194 4Ap J 99. . 210S NEW ORBITS for THE

194 4Ap J 99. . 210S © American Astronomical Society •Provided bytheNASA Astrophysics DataSystem 4 2 is GaposchkhTssecondaryminimum.ThesystemconsistsofaveryluminousBstarwithstrongabsorp- of 0.065.Thevelocity-curveshowsaconspicuousrotationeffectatphasesonbothsides6days—which strong linesoíMnu. except foranunexplaineddifferenceinyandasmalla>.Thespectrumcontainsunusually 3 McDonaldCassegrainplates.TheelementsagreewellwiththoserecentlydeterminedatVictoria plates. Theelementsagreewiththosedeterminedin1928atMountWilson,exceptthateis0.10instead tion linesandanequallyluminousstarwhichisassociatedwithemittingnebulosityofBBei. of the55spectrogramsobtainedbetweenJuly9andSeptember15,1943,showsanumber grams takenattheMcDonaldObservatorybrightlineswereveryweak;and,al- have sometimesbeenlikenedtothebrightlinesinspectrumofßLyrae.Acompari- respectively. SincethenthevariablehasalsobeenobservedbyHertzsprungandhisas- longitude ofperiastronis23.2°.HumasonandNicholsondiscoveredthevariability and anorbitwaspublishedin1928byHumasonNicholson.Theperiodisalmostex- of veryinterestingfeatureswhichmayhelpinelucidatingsomethepuzzles though onlaterdatestheybecamesomewhatstronger,neverevenremotelyre- son ofthetwostarsis,however,atfirstquitedisappointing.Onfewspectro- period of12.004211daysandhavecomputedmyphaseswiththehelpthesedata.The pearance ofthebrightanddarklines. istics ofthesevariations.ThroughtheirkindnessIwasabletoexamineseveral the emissionlinesandalsoofabsorptionestablishedprincipalcharacter- actly 12days,andtherangeinvelocityis385km/sec.Theeccentricity0.065, abound inourdiscussionsofBestars. sembled theconspicuousenfissionlinesofßLyrae.However,amorecriticalexamination fines exceptthoseofinterstellarcalcium,fromthe calciumfines,andfromtheHfines. kin. IhaveadoptedGaposchkin’sepochofprincipalminimum,JD2428429.79,andhis sociates, whofindaßLyrae-typefight-curveandcloselyconfirmtheresultsofGaposch- spectrograms. Thereseemstohavebeennoappreciablechangeintheintensityorap- radial velocitiesarefistedinTable1.Separatecolumnsgivethemeanresultsfromall an eclipsingvariablewithtwoalmostequalminima,whosedepthsare0.43and0.42mag., dated December9,1940. 3 4 3 HD 78316(kCancri).—Aneworbithasbeenderivedfrom84Yerkessingle-prismspectrogramsand ED 163181.—Aneworbithasbeenderivedforthiseclipsingvariablefrom55McDonaldObservatory- This starisofspecialinterestbecauseithasbrightlinesvariableintensity,which The binarynatureofHD163181wasdiscoveredattheMountWilsonObservatory, The startookanaddedimportancewhenS.Gaposchkin,in1938,announcedthatitis i V453Sco;a=Ô-32°27'(1900),mag.6.6,Sp. B0. *Ap. /.,67,341,1928. QuotedbyMrs.C.P.GaposchkinfromaletterProfessor HertzsprungtoDr.HarlowShapley HarvardBull.,No.909,p.20,1938;Ap./.,89,125and 322, 1939. * ContributionsfromtheMcDonaldObservatory,University of Texas,No.86. NEW ORBITSFORTHESPECTROSCOPICBINARIES HD 163181AND78316(76kCANCRI)* McDonald andYerkesObservatories Received November10,1943 Otto Struve 1 ABSTRACT HD 163181 210 194 4Ap J 99. . 210S July Sept. Aug. Date © American Astronoihical Society •Provided bytheNASA Astrophysics DataSystem 1943 30. 29. 28. 27. 27. 25. 25. 20. 28. 15. 15. 14. 14. 14. 13. 13. 12. 12. 29. 28. 23. 22. 22. 13. 15. 14. 13. 18. 14. 14. 13. 12. 12. 11. 11. 13. 4. 8. 6. 8. 8. 8. 6. 5. 9. 9. 5. 7. 7. 8. 8. 6. 9. 9. 7. 4:14 2:06 2:59 2:17 2:19 3:26 4:36 2:30 2:49 2:16 2:51 2:15 2:47 2:03 2:20 2:14 2:24 2:00 2:27 2:54 2:30 2:30 2:12 2:53 3:43 3:06 4:15 4:50 4:32 2:50 2:39 2:24 2:46 1:44 1:47 1:54 1:54 2:49 2:34 2:37 6:14 6:09 6:07 1:54 6:55 5:28 5:51 5:56 6:22 5:26 6:16 5:33 5:48 7:15 7:26 U.T. 2430000+ 982.59 981.62 981.60 980.60 980.57 982.60 981.57 980.62 979.59 979.58 975.58 974.62 965.60 975.64 975.62 975.59 974.59 973.60 973.58 972.69 972.68 971.58 966.59 964.60 963.62 963.60 961.60 961.59 956.62 951.62 950.66 949.62 945.60 944.63 945.61 944.62 943.61 942.61 934.68 933.70 933.69 928.76 927.80 927.79 923.73 919.76 919.74 918.75 917.73 916.77 914.76 918.73 917.76 916.81 914.74 NEW ORBITSFORSPECTROSCOPICBINARIES211 JD 11.94 11.91 10.92 10.90 10.01 10.00 10.93 10.92 II. 9 9 10.95 Phase . RadialVelocitiesofHD163181 I. 1 3 4.91 4.90 0.96 0.94 0.91 0.90 6.94 6.92 5.92 6.89 5.94 5.89 3.91 0.94 0.92 8.90 2.92 7.92 7.91 0.95 5.95 4.94 4.09 6.94 6.93 5.96 5.95 3.94 4.07 8.01 2.09 5.10 3.10 3.07 0.16 1.92 7.02 1.12 8.07 5.08 0.08 7.03 2.15 2.11 Qual v.p. v.p. g g g g g g g g g g g g g g g g g g g * g g g g g g g P P g g g g P g g g g g g g g g g g g g g g g g g g g Lines No. 20 20 of 17 16 17 13 15 13 15 14 17 17 18 16 16 17 16 14 21 16 16 15 12 14 10 12 10 15 15 13 18 15 13 15 16 16 13 15 15 17 18 15 15 14 TABLE 1 8 8 8 9 8 6 8 2 5 6 8 + 0.2 + 95.8 + 102.8 + + + + 159.5 -189.9 -188.9 -152.6 + 164.6 + 91.4 -143.3 -128.7 - 15.9 -169.5 + 3.5 + 10.1 + 1.1 - 6.6 - 90.8 - 87.7 -169.2 -200.2 -208.6 + 8.2 -187.5 -177.0 -163.4 + 6.7 + 112.3 +149.8 + 113.3 + 78.0 + 61.9 + 135.5 + 30.9 + 48.9 + 158.0 + 162.4 + 156.2 + 107.4 + 111.8 - 77.9 -157.7 -121.6 -137.4 - 8.7 -146.1 -196.0 -128.3 -217.0 - 60.1 - 72.6 Lines All 10.7 16.7 0.6 9.1 Velocities inKm/Sec Lines No. of + 2.2 + 2.1 + 2.2 + 18.1 + 2.4 + 4.0 + 12.0 + 3.1 -16.1 - 6.9 - 6.8 - 6.8 + 6.4 -13.0 -15.8 - 0.3 - 1.8 + 11.7 -15.1 -38.0 -10.7 + 4.2 +10.5 + 3.1 + 2.1 -12.0 -60.3 - 6.1 + 16.2 -15.0 - 5.3 - 0.9 - 4.9 - 4.5 - 7.3 Ca n 0.0 Lines No. of + 98.2 +116.5 -198.3 -187.3 -140.5 -142.4 -142.1 - 12.8 - 15.7 - 12.3 - 9.1 +164.8 +155.8 - 3.9 - 21.3 - 29.2 -100.8 - 89.5 -156.8 + 78.5 -160.2 -222.9 -194.2 -202.0 + 87.5 - 23.2 + 166.7 + 56.7 + 44.7 + 140.5 - 20.5 -154.5 -152.5 - 0.1 +102.7 + 10.8 + 162.0 + 165.1 + 150.8 - 20.1 - 85.8 -190.7 -112.0 -119.2 - 5.2 + 7.2 + 102.0 + 108.6 - 11.4 -178.2 -146.2 -170.0 -202.0 - 73.9 - 81.0 +18.9 +23.4 + 4.0. + 9.0 + 14.8 + 5.3 + 7.5 + 5.4 - 1.3 - 0.7 -18.8 -29.5 + 12.1 + 4.6 - 8.1- - 2.2 - 6.9 + 0.2 + 5.6 - 8.1 - 2.5 -12.1 +27.7 -10.6 + 2.8 - 2.4 + 0.5 + 2.9 + 18.9 - 8.3 - 5.7 -20.4 - 6.0 +33.3 +28.2 + 2.9 + 10.4 - 2.0 - 2.2 -13.7 +22.0 + 3.2 + 3.5 - 9.0 -16.2 - 3.3 -13.9 - 0.3 - 7.8 -22.0 - 4.7 -10.3 -10.8 - 3.1 - 2.2 O-C 212 OTTO STRUVE

The plates were grouped in 13 normal places as is shown in Table 2. The values of O—C are those which result from a least-squares solution. Table 3 shows a comparison of the velocities from the H lines with those from all stellar lines. The velocity curve is given in Figure 1, where the observations refer to the normal points. Because the period is so nearly 12 days, the observations are grouped at intervals of about 1 day, and there are no intermediate observations. The orbital elements were derived by means of Schlesing- er’s least-squares method and are shown in Table 4. The probable error listed in the line

TABLE 2 Normal Places

Mean No. of No. Limits of Phase Phase Plates 0-C 1. 0.08- 0.16 0.120 + 0.65 2. 0.90- 0.96 0.931 - 0.16 3. 1.12- 2.15 1.753 + 0.48 4, 2.92- 3.10 3.030 - 5.80 5. 3.91- 4.09 4.002 - 3.82 6, 4.90- 5.10 , 4.986 + 10.92' 7. 5.89- 5.96 5.935 +21.03 Rotation 8. 6.89- 7.03 6.953 -11.171 effect 9. 7.91- 8.07 7.978 - 6.82, 10. 8.90- 8.900 + 5.60 11. 10.00-10.01 10.005 + 6.Ó5 12. 10.90- 10.924 -10.26 10.95 13. 11.91- 11.947 - 3.37 11.99

TABLE 3 Comparison of H and Other Lines

No. No. Hb -\-Hy Normal Mean of H—(All Normal Mean of H—(All 2 Place Phase Plates Lines) — ( All Lines) Place Phase Plates Lines) — ( All Lines) 0.120 -11.0 + 2.2 8. 6.953 7 + 2.9 +25.9 0.931 -26.4 -26.4 9. 7.978 4 + 7.0 - 0.6 1.753 -15.7 -18.6 10. 8.900 1 + 6.6 +22.2 3.030 - 2.1 - 8.3 11. 10.005 2 -14.7 + 9.6 4.002 + 6.5 +13.3 12. 10.924 5 + 4.4 +32.3 4.986 - 9.5 -15.4 13. 11.947 3 - 6.6 - 0.8 5.935 - 5.6 + 3.4 for 7 is that for Schlesinger’s unknown T. The value of T is given in days counted from zero phase at JD2428429.79. The normal points were all given equal weight, in order to reduce the influence of physical factors, such as rotation, which tend to distort the velocity-curve. The value of 2Az>2 was reduced from 1793 to 966. The probable error of a single normal point is +7 km/sec; that of a single plate is +8.7 km/sec. These values are large; Humason’s and Nicholson’s observations give an even larger value, namely, + 9.2 km/sec for one plate. The agreement of my elements with those of Humason and Nicholson is very satis-' factory. There has been no change in co, and the only significant difference is in e. The Mount Wilson value was 0.065. The observations show at once that there is a pro-

© American Astronomical Society • Provided by the NASA Astrophysics Data System 194 4Ap J 99. . 210S km/sec. Theobservationsalsosuggestthattherotation effectbeginsatleast1fullday residuals ofconsiderablesizeatthatphasewhich thevelocityisapproximately—26 0.09. P—1.33days,forthe entiredurationofeclipse.Iftheobservationsaffected bythe before mid-eclipseandlasts atleast1fulldayafterit.Thelimitsmaywell beasgreat is slightlyunsymmetricalwithrespecttothispoint. Bothvelocity-curvesgivepositive not originallyinterpretedinthisway.Withmyelementsthetimeofmid-eclipseshould Whichever valueweuse,thereisastrongindication thatthespectroscopicrotationeffect The MountWilsonelementsgive occur whenz;+co=90°,ortheradialvelocity is it withoutanaccuratevelocity-curveovertheentirerangeofphasesconcerned.The as ±1.5days.Thisdoes notaccordwellwithGaposchkin’sphotometric result,D= effect isalsounmistakablypresentintheMountWilsonvelocity-curve,thoughitwas nounced rotationeffect.Theresidualsarepositiveatphasesnear5and6daysnega- so thattherotationalvelocitymustbeconsiderable.Itwouldprematuretocompute tive atphasesnear7days.Therangeintheseresidualsisoftheorder±25km/sec, © American Astronomical Society •Provided bytheNASA Astrophysics DataSystem K= 189.5km/sec P=12.004211 days T=4.78 days co =+46.2° 7 =—42.8km/sec e=0.04 NEW ORBITSFORSPECTROSCOPICBINARIES213 Preliminary y Kecosco=—30.3km/sec. Fig. 1.—Velocity-curveofHD163181 7 +Zecosco=—22.2km/sec. Orbital ElementsofHD163181 K —192.6km/sec P =12.004211days T =4.04days co =+23.7° 7= —39.8km/sec 6 =0.10 TABLE 4 Final ± 0.76day ± 0.014 ±22.2° ± 2.9km/sec ±3.5 km/sec Probable Errors 194 4Ap J 99. . 210S versal, sothat,ifthereis anyabsorptionfineduetotheunknownstar’sshell, itmust are strongandgreatly displacedtowardthered.Theyaresingle,without centralre- certain ofanyabsorption fines oftheunknownstar.Atphase8.0daysemission fines very slightly,themotionofunknownstar.The questioniswhetherornotwecanbe blended withthoseoftheunknownstarandthat residualsoftheHvelocitiesreflect, real, isnotasimplerotationeffect.Itmoreprobable thattheHfinesofBstarare measure arethoseoftheBstar,notunknown star.Hencethephenomenon,ifitis sent, sothattherecanbenoserioustroublefrom blending.ButtheHfineswhichwe residuals atphases8.9,10.0,and10.9days,negative residualsatphases0.9,1.8,and Table 3mightsuggestthatsuchaneffectdoesexist: forHbandHywehavepositive 3.0 days.Duringthelargerpartofthisinterval emissionfinesareveryweakorab- from theabsenceofanydilutioneffectinfinesHeiorintensityMgn phases, areweakenedbythesuperpositionofstrongcontinuousspectrumun- siderations. Inthevicinityofmyphase0orHumasonandNicholson’s9.5days, v +co=270°.MyobservationsinthisrespectagreewiththoseofHumasonandNichol- duced byabsorptionofthecontinuousfight unknownstarintheemittingshell. tation effectinthemeasuresofHabsorptionfines, ifthelatteraretoanyextentpro- such aneffect.ButthebrightHfinesdisappearat phase0.Hencewemightexpectaro- the twosidesofphase0.TheBstarisinfront,anditsfinesshould,course,notshow because ofthegapsinourobservations.Itmaybeannular,atphasesnear0days we mustconcludethattheunknownstarisnotresponsibleforabsorptionfinesat fines belongingtotheunknownstarareobservedatelongations—inspiteoffact known component.Thislatter,then,doesdisplayacontinuousspectrum,eventhoughit the absorptionspectrumofastar,notthatshell.Itisclearthisstarmustbe 4481. Thesefinesarenormal.TheHstrongandnarrow,buttheylackthetypi- enhancement appearstobeverycloselythesameforalllines.Thecriticaltestcomes lines, whichareweakatallotherphases,greatlyenhanced,andtheamountofthis mid-eclipse. TheBstarstandsinfront,theunknowniseclipsed.absorption This apparentparadoxisoftheutmostinterest,anditssolutionwillundoubtedlylead pal eclipsefalls0.5PearlierthantheofBstarwhichwehavejustdiscussed. weak andshowtherotationeffect,buttheyneverquitedisappear.Sincenoabsorption in front,andtheemissionfinesarelittleaffected.TheabsorptionofBstar ing risetothemareeclipsedatphase0. does notshowabsorptionfinesatthetimesofelongation.Atourphase0,whenab- the normalB-typecomponentofbinarysystemwhoseabsorptionfines,atallother cal sharpnessofHfinesproducedinatenuousshell.Weconcludethatweareobserving the absorptionlinesareverystrong.ThisispreciselyphaseofGaposchkin’sprincipal to importantresults. son, whosearchedfortheothercomponentbutcouldnotfinditsabsorptionspectrum. Yet thereisnoindicationinthespectrumofasecondsetabsorptionlineswhichwould is consistentwiththeinternalprecisionofmeasurements. the emissionfinesdoseemtodisappearcompletely. phase 6days.Theeclipseisprobablynottotal,thoughwecannotbequitecertainofthis that Gaposchkinfindsforitstotalfight=0.46asagainstL\0.54theBstar— conclude thattheunknownstarhasemissionfinesinitsspectrumandgasesgiv- sorption finesoftheBstararestrong,emissionveryweakorabsent.We of Gaposchkin’speriod,oratphase0.6day,ifweagainusethespectrographiccriterion correspond tothatbodywhichiseclipsedatphase0,ifwedependuponanextrapolation 214 OTTOSTRUVE rotation effectareexcluded,theprobableerrorofoneplateisreducedtoavaluewhich © American Astronomical Society •Provided bytheNASA Astrophysics DataSystem The generalrunofthevaluesO—CinTable2showsthatthereisnorotationeffecton At phase6days,whenGaposchkinobserveshissecondaryeclipse,theunknownstaris I believethatasatisfactorysolutionoftheparadoxisobtainedfromfollowingcon- Probably themostremarkableresultof.Gaposchkin’sworkisfactthathisprinci- O^i (T) CM o co © American Astronomical Society •Provided bytheNASA Astrophysics DataSystem

PLATE IX

The Spectrum or HD 163181 194 4Ap J 99. . 210S 6 6 r by thelongerexposuresnecessary atminimumlight—suggeststhattheH shell isnever until phase11days,beingthenclearlydouble,with thestrongercomponentonred far astoquestion,withWilson,theexpanding-shell hypothesisinthecaseofanormal metric elements.Theseconsiderations,together with O.C.Wilson’sresultsforHD to fitwithintherathersmallamountofspaceallowedbyGaposchkin’srevisedphoto- let shift,whichshouldprobablybepresentinanexpandingshellthatissmallenough plain it,butwearefullyawarethatitconstitutesanimportantprobleminastrophysics. side. Atphase0theyaredoubleandveryweak.But theirpresence—enhanced,perhaps, has forbiddenlines.Ifsuchlineswerepresent,their velocitiesshouldnotbeaffectedby doubtedly isimportantinsomestars.Unfortunately, neitherHD163181nor193576 Wolf-Rayet star.Self-absorptioninthemannerconsidered byBealsforMWC374un- Even ifsuchexpansiondoesexist,itwouldbedifficulttoavoidthecomplicationofavio- evidence ofexpansioninthegaseousshellunknowncomponentHD163181. lines ofthePCygni-typestar,MWC374.Inthisremarkablespectrumshiftisprob- general phenomenonamongspectroscopicbinariesofveryearlytype.Wecannotex- binary, andthedisplacedemissionlinesoscillateinjSiasewithabsorptionlines.In self-absorption, andthetestwouldbeeasy. ably causedbyself-absorptioninadeceleratedexpandingshell.Thereis,atpresent,no mately equallyluminousstars,andtheyoscillatewithaphasedifferenceof0.5Pfromthe HD 193576.In29CanisMajoristheredshiftoccursinbrightercomponentof intensity astheBstarandthatithasveryweakabsorptionlineswhichareneverseenat But atphase10daystheyaremuchweakened.They remainaboutconstantinintensity It isalmostcertainlyrelatedtotheredshiftobservedbyC.S.Bealsinemission absorption lines.Itisquiteprobablethattheredshiftofemissionlinesafairly HD193576 andHD163181thedisplacedemissionlinesbelongtooneoftwoapproxi- massive one,becauseitispossiblethatthevelocityofemissionlinesdoesnotrepre- of theabsorptionlines.Thisdoesnotmean,however,thatunknownstarismore below thecontinuousspectrumbyhigh-levelgases.Wemust,therefore,leaveundecided residual isnearlyzero:therenoabsorptionfromtheshell.Atphasesnear6days set ofemissionlineswhosevelocitiesaresystematicallyshiftedtowardtheredbyperhaps the elongationsoratsecondaryminimum.Itdoes,however,haveassociatedwithita I suggestthattheunknownstarhasacontinuousspectrumofapproximatelysame sion thattheemissionlinesdoshiftbuttheirrangeisconsiderablysmallerthan out beinginfluencedbythedisturbingpresenceofabsorptionlines.Itismyimpres- tem isnotquitecertainbecauseitexceedinglydifficulttomeasurethebrightlineswith- 0.5P. Whethertheemissionlinealsoshiftswithrespecttocenterofmasssys- sion linetakenasawholeshifts,relativetotheabsorptionline,withphasedifferenceof ed conspicuouslytowardthered.Therecanbenoquestionthatvelocityofemis- ward thevioletofabsorptionlines.Atphasesaround9daystheyarestrongandshift- the twoelongations.Atphasesaround3and4daystheyareveryweakshiftedto- whether theabsorptionsarerenderedshallowbysuperpositionofemissionorcut blend withtheHlinesofBstarandshiftittowardred.Butatphase8.0days the originofHresidualsinTable3. emission linesaredouble.ButtheeclipseofBstarispresumablyannular,orpartial, sent thevelocityofunknownstar. so thattheHabsorptionlinesmaybelongtoit.Theemissionaretooweakshow 193576, leavetheproblemofredshiftunsolved, thoughIshouldnotwanttogoso 1.5 A,asaretheemissionlinesof29CanisMajorisorWolf-Rayetcomponent 6 6 © American Astronomical Society •Provided bytheNASA Astrophysics DataSystem O.C.Wilson,Ap.95,402, 1942.J.R.A.S.Canada,37,241,1943. The HemissionlinesofHD163181arestrongat phases7,8,andprobably9days. Why, then,istheappearanceofemissionlinessodifferentattwomaxima? Our pictureiscomplicatedbythefactthatemissionlinesarequitedissimilarat NEW ORBITSFORSPECTROSCOPICBINARIES215 194 4Ap J 99. . 210S 8 9 ble toseparatesuchaneffectfromtheproposedredshift. lines. Buttheweakeningofemissionlinesatphase10days,only1dayafterelonga- the latterareconsistentwiththisinterpretation,becauseofweaknessemission with theabsorptionlinesofBstar.Itisdifficulttodecidewhetherintensities of thegeneralredshift,whichwouldbringtheirstrongest,portionroughlyincoincidence totally eclipsedbutremainsasaluminousfringearoundtheunknown ting gasesarenotsymmetricallydistributedaroundtheunknownstar,butitispossi- ready wellunderway2fulldaysbeforemid-eclipse.Itisentirelypossiblethattheemit- 216 OTTOSTRUVE tion, wouldthenindicatethattheeclipseofemittingshellunknownstarisal- star. Atphases1,2,3,and4daystheemissionlinesareveryweak,presumablybecause spectrograms takenattheYerkesObservatory.Anorbit,basedupon25plates, velocity-curve wassufficientlyunsymmetricaltoplacethestaramongthoseobjectsfor ence incoisindicated.Sincethisdifferencenot consistent withthefactthatvalue results oftheirmeasurements32high-dispersion spectrogramsofkCancritakenbe- in thelineofapsidesduringlast36years.Anewseriessingle-prismspectrograms was obtainedbyN.Ichinohein1907.Theeccentricityofthisorbit0.15,andthe Victoria orbitisgreaterthanthatobtainedbyme. Nevertheless,therearecertainfea- this workatVictoriawasmuchgreaterthanthatused atYerkes,andtheprecisionof was startedattheYerkesObservatoryin1929and wascontinued,withinterruptions, which anewdeterminationisdesirableinordertotestwhethertherehasbeenchange appreciably inshapefromthatcomputedwiththe Victoriaelements,andasmalldiffer- time theobservationsmadeatVictoria.Insecond place,myvelocity-curvediffers Yerkes observationshaveagapbetween1935and 1941 andthereforedonotduplicatein tures inmyvelocity-curvewhichappeartojustify itspublication.Inthefirstplace, tween MarchandMay,.1938,withtheVictoria72-inch reflector.Thedispersionusedfor until May,1943.InthemeantimeJ.A.Pearceand PhoebeRiddlehavepublishedthe 7h 8 9 © American Astronomical Society •Provided bytheNASA Astrophysics DataSystem The binarynatureofkCancriwasdiscoveredin1904byFrostandAdamsfrom5 a=92“3;Ô+11°4'(1900), mag.5.1,Sp.B8. Ap.25,315,1907. Pub.A.A.S.,10,65,1940. 7 HD 78316(76KCANCRl) 194 4Ap J 99. . 210S 2 obtained byPearceandRiddleisalmostidenticalwiththevalueIchinohe, km/sec, theminimumofvelocity-curvewasnotobservedsufficientlytoallowavery there isnoevidenceofarotationthelineapsides.ButIchinohe’sorbitbasedupon Victoria valueofthetimeperiastronpassage.Theobservationswerearrangedin15 The phaseswerecomputedwiththeperiodobtainedatVictoria,andzerophaseis critical determinationofco. a relativelysmallnumberofplates,and,whilehismeanerrorperplateisonly+4.2 fect ofphysicalcauseswhicharesometimesknown todistortavelocity-curve.Thevalue normal places(Table6),andthevaluesofO—Cwere computedseparatelyforeachplate from JD2429012.081.The probableerrorlistedinthesecondlineisthat obtainedfor of SAflwasreducedfrom123.8to63.6,andtheprobable errorofonenormalvelocityis equal weightsforallnormalpoints.Thiswasdone inordertominimizethepossibleef- with thehelpofpreliminaryelementsandwere thenaveragedforeachnormalplace. in T^ble7.Thetimeofperiastron passageisexpressedindayscountedfrom phase0,or A least-squaressolutionwasmadefollowingthe method ofSchlesingerandadopting Schlesinger’s variableF. The principaldifferencefromtheelementsofPearce andRiddle 1R9180.. ± 1.6km/sec;thatofoneplateistheorder±3.5 km/sec.Theelementsareshown 9995.. 9188.. 9197.. 10005. © American Astronomical Society •Provided bytheNASA Astrophysics DataSystem 10884. 11687. 11663. 12717. 12700. 12676. 12669. 12654. 11707. 11705. 11702. 11694. 11668. 12759. 13009. 12707. 12659. 12648. 11678. 11672. 12741. 12736. 12732. 12724. 13020. 12757. 12753. 12752. 12749. 12747. 13019. 12992. 12991. 12784. 12772. 12766. 13039. 12976. 12794. The 84newYerkesobservationsand3McDonaldarelistedinTable5. Plate 1929 Apr.14 1941 Nov18 1933 Dec.6 1932 Jan.19 1942 Jan.6 1935 Mar.11 Date Apr. 7 Jan. 24 Jan. 23 Jan. 13 Mar. 24 Mar. 16 Dec. 19 Apr 16 Apr. 15 Apr. 15 Apr. 1 Mar. 30 Feb. 21 Feb. 12 Dec. 14 Dec. 11 Dec. 6 Nov. 20 Apr. 13 Mar. 29 Mar. 27 Mar. 15 Dec. 12 Dec. 4 Nov. 6 Apr. 29 Apr. 13 Apr. 12 Mar. 19 Mar. 19 Mar. 5 Feb. 20 Dec. 4 Nov. 27 Nov. 17 Nov. 17 Feb. 13 Mar. 13 NEW ORBITSFORSPECTROSCOPICBINARIES217 hin 11 35 10 11 46 11 34 10 31 11 50 10 43 10 46 11 47 11 34 11 48 4 4 3 5 8 20 350 U.T. 9 56 3 3 3 3 5 8 20 8 00 6 35 7 45 Radial Velocitiesof76kCancri < Phase 4.686 4.984 4.356 0.696 5.168 2.597 0.575 5.028 5.585 3.483 4.488 4.511 4.464 0.252 2.468 0.121 4.519 6.344 6.307 5.958 3.450 3.606 4.792 4.459 2.704 5.978 3.773 3.643 2.444 3.404 3.893 0.295 2.082 5.735 5.168 2.617 5.842 6.063 5.453 6.382 2?450 1.314 1.781 1.113 (Elm/Sec) +77.3 +52.6 +72.4 +27.0 +80.9 +57.6 +51.1 +52.4 +64.1 +21.7 +30.6 +74.5 +21.4 +25.6 +57.1 +78.2 +62.2 +66.3 +43.1 +53.5 +58.1 +72.7 +79.0 +60.1 + 4.7 +41.0 +62.3 +63.2 -55.7 -44.2 -49.8 -59.9 - 0.3 -52.8 -51.3 - 2.5 -13.6 -49.7 -41.1 - 4.8 -52.3 -22,7 -39.1 -23.2 Vel. TABLE 5 CG2053.. CQ1959.. CQ1945.. 1R13108. 1R13040 ,13131. 13091. 13090, 13142. 13134. 13098. 13097. 13092. 13089. 13079, 13078. 13057, 13056 13050, 13179. 13171. 13162. 13155. 13154. 13144. 13143. 13132. 13130. 13126. 13110. 13109. 13049 13180. 13170. 13164. 13163. 13156. 13127. 13119. 13118. 13117. 13111. 13135. Plate 1943 Jan.15 1942 Dec.12 Date Apr. 8 Apr. 8 Apr. 1 Apr. 22 Apr. 22 Apr. 21 Apr. 21 Apr. 21 Apr. 8 Mar 22 Mar 4 May 3 May 3 May 1 Apr. 22 Apr. 1 Mar. 31 Mar. 31 Mar. 31 Mar. 30 Mar. 22 Mar. 22 Mar. 21 Mar. 21 Mar. 21 Mar. 21 Mar. 13 Mar. 2 Feb. 19 Feb. 18 Feb. 18 May 1 Mar. 30 Feb. 19 Feb. 18 Feb. 18 Feb. 16 Feb. 16 Jan. 15 Dec. 16 Dec. 16 hm 11 05 10 03 ll36 3 45 2 18 2 48 5 25 1 45 8 11 8 05 6 51 5 38 8 51 7 52 8 14 7 10 8 10 7 59 6 85 7 06 U.T. 40 46 43 39 49 45 43 56 52 37 08 53 34 45 48 40 43 57 54 15 13 25 16 Phase 0.724 0.767 2.438 0.738 0.678 5.092 5.048 2.525 2.481 6.183 6.139 6.089 5.181 5.113 3.624 1.488 1.358 1.307 1.264 3.580 3.534 2.614 2.570 4.851 4.586 1.532 1.439 2.518 2.470 4.535 3.694 2.847 3.669 3.628 3.574 3.526 3.651 6«? 108 1.068 1.643 1.598 1.606 1.562 (Km/Sec) +47.1 +14.0 + 3.2 +32.9 +37.1 +54.9 +57.2 +82.1 - 1.4 - 3.4 -15.3 -42.2 +20.8 +24.1 +78.3 +84.6 -47.6 -50.1 - 8.9 -36.9 -43.2 -41.5 +67.7 +56.7 +59.9 -33.2 +53.8 +46.5 +74.3 +52.1 +61.7 +87.8 +78.3 +10.1 + 4.4 -25.4 +85.3 +82.1 +13.2 +13.9 +79.2 +80.4 -23.6 Vel. 218 OTTO STRUVE is in the value of 7, for which they find +25.3 km/sec. It is possible that this difference of 4.2 km/sec is due to a systematic difference between the two instruments, though the amount is alarmingly large. The published systematic differences for stars of classes ES- AS are: Yerkes — Lick = +0.95 km/sec , Victoria — Lick = —0.95 km/sec .10

However, these determinations (which incidentally are in the wrong sense) probably do not apply in the present case. In September, 1941, a change was made in the Bruce spec-

TABLE 6 Normal Places

Com- Com- No Limits of Mean No. of puted O-C No. Limits of Mean No. of puted O-C Phase Phase Plates Velocity Phase Phase Plates Velocity 0.12-0.30 0.22 -53.8 +2.3 3.65-3.89 3.73 +77.5 +2.1 0.58-0.77 0.70 -44.7 -2.3 4.36-4.59 4.49 +60.4 +0.3 1.07-1.36 1.24 -13.6 +0.5 4.69-4.85 4.78 +47.8 -1.3 1.44- 1.55 + 7.3 +0.7 14.98-5.18 5.10 +30.0 .64-2.6 1.78-2.08 1.93 +31.0 +1.5 5.45-5.84 5.65 - 7.1 +3.6 2.44- 2.47 +57.2 -3.1 25.96-6.18 6.07 -35.0 .52+0.1 2.57-2.85 2.66 +64.1 -1.4 6.31-6.38 6.34 -47.8 -3.4 3.40-3.64 3.55 +78.2 + 1.0

TABLE 7 Orbital Elements of 76 « Cancri

Preliminary Final Probable Errors P = 6.39316 days P = 6.39316 days 7= +21.1 km/sec 7= +21.1 km/sec + 0.8 km/sec K= 66.9 km/sec K= 66.1 km/sec + 0.6 km/sec co= 161.0 co= 174.6° + 2.4° 6= 0.14 e= 0.137 + 0.086 T= 0.00 r=+0.21 day + 0.065 day

trograph which resulted, according to Hynek’s measures of f Tauri, in the following difference : Yerkes new — Yerkes old = —2.0 km/sec .

Measures of several plates of standard-velocity stars have given a similar result. Since formerly the systematic error of the Yerkes plates was +1.0 km/sec, it should now be about —1.0 km/^ec. It is difficult to account for the rest of the discrepancy. The plot of the velocities in Figure 1 certainly does not warrant the idea that there has existed since 1941 a systematic difference of —4 km/sec, which did not exist in 1929-1935. If there is a difference, it is in the opposite sense. It is not impossible that the velocity of the center of mass of k Cancri is slightly variable. This question must be postponed until more ma- terial on the systematic errors is available. In spite of this uncertainty, the other elements are known with considerable precision. There is no spectrum of the secondary component, and it is probable that what Ichinohe

' 10 J. H. Moore, Pub. Lick Obs., 18, xi, 1932.

© American Astronomical Society • Provided by the NASA Astrophysics Data System 194 4Ap J 99. . 210S 3289.40. 3329.38. 3327.78. 3322.68. 3303.19. 3298.21. 3295.86. 3488.79. 3403.39. 3389.74. 3365.98. 3363.02. 3360.05. 3358.39. 3355.78. 3349.02. 3339.79. 3337.51. 3313.71. 3495.80. 3494.41. 3492.48. 3490.42, 3483.11. 3482.03, 3475.55. 3474.13. 3471.76. 3468.72. 3460.21. 3457.78. 3442.04. 3433.52. 3422.83. 3421.11. 3408.70. 3399.89. 3394.51. 3387.67. 3383.52. 3379.62. 3376.55, 3372.52. 3368.03. 3341.71. 3336.28. 3335.12. 3464.45. 3451.69 3446.07. 3439.05. 3436.28, 3428.13. 3419.47, 3374.62. Wave Length © American Astronomical Society •Provided bytheNASA Astrophysics DataSystem Int. 0 0 0 0 0 0 4 3 3 0 0 Onn 0 0 0 0 0 0 2 2 5 2 2 7 7 7 1 1 1 1 1 1 In 1 1 1 1 In 1 1 1 1 1 In 1 1 1 In 1 1 In 1 1 1 1 NEW ORBITSFORSPECTROSCOPICBINARIES219 {Fe i8.51) List ofLinesintheSpectrumkCancri Fe ii Mnu 1.98 Feu 6.11' Cm Feu Fe H4.67 Niu 1.35 Fe n8.68 Mn ii5.84 Fe i0.60,Tiu1.06 Mn n8.68 Mn ii4.06,n4.15 Fe n4.50 Mn n0.33 Fe n1.23,u1.32 Mn n8.96 Cr ii3.32 Cr ii9.82 Cr ii Feu Feu Cm 7.62 Co ii6.40 Cm 3.30 Cm 2.74 Cm 1.20 Cm 9.54 Cr ii6.27 Cr ii8.05 Cru Cr ii Mn n2.91 Cr ii8.94 Cm 8.77 Feu Cru Cm 5.13,Fen5.25, Ti n4.55,Cr4.32 Ti ii4.35 Tiu 2.80 Tiu Tin Ti n Ti n7.84 Ti ii3.76 Ti ii Ti ii Ti H Ti ii Tiu 6.18 Fe n5.74 Identification 0.30 9.80, Tiu0.34 6.76, Cm8.35 8.50 6.26 9.00, Tiu9.41 9.5 2.86, Feu3.47 6.32 2.93, Fen3.07 9.35 7 85 \5.22/, Cm5.28, /5.17\ 7.89 5Í42, Feu5.81 1.84 14.00 Cr ii5.46 TABLE 8 ,3896.00 3708.40. 3710.37. 3697.11. 3706.06 3704.03, 3687.09. 3685.48, 3627.55 3734.61. 3730.21 3722.15. 3717.58. 3715.66. 3712.18. 3691.71 3682.55 3631.59 3821.84 3798.16 3781.27 3774.73 3765.84 3761.48 3750.28 3743.44 3735.33, 3825.22 3778.44 3770.70 3767.61 3763.93 3585.90 3576.87 3563.39 3827.69 3812.49 3759.50 3741.67 3736.77 3535.83 3516.29 3510.38 3862.84 3844.35 3835.52 3832.49 3814.50 3507.93 3499.71 3856.27 3854.00 3847.30 3497.60 3889.34 3879.25 3878.08 Wave Length lOnn lOnn lOnn lOnn lOnn Int. 4nn 3nn 2nn 5nn 2 Onn 0 0 0 4 2 0 2n 2 3 2 7nn 1 0 0 4 0 2 1 1 1 4 1 In Inn 2 In 1 Onn 5 5‘ 1 1 0 3 3 In 1 1 1 1 1 1 1 1 1 1 {Fe i8.58) E 134.37 ZZ 141.94 Mn n8.06 E 163.86 Fe i3.37 E 151.97 Mn n9.88 Ca n6.04,Tih6.23 H 177.15 E 196.83 H 181.56 Fe n7.17 Feu 1.97 Feu 1.51 E 110.63 Fei 7.21 Fe i3.80,n4.11 E 120.15 Ca ii6.92 Cm 5.19,5.45 Fe n4.91 Feu 4.12 Fe i2.96 E 107.90 Cr ii5.62 H 202.81 Mnu 3.98 E 95.39 Mgi 2.31 Feu 7.08,Fen7.69, Cm 1.49,1.72 Si n2.60 Si ii6.03,Mn6.53 Si h3.67 Tiu L65 Ti n5.20 Tiu 1.32 Ti u9.30 Cr ii5.31 Ni n6.76 E 89.05 Fe i8.02 Fe n5.82 Fe h9.88 V u8.36 F ii4.34,Tiu4.65 Fe u7.39 Mn n7.53 V H7.32,Mgii8.24 Ti n5.41 7* n0.85 F n6.16 Fe i7.83 Identification oCO

J 220 OTTO STRUVE

TABLE 8—Continued CT) Wave Length Int. Identification Wave Length Int. Identification 3898.06. 2n Fe i 7.89, Fe i 8.01 4192.51. 0 Ni u 2.02 3900.68. 2n Ti il 0.54, Al ii 0.68 4198.12. 0 Sin 8.25, Fe i 8.31 3902.56. In Fei 2.9$ 4200.40. In Ti n 0.40, Mn n 0.25 3905.56. In Cr n 5.64, Si i 5.53 4202.65. In Fe i 2.03, V n 2.34 3906.27. 0 Fe ii 6.04 4205.49. 2 V ii 5.09, Fe n 5.48 3913.63. 0 Ti ii 3.46 4206.63. 3 Mn ii 6.43 3918.99. 0 Fe ii 8.51 4215.53. In Sr n 5.52 3920.85. 2 Fe I 0.26 4219.93. 0 Fe i 9.36 3926.33. 1 Mn ii 6.47 4222.59. 1 Fei 2.22 3930.99. 1 Fe i 0.51 4229.62. In 3933.72. 7 Ca n 3.67 4233.29. 3 Fen 3.17 3935.96. 1 Fe ii 5.94 4235.22. 0 Fei 5.94 3939.04. 0 Fe ii 8.97 4237.72. 1 3941.25. 1 Fe i 0.82 4239.33. 2 Mn 9.72 3942.52. 0 Fe i 2.44 4240.53. 1 Al ii 0.75 3943.84. 3 Mn ii 3.82 4242.48. 3 ^ Cr n 2.38, Mn n 2.38 3958.58. 0 Cr n 8.07 4244.36. 1 Mn n 4.25 3968.38. 3 Ca n 8.47 4246.91. 1 Sc n 6.83 3969.98. lOnn E e 0.08 4248.17. 1 Mn ii 7.95, Fei 8.23 3981.35. 0 Fe i 1.77, Feu 1.61 4251.65. 2n Mn n 1.77 3983.86, 5 Eei3.96, Cr i 3.91 4253.18. 3 Cr n 2.62, Mn n 3.02 3999.79. 1 Mn ii 00.06 4256.02. 3 4002.30 0 Fe ii 2.55 4259.38. 3 Mnn 9.26 4006.48 1 4262.11. 2 Cr ii 1.92 4013.82 0 Fe i 3.82, Fe i 3.79 4264.07. 0 Fen 3.89 4024.26 2n Fe ii 4.55 4267.26. 3n C n 7.02, Cn 7.27, 4026.11 3nn Sei 6.19 Fe i 6.97 4028.15 0 Ti n 8.34 4269.26. 0 Cr n 9.28 4030.27 1 Fe i 0.49, Mn i 0.75, 4273.39. 1 Fe ii 3.32 Cr n 0.28 4275.75. 0 Cm 5.57 4032.59 2 Sen 2.95 4278.55. 1 Fen 8.13, Cm 8.10 4034.38 0 {Mn i 4.49) 4282.45. 3 Mn n 2.50, Fei 2.41 4038.98 1 Mn n 8.73 4284.26. 1 Cm 4.21 4044.32 1 Fe n 4.01 4288.07. 0 Fin 7.88 4045.80 0 Fe i 5.81 4292.25. 3 Mn n 2.25 4050.29 1 4296.43. In Fe n 6.58 4063.51 0 Fe i 3.60 4300.23. 3 Ti n 0.05, Mn n 0.22 4066.93 0 Ni ii 7.05 4303.43. 2 Fe n 3.17 4070.82 0 Cr ii 0.90 4305.89. 0 Fe i 5.45, Sc n 5.71 4075.17 0 Sin 5.45, Cr n 5.63 4308.27. 2n Fin 7.90, Fei 7.91 4076.47 1 4314.61. 1 Fe ii 4.29 4081.13 1 4318.47. 0 Fe n 8.22 4085.10 1 {Fe i 5.32) 4322.73. 1 4101.51 lOnn E 8 1.74 4325.49. 2 Fei $.16 4104.75 0 Mn ii 5.01 4326.78. 3 Mn n 6.76 4110.50 2 Cm 1.01 4340.58. lOnn S T 0.46 4120.74 In Sei 0.81 4344.17. 4 Ti n 4.29, Mnn 4.03 4122.78 On Fe ii 2.64 4345.80. 1 4127.97 6 Si ii 8.05 4348.35. 1 Mn n 8.49 4130.89 6 Si n 0.88 4350.62. 0 Ti ii 0.83 4136.94 4 Mn ii 6.91 4351.94. 1 Fen 1.76 4140.29 0 Mnu 0.19 4357.32. In Fe n 7.57 4144.05 0 Sei 3.76, Fe i 3.87 4363.39. 0 Cm 2.93 4163.89 On Ti n 3.65 4377.42. In {Mo n 7.76) 4171.82 In Tin 1.90 4379.85. 1 Mn n 9.74 4173.79 2n Sen 3.45 4385.30. 1 Fe n 5.38 4177.69 1 Ser7.60, F n 7.55, 4388.11. 0 Sei 7.93, Fei 7.90 Fen 7.70 4409.46. 1 Ti ii 9.22, Ti ii 9.52 4178.78 3 Fen 8.85 4414.66. 0 Fei 5.12 4183.85 On V n 3.43 4416.94. 1 Fe u 6.82

NEW ORBITS FOR SPECTROSCOPIC BINARIES 221

TABLE 8—Continued

Wave Length Int. Identification Wave Length Int. Identification 4420.89. 1 4508.32. Fe n 8.28 4431.22. 0 Fen 1.63 4511.81, (O n 1.82) 4434.10. 0 Mg n 3.99 4515.59 Fe ii 5.34 4451.66, 0 Fe n 1.54 4519.28, {Fe n 0.22) 4462.88. In 4522.93. Feu 2.63 4466.11. On Fe i 6.55 4525.76, 4468.20. In Ti h 8.50 4528.33, Fe i 8.62, V n 8.51 4471.68, 3n He i 1.48, Re i 1.68 4549.45, Fe ii 9.47, Ti ii 9.63, 4475.41. 1 Fe ii 9.21 4478.65. 2 Mn ii 8.74 4555.68. Fe ii 5.89 4481.48. 6 Mgu 1.33 4558.32, Cr n 8.66 4483.78. 1 4583.50, Fe H 3.83 4491.47. 1 Fe H 1.40 4588.41. Cr n 8.22 4499.32. 0 Fe n 9.71 4590.24. Ti n 9.95, Cr n 9.89

thought was the spectrum of the secondary was really due to faint lines oí Mn n, belong- ing to the primary, which were not identified in 1907. There is, in particular, a line of Mn n at X 4478.74, which looks deceptively like a component of Mg n 4481.33. I have measured a considerable number of fines on two McDonald Observatory spectrograms on Eastman Process emulsion. The fist of these fines is given in Table 8. The appearance of the spectrum is that of a typical Mnu star with sharp fines.