirce testified Troy pound of the Philadelphia Mint was hollow, beforethe Commissionon January24, 1885.66 In and its specific gravity and consequently its exact conformitywith the resolutionof the American mass were unknown. What was certain was that MetrologicalSociety of December30, 1884, Peirce the Troy weights issued by the Office of Weights spoke in favor of strengtheningthe Office of and Measures did not bear the exact relation to Weights and Measures. To that end, he pro- the avoirdupois weights which was legal in Eng- posed that the Officeshould be empoweredto land. Peirce said that such a state of confusion grant certificatesfor weights and measures that was sufficientof itself to prevent the use of the would be valid in the courts. As an illustration English system in scientificwork.62 At the meet- of the uncertainstate of weightsand measuresin ing of December 30, 1884, the following resolution the United States, he said of the Troy pound of was adopted, the Mint, the standardof United States Coinage, and made by Captain Kater, which was obtained Resolved, that a Committeebe appointed to address the Secretary of the Treasury, and to memorialize in 1827 when Gallatin was Minister to Great Congress, upon the need of establishingan efficient Britain, "it is a weight which now would be bureau of weights,measures and physical units.63 deemed badly made, because theirbuoyancy can- not be accuratelyascertained. The Mint pound The Proceedings of the meeting on May 20, 1885, was not in state, weighed vacuo by Captain Kater and it never has ben weighed since." Peirce stated the Secretary called the attentionof the Committee thatthe Britishpound was one made in 1855 and to Memorialize Congress to the resolutionpassed at that there was no doubt about the relationsof the last meeting on the suggestion of Prof. C. S. Peirce, of the Coast Survey.64 our pound to the pound in England. But we did not knowthe Troy pound in relationto the British VIII pound. Peirce also discussed the relativeadvan- tages of the metricand Peirce was in charge of the Office of Weights Englishsystems of weights and measures. He said that the and Measures from October 1, 1884, to February yard was the 22, 1885, after which time he declined to serve. distancebetween two lines at 62? F, whereasthe meter was the One of his activities while in charge of the Office distance between the ends of a bar at the was reported by the Superintendent in the Report freezingpoint of water,and thus com- for the fiscal year ending June 30, 1885. parison was renderedmore difficultin the case of the meter. During the winter,under instructions from the Super- Peirce's proposal for a National Bureau of intendent,Mr. Peirce proceeded to Boston, Provi- Standards,as expressed the dence, Hartford, New York and Philadelphia, and by resolutionof the AmericanMetrological Society, and as impliedby 61 Ibid., 5 (1885) : p. 46. his testimonyto the Allison Commission,had to 62 Ibid., p. 47. 63 Ibid., p. 48. 65 Report of the Superintendent, 1885 (Washington, 64 I1)id.,p. 83. 1886),p. 38. 46 VICTOR F. LENZEN [PROC. AMER. PHIL. SOC. wait manyyears for its realization. In Germany several mintsduring the previous year. As re- there was establishedin 1887 a Physikalische- quired by law, the Commissionfor the year 1888 Technische Reichsanstalt,and in England there met the second Wednesday of February, 1888, was foundedin 1899 the National Physical Lab- the eighth,at the Mint in Philadelphia,and in oratory. Finally, on April 18, 1900, Lyman J. the presence of the Director of that Mint pro- Gage, Secretary of the Treasury, proposed to ceded to test the finenessand weightof gold and Congressthe establishmentof a National Stand- silver coins reserved during the calendar year ardizing Bureau, in order to free the United 1887 at the mintsin Philadelphia,San Francisco, States fromdependence upon European nations and New Orleans.69 forstandardizing work. The Hearings beforethe Additional recognitioncame to Peirce by his Senate Committeeon Commerceduring the sec- selection to write the definitionsof terms for ond session of the 56th Congress,December 28, weightsand measuresfor the CenturyDictionary, 1900, broughtforth many resolutionsof support firstcopyright 1889. For the English translation from Scientificmen, Universities,and Profes- of the book by Ernst Mach, Die Mechanik in sional Societies.67 The supportingdocuments of ihrer Entwickelung(1883), Peirce rewrotethe earliestdate were two; a resolutionof the Elec- sectionon unitsand measuresso thatit would be trical Conferenceat Philadelphia,September 11, suitedto English speakingreaders of the Science 1884, and the resolutionof the AmericanMetro- of Mechanics (1893). logical Societyof December 30, 1884, which had been suggestedby Charles S. Peirce. Ix In view of his experiencein Metrology,Peirce In the presentpaper we have set forththe de- was appointed to the Committeeon Weights, velopmentof weightsand measuresin the United Measures, and Coinage of the National Academy States during the nineteenthcentury, as back- of Sciences, upon his election to the Academy ground for an account of the contributionsof duringthe meetingof April 17-20, 1877.68 Charles S. Peirce. He played his role as a mem- Further evidence of his standingin the field ber of the United States Coast and GeodeticSur- was his appointmentby PresidentGrover Cleve- vey,which during that century included the Office land, January9, 1888, to the Assay Commission of Weights and Measures. When Peirce joined for United States Coinage. An Assay Commis- the Coast Survey in 1861, it was perhaps the sion was appointedevery year to examineand test principalinstitution in the United States for re- the finenessand weightof coins reservedby the search in the physicalsciences. Peirce's role can in of his 66 Testimonybefore the AllisonCommission appointed, be understoodonly relationto that co- "To Considerthe Present Organizationof the Signal workers in the Survey, men such as Hassler, Service,. . ." Senate MiscellaneousDocuments, No. 82, Bache, Hilgard, and others. The painstakingef- Forty-NinthCongress, First Session (Washington,1886), fortson the part of these pioneer scientiststo pp. 370-378. create authenticstandards of measure laid the 67 Hearing beforethe Subcommitteof the Committee on Commerce.U. S. Senate,Fifty-Sixth Congress, Sec- foundationsfor the quantitativedevelopment of ond Session,December 28, 1900. Senate DocumentNo. American science. 70,pp. 1-57; specificallyp. 37. 68 Proc. National Acad. of Sciences, Part 2, 1 (1877) 69Reportof the Directorof the Mint, for the Fiscal p. 125. Year endedJune 30, 1888 (Washington,1888), pp. 53-54.