JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 97, NO. B13, PAGES 19,651-19,661, DECEMBER 10, 1992

Paleomagnetismof 122 Ma Plutonsin New England and the Mid-Cretaceous PaleornagneticField in North America' True Polar Wander or Large-Scale Differential Mantle Motion?

MICKEY C. VAN FOSSEN AND DENNIS V. KENT

Lamont-DohertyGeological Observatory and Departmentof GeologicalSciences, Columbia University, Palisades, iVew York

A palcomagneticstudy of CretaceousWhite Mountainsplutonic complexes in New Hampshireand Vermont yields high unblockingtemperature, dual polarity magnetizationsin differenttypes of igneousrocks. The resultingpole position for threeplutons (71.9 ø N, 187.4ø E, A95 = 6.9ø, age= 122.5Ma) agreeswith pre- viously publishedmid-Cretaceous poles for North America, which togethergive a mid-Cretaceousstand- still referencepole slightlyrevised from Globermanand Irving [1988]at 71.2ø N, 194.1ø E (A95 = 3.7ø,N = 5 studies).We argueon the basisof the wide geographicdistribution of thesestudies, the varietyin tec- tonic settingsand rock types,positive reversal tests, and an overall reversalpattern consistent with geo- magneticpolarity time scales,that this mean pole representsthe North Americanmid-Cretaceous reference field for nominally 36 m.y. (124 to 88 Me). The standstillpole limits to within +4 ø, the motion of the North Americanplate relativeto the Earth'sspin axis. During the samemid-Cretaceous interval, the New Englandhotspot track (124 Ma MonteregianHills, 122.5-Ma CretaceousWhite Mountains,and 103- to 84-Me New Englandseamounts) requires 11ø't:4ø of nonh-polewardmotion of North America,in direct conflict with the palcomagneticstandstill. A similar (-13 ø) discrepancyis independentlydemonstrated betweenthe spin axis and the Tristan da Cunha hotspottrack on the African plate during the mid- Cretaceousinterval. •he hotspot/spinaxis discrepanciesended by -90 Ma when it is shownthat both Atlantic hotspotsagree with North Americanand African dipole palcolatitudesand present-daylocations. Nondipolefields are an unlikelyexplanation of the uniformmotion of thesetwo widely separatedhotspots with respectto the spinaxis, leaving as possibleinterpretations true polar wanderand large-scale(but dif- ferential) mantle motion. The southerlymotion of the mid-CretaceousLouisville hotspotrelative to the spinaxis is ostensivelyat oddswith what wouldbe predictedunder the true polar wanderinterpretation and points to differential mantle kinematics. The motions of the three widely separatedmid-Cretaceous hotspotswith respectto the spin axis may be relatedto the recentlyproposed increase in global oceanic lithosphereproduction rates which gave rise to the mid-Cretaceous"superplume."

INTRODUCHON activity [Larson, 1991; 7•rduno et al., 1991], and fast plate motion in the mantle hotspot framework [Morgan, 1972, It has been long recognizedthat Cretaceouspalcomagnetic 1983; O'Connor and Duncan, 1990]. It is less clear, however, polesfor North Americatended to fall relativelyclose together how to characterizethe mid-Cretaceousin terms of geometry, in the general area of the Bering Strait [e,g., Irving, 1964; amount,and timing of relative motionbetween the hotspotand McElhinny, 1973; Mankinen, 1978]. This tendencywas most palcomagneticreference frames. Previousstudies that include recentlyconfirmed by Globermanand Irving [1988] who calcu- the mid-Cretaceousinterval are primarilybased on global syn- lated a mid-Cretaceousreference pole at 71øN, 196øE(A95 = thesesof palcomagneticand relative plate reconstructiondata 4.9ø, k = 352) on the basisof four publishedstudies on rocks sets and comparisons with models of plate motion in the rangingin age from circa 88 Ma to 124 Ma. The mean pole hotspot reference frame [e.g., Livermore, et al., 1984; position may therefore representthe palcomagneticfield Andrews, 1985; Gordon and Livermore, 1987; Besse and relativeto cratonicNorth Americafor some36 million yearsof Courtillot, 1991]. The complexityof theseanalyses may con- the Cretaceous,a "standstill"in apparentpolar wanderduring tributeto the currentlack of consensuson the subjectand, fur- which there was no discernibleplate motion relative to the thermore,Livermore et al. [1984] suggestedthat the pre late geographic axis according to the geocentric axial dipole Cretaceousmotion between the hotspotand palcomagneticref- hypothesis. Bracketing the mid-Cretaceousstandstill are Late erenceframes might be betterattributed to the errorsin hotspot Jurassic/EarlyCretaceous poles [May and Butler, 1986; Van track identificationsand/or palcomagnetic data. Fossen and Kent, 1992] and Late Cretaceouspoles [Diehi, In this study we examine North American mid-Cretaceous 1991] which requirerelatively fast North Americanapparent apparentpolar wander througha palcomagneticstudy of the polar wander (~lø/m.y.). youngest group of White Mountain Series plutons in Nee The mid-Cretaceousis itself an interestinginterval in the England(mean age = 122.5 Ma). This independentdetermina- annalsof global change. Aside from being dominatedby a tion provides an additional test of overall consistencyof the ~30-m.y. intervalof uniformnormal geomagnetic polarity (the North American mid-Cretaceouspalcomagnetic poles whose CretaceousNormal Polarity Superchron,corresponding to the reliability can be further assessedby comparingreverse and CretaceousQuiet zonesin the oceans),the mid-Cretaceoushas normal polarity magnetizationsamong the new 'andpublished beencharacterized as a periodof fastseafloor spreading [Larson pole positiondata for a regional-scalereversal test. As we will and Pitman, 1972; Larson, 1991], increasedmantle plume show,the North Americanrecord of the mid-Cretaceouspalco- magneticfield 'andhotspot activity alsoprovides a more direct Copyfight1992 by the AmericanGeophysical Union. experimentto testvarious ideas on the natureof the palcomag- Paper number 9ZIB01466. netic and hotspotreference frames. 0148-0227/92/92JB-01466505.00

19,651 19,652 VAN FOSSENAND F•NT: MtD-•ACEOUS PALEOMAGNETICFIEIX)

TIlE CRETACEOUSWI•IITE Mou•rr•s MAG•A SE•S 4 Ma). Andesite and basalt flows within the Ossippee Mountainsring dike have been sampledat six sites (minimum Geological Setting age of 121 q- 4 Ma based on K-Ar biotite age of kntruding Conway granite). These flows were tilted throughdifferential The Cretaceous alkaline intrusions in southern New subsidenceabove the intruding magma chamber [Kingsley, Hampshireand southeasternVermont are the youngestgroup of 1931; Billings, 1956]. Gabbro and a basalt dike at Mount igneous rocks associatedwith the White Mountains Magma Tripyramid have been sampledat three sites. Unfortunately, Series (Figure 1). Their emplacementfollowed an initial late only a commercialradiometric date appearingon the map of Triassic phase of igneousactivity at ~230 Ma and a Jurassic Hatch and Moench [1984] is available (112 :t: 5 Ma) and, phaseat ~175 Ma [Foland and Faul, 1977]. The Cretaceous althoughHubacher and Foland [1991] do not list Tripyramidas granites, monzonites, gabbros, basalts and andesitesoccur one of the Cretaceous intrusions redated, we assume that it falls predominantlyin five major plutonic and ring-dike complexes within the distinct interval of Cretaceousmagmatism centered that intrude lower to middle Paleozoic metamorphic and at 122.5 Ma. igneous formations. These complexes are (north to south, Thermal and AF vector demagnetizationprofiles of White Figure 1) Mount Tripyramid, Ossippee Mountains, Mounta'mssamples are straightforwardwith the exceptionof Merrymeeting,Mount Ascutney,and Mount Pawtuckaway. those from the Merrymeeting gabbro. Samples from K/fi biotite datingof numerousCretaceous White Mountains Tripyramid, Pawtuckaway,and Ascutneycontain a single com- plutons and stocks, including the five major p!utonic com- ponent of magnetizationwith high unblocking temperature plexes sampled here for palcomagnetism,suggests an age (-570 ø C) and high coercivity(Figures 2a, 2b, and 2c). This rangeof about118 Ma to 125 Ma [Foland eta!., 1971; Foland magnetization is of reversed polarity (declination =154 ø, and Faul, 1977]. Recently, a number of CretaceousWhite inclination =-56ø) in Mount Ascutneyand Mount Tripyramid Mountainsplutons have been restudied using the 40Ar/39Ar gabbros,and is of normal polarity (declination=343 ø, inclina- biotitemethod by Hubacherand Foland [ 1991]. They suggesta tion =+59ø) in Mount Pawtuckawaymonzonite. shorterperiod of magmatic activity, perhapsonly 3 m.y. in duration,centered on 122.5 Ma. Thesedata, togetherwith pub- lishedzircon fissiontrack (mean= 113 +_10 Ma [Doherty and Lyons, 1980]) and apatitefission track ages(mean = 103 :t: 15 White Mountains Magma Series Ma [Zimmermanet al., 1975;Doherty and Lyons,1980]), sug- gest that theseplutons were emplacedat shallowdepths (-3.5 Cretaceous.:...... • Jurassic + Triassic km) and were subject to monotonic cooling since crystalliza- tion.

Palcomagnetism '... In the pioneering study of the White Mountains Magma Series by Opdyke and Wensink [1966], a pole position was calculatedfrom a variety of plutons consideredthen, with the exceptionof the CretaceousMount Ascutney,to be entirely Jurassic (-180 Ma). The wide age range of the White Mountains Magma Series has been subsequentlydocumented [e.g., Foland and Faul, 1977] and renewedpalcomagnetic stud- ies of the Middle Jurassicand Triassic plutonshave revealed a sensiblerelationship between inferred magnetizationage and radiometricage [Van Fossenand Kent, 1990; Wu and Van der Voo, 1988]. Our experimentalplan for the Cretaceousplutons was to apply fiartheralternating field (AF) and new thermal demagnetizationstudies to samplesfrom the original Opdyke and Wensink collection (10 sims), as well as demagnetization tppeeM s. experimentson our own samples(7 sims). Becausethe peak AF field appliedby Opdykeand Wensinkwas generally only 30 Mt. mT and because 'the thermal method was not used, further work was necessaryto addressthe issue of contaminationby sec- ondarymagnetizations with modem paleomagneticmethods. The paleomagnetic sampling covers the five major Cretaceous White Mountains intrusions. While the recent t. Pa 40Ar/39Ardating of Hubacherand FoIand [1991] provides a •f ...,i:•'M•;•'•;•'•..i!•ySMerrymeettng refined temporalassessment of Cretaceousmagmatism. The specific and earlier published K/At age information from Foland and Faul [1977], unless otherwise indicated, is also quotedfor reference. Gabbro has been sampledat three sites from Mount Ascutney(K/Ax biotite age = 119 :t:4 Ma) as well as at three sims from the Merrymeetingpluton (K/fi biotite Fig. 1. Locationmap of the White MountainsMagma Series'm New age = 117 :t:4 Ma). To the southat Mount Pawtuckaway,mon- England. The five majorCretaceous igneous complexes sampled for zonite has been collectedat two sites(K/Ar biotite age = 121 :!: paleomagnetismare highlighted. VAN FOSSF.N AND KENT: MID-CRETACEOUSPALEOMAGNETIC FIELD 19,653

A MountTripyramid - gabbro B MountPawtuckaway - monzonite c Mount Ascutney- gabbro

30mT W,Up W,Up W,Up 2

S N

450øC 400øC 80rot

O.I AIm

25mT S N 575øC• .N

S••• EDn E,Dn

D OssippeeMountains- basalt flow E OssippeeMountains- basalt flow Merrymeeting- gabbro IfMA.4A .1WB I A

2•x•3øC W,Up 430øC W, Up 0oc 220øC 500øC

S , (,570øC''.:::::'.:N

S • • • • • N O,2A/m

EDn E,I)n E Fig. 2. Representauvedemagnetograms showing AF and/or thermal demagnetizationof various CretaceousWhite Mountainslithologies. (a) Gabbrofrom Mount Tripyramid,(b) Mount Pawtuckawaymonzonite, (c) Mount Ascutneygabbro, (d, e) basaltflow from the OssippeeMountains, and (f) unstableMerrymeeting gabbro. Open/closedsymbols projected on to vertical/horizontalplanes in geographiccoordinates.

The Ossippee volcanic rocks also contain a high stability, (Figure 2f). Althoughthe Merrymeetingmagnetization is most reversed polarity magnetization (declination =162 ø, inclina- probably of reversedpolarity, in the absenceof clear origin- tion =-49 ø) but to slightly higher unblocking temperatures bound demagnetization trajectories the directional data from (-600 ø C). In addition,five of the six Ossippeesites show a the Merrymeetingsamples were excludedfrom furtherstudy. consistent normal polarity overprint (declination =349 ø, Isothermal remanent magnetization (IRM) experiments on inclination =+70 ø) that is usually removed with applied tem- representativesamples from the CretaceousWhite Mountain peraturesof-200øC (Figure2d), but which did extendto >400ø plutonic complexes (including the unstable Merrymeeting C in somesamples from sitesKMA, C, madF (Figure 2e). This gabbro) show efficient acquisitionof magnetizationand satura- low unblockingtemperature component is parallel to the pre- tion in moderate(-150 mT) magneticfields (Figure 3). These sent field direction and most likely representsa remanenceof data togetherwith the observedpeak unblockingtemperatures recent origin. However, the high unblocking temperature of about 570øC suggestthat the predominantcarrier of rema- componentalso appearsto be secondarywith respectto tilt on nence is (titano)magnetite. Slighfiy higher unblocking tem- the basis of the McFadden [1990] discordancy test, even peratures(-600øC) for the Ossippeevolcanic flows suggests thoughthis reversedmagnetization is obviouslynot compati- the additionalpresence of some (titano)hematitein these sub- ble with the presentfield direction. aerial rocks, althoughthe IRM curveswould suggestthat this The origin of the unstablemagnetization at Merrymeeting is contributionis not a significant one. not known. Opdyke and Wensink [1966] reported a reversed Interpretationand pole position magnetizationdirection for these three gabbro sites following blanket AF treatments between 30 mT and 50 mT, but vector A mean pole position has been calculatedusing eight site analysisof their data, includingsome samplesthat were sub- mean magnetizations(Figure 4) convertedto virtual geomag- jected to alternating fields of 160 mT, indicates that a stable netic poles (VGPs) from Mount Tripyramid, Mount Ascutney, magnetizationdirection was not satisfactorilyachieved. Our and Mount Pawtuckawayplutons (71.3 ø N, 187.5ø E, A95 = additional thermal demagnetizationexperiments were also 4.2ø, K = 172; Table I and Figure 5a). The Ossippeepole unable to resolve consistent remanence in these samples (69.8 ø N, 160.3 ø E) has been excluded from this calculation 19,654 VAN FOSSEiNAND KENrr: MID-CRETACEOUSPALEOMAI3NE7IC FIELD

• , I , • I , • I • , I • ,,, I, , MID-CRETACEOUS REFERENCEPOLE FOR NORTH AMEPdCA 1.0 The four publishedpole positionsused by Globerman and Irving [1988] in their calculationof a mid-CretaceousNorth 0.8 American referencepole are (Figure 5b): (1) Arkansasintru- sionspole at 74ø N, 193ø E [Globermanand Irving, 1988, age= g 0.6 100-88 Ma], (2) Niobrara Forrnationpole at 66ø N, 192ø E ' , [Shire and Frerichs, 1974] (age = Coniacian-Santonianor 84 0.4 2 / tl / I Ossippeefiowso to 88.5 Ma on the Kent and Gradstein 1986 time scale), (3) 4 ? I Meymeting MonteregianHills pole at 73ø N, 191ø E [Foster and Symons, • / • / I •o•asc• 1979](40Ar/39Ar date = 124:t: 1 Ma [Folandet al., 1986]or 0.2 I K/At = 118 to 136 Ma [Eby, 1984]), and (4) Newfoundland dikespole at 71ø N, 207ø E [LaPointe,1979, age - 129 Ma]. 0.0• ' '...... The 122.5 Ma White Mountainspole (71.3ø N, 187.5ø E, A95 0 3b 6b 9b = 4.2ø) falls within the age range of the above poles and is B [mT] indistinguishableat the 95% confidencelevel from the mean pole positioncalculated by Globermanand irving [1988] at 71ø Fig. 3. Iso•emal •manent magnetization(•) acquisitioncu•es for representativesamples from the CretaceousWhite Mountains N, 196ø E (A95 = 4.9ø, N = 4 studies).This agreementprovides plutons.J/J• = no•ah•d •; B = directinduc•nce in m•tesla. independentsupport for a mid-Cretaceousstandstill interval which, as suggestedby Globermanand Irving, is reasonably bracketed by the Arkansas intrusions (88 Ma) and the because it is far sided relative to the Tripyramid-Ascutney- MonteregianHills intrusionsof Quebec(118 Ma to perhaps Pawtuckawaymean (-9 ø arc distance,Figure 5a) and especially 136Ma). Consideringthe 40Ar/39Ax date of 124Ma [Foland becauseit lacks a direct radiometricage constraintfor maxi.. et al., 1986] for the MonteregianHills as the more definitive mum age. Basedon the geologicalevidence [Kingsley, 1931; age, we suggestthat the most likely durationfor the standstill Billings, 1956] and the K-Ar age of the intruding granite is -36 m.y. (124 Ma to 88 Ma). [Foland and Faul, 1977], the Ossippeeflows are no younger Rather than signifyingno apparentpolar wander for 36 m.y., than 121 Ma. There is an additional constraint on rninimum the close agreementof the mid-Cretaceouspaleopoles from ageof any thermalremagnetization from a zkcon fissiontrack North America could represent a widespreadremagnetization ageof 107 Ma [Dohertyand Lyons,1980] andwe note thatthe event at somelater stageof the mid-Cretaceous.In this regard, Ossippee pole does not correspond with younger (late the paleomagneticreliability of the three older standstillpoles Cretaceous-Tertiary)published North Americanpoles. The far (Newfoundlanddikes, MonteregianHills, and White Mountains sidedness could be the result of a shallowing bias on the poles) would be of particular concern. There are no stability reversed magnetization through incomplete removal of the tests available for the Newfoundland dikes [LaPointe, 1979] steep norrnal polarity overprint found only at Ossippee. which makes it difficult to judge the reliability of the Alternatively, existingradiometric age control allows the rock Newfoundland pole without a comparison to coeval poles. age and possibly also the magnetizationacquisition age to be However, as in the caseof the 122.5-Ma White Mountainspole considerablyolder than the Tripyramid-Ascutney-Pawtuckaway of this study,the 124-Ma MonteregianHills pole [Foster and rocks/magnetization, and in this regard we note that the Symons,1979] is basedon dual polarity magnetizationswlfich Ossippeepole falls near to older early Cretaceous/lateJurassic passthe reversaltest (A-class, in this case). Another argument North Americanpoles [Irving and Irving, 1982; Gordonet al., against remagnetization in the White Mountains and 1984, see Figure 5a]. MonteregianHills rocks are 40Ar/39Ar ages and apafite or zir- In the absenceof direct age dating of the Ossippeevolcanic con fission-trackages which show no evidence for later ther- flows, the Tripyramid-Ascutney-Pawtuckawaymean is selected mal disturbances[Foland et al., 1986;Eby, 1984;Hubacher and as the representativeCretaceous White Mountainspole. The Foland, 1991; Doherty and Lyons, 1980; Zimmerman et al., dual polarity of the high unblock'mgtemperature magnetiza- 1975]. Moreover, dual polarity magnetizationsin theserocks tions suggeststhat adequate time is representedto average are consistentwith what would be expectedfrom geomagnetic paleosecularvariation and, along with the mean of 40Ar/39Ar polarity tirne scalesfor times prior to 118 Ma [e.g., Kent and age determinations(122.5 Ma), is consistentwith acquisition Gradstein, 1986; Harland et al., 1990], and therefore these during the ti!rte of geomagneticfield reversalsjust prior to the poles carry no suspicion of remagnetization during the CretaceousNormal Polarity Supercitron(-I18 to 84 Ma on the CretaceousNormal Superchron. Finally, the antipodeof the Kent and Gradstein [1986] time scale). The mean of the two mean of two reversed polarity VGPs (mean reversed Pawtuckawaynormal polarity VGPs is antipodalto within 7.5ø Monteregian Hills and mean reversedWhite Mountains plu- of the rneanof the six Triapyramidand Ascutneyreverse polar- tons; 70.1ø N, 180.8ø E) is includedwithin cone of 95% confi- ity VGPs (Table 1 and Figure 5a). For thesedata the critical denceabout the rneanof five normal polarity VGPs (meannor- angle is 13ø, at or abovewhich a null hypothesisof a common mal MonteregianHills, Mount Pawtuckawayfrom this study, mean can be rejected[McFadden and McElhinny, 1990]; the Arkansas intrusions mean, Niobrara formation mean, and mean CretaceousWhite Mountains pole thereforepasses a C-class of the Newfoundlanddikes; 72.8ø N, 190.0ø E, A95 = 5.9ø). reversal test for an isolated observation. We attribute the small Basedon this positiveregional-scale reversal test, we suggest disagreernentof normaland reverse polarity VGPs to paleosec- that no polarity-dependentbias exists in the mean standstill ular variationand regardthe eight VGPs as vectorsdrawn from pole. one population. Even if the departurefrom antipodalityis due A remaining concern,however, is that post-Cretaceoustilt- to slight contamination,the averageof the normal and reverse ing of the igneousrocks from which the constituentpole posi- polarity directionsshould cancel any bias. tion data were derivedsomehow acted to bring the paleomag- VAN FOSSENAND KEI'•: MID-CRETACEOUSPALEOMAGNETIC FIEt.D 19,655

A. N B.

• 150ø w E S • • • "• f '• '• •' r'E Mo:TFipyFamid JWK,•, Ml o o o

o

210ø • 150ø 150 ø s C. Fig. 4. Stereographicprojection of site mean •nagnetizations(squares) with solid/opensymbols indicating lower/upper hemisphere.(a) Normalpolarity magnetization from Mount Pawtuckaway (two site means; circles indicate sample directions) and reversedpolarity magnetization from OssippeeMountains flows (five sites). The low unblockingtemperature normal polarityoverprint at Ossippeeand present-day (PDF) anddipole field directionsat the samplingsites are alsoshown. (b) Mount Ascutneyreversed polarity magnetization (three sites), and (c) Mount Tripyramidreversed polarity magnetization (threesites; circles indicate sample directions).

netic poles fortuitously into close mutual agreement. Direct control on palcohorizontalis available only in the Niobrara shales [Shire and Frerichs, 1974] althoughin the case of the Arkansas intrusions,Globerman and Irving [1988] noted that vertical mid-Cretaceous dikes in that field area intrude flat- lying Lower Cretaceousstrata as well as Paleozoicbasement. The other three igneous poles (Monteregian Hills, White Mountains, and Newfoundland) lack direct control for a palco- horizontal reference: the Monteregian plutons intrude Grenville basementwhereas the White Mountainsplutons and Newfoundlanddikes cut lower and middle Paleozoiccrystalline rocks, respectively. However, at each of these four igneous provinces,the mid-Cretaceousintrusions that were studiedrep- resent the last tectono-magmaticevent recognizedat that par- ticularlocality (the Niobrarashales underwent Laramide defor- mation). Thus while tilting cannotbe excludedat every local- ity, we judgeit to be highly improbableto accountfor the mid- Cretaceousstandstill pole.

,, . Fig. 5. (a) CretaceousWhim Mountainspole positionscompared to North Americanapparent polar wanderpath of Irving and Irving [1982]. The meanpole (KWM) is calculatedusing Mount Ascutney (A), Mount Tripyramid(T), and Mount Pawtuckaway(P) poles,while the OssippeeMountains (Oss) pole, falling nearolder North Americanpole positionsand havingno direct radiometricage constraint,is omitted from mean. (b) New mid-Cretaceous "standstill" pole for North Americaat 71.2ø N, 194.1ø E (A95 = 3.7ø) usingthe meanCretaceous White Mountains CKWM) pole from this study and four other North American cratonicpoles listed by Globerman and Irving [1988]: 1 = Arkansas intrusions[Globerrnan and Irving, 1988], 2 = Niobrara for- marion [Shiveand Frerichs, 1974], 3 = MonteregianHills [Foster and Symons,1979], 4 = Newfoundlanddikes ILsPointe, 1979]. 19,656 VAN FOSSENAND •: M!D•ACEOUS PALEOMAGNETICF!•.n

TABLE !. Site Mean Directionaland Pole Position Dat. a..From the CretaceousWhi.te Mountains Magma Series Magnetization NorthPole Site ..n R k ct95(A95) ..De.c... Inc. pa/eo.A Latitude Longitude dp tim

Mount Tripyramid(43.97ø31, 288.53øE) JWK 5 4.9543 88 8.2 152.6 -58.8 39.5 69.2 195.8 9.1 12.2 • 5 4.9795 195 5.5 156.8 -59.0 39.7 72.3 192.7 6.1 8.2 JWM 5 4.8612 29 14.5 157.2 -50.9 31.6 68.3 171.5 13.2 19.6 Tfipyramid [3] 2.9929 282 7.4 155.6 -56.3 36.8 70.2 185.7 7.6 10.6

OssippeeMountains (43.80ON, 288.72øE) KMA 5 4.9728 147 6.3 160.9 -52.0 32.6 71.3 168.3 5.9 8.7 Tilt corrected 4.9725 145 6.4 150.1 -75.5 64.4 256.8 10.7 11.7 KMB 5 4.9761 168 5.9 166.8 -53.5 34.1 75.9 159.5 5.8 8.3 Tilt corrected 4.9761 168 5.9 164.8 -77.5 66.2 273.5 10.4 11.1 KMC 2 1.9920 -- -.- 155.8 -45.5 27.0 64.2 165.9 -.- -.- Tilt corrected 1.9920 -- -.- 124.0 -77.7 52.7 255.5 -.- KME 4 3.9710 104 9.1 157.2 -41.6 24.0 62.8 159.4 6.8 Tilt corrected 3.9710 103 9.! 135.6 -74.9 58.7 249.0 15.1 16.5 KMF 6 5.6316 13 18.9 170.5 -50.4 31.1 75.3 142.5 17.0 25.3 Tilt corrected 5.6309 14 18.9 148.5 -26.8 50.0 160.8 11.1 20.5 Ossippee [5] 4.9747 157 6.1 161.9 -48.7 29.7 69.8 160.3 6.8 10.6 Tilt corrected 4.6965 13 21.9 146.0 -67.5 66.2 226.6 30.4 36.5

Mount Ascutney(43.45 øN,287.50 øE) KA1 6 5.9842 317 3.8 162.7 -59.2 40.0 76.7 188.3 4.2 5.6 OWA 12 11.9710 384 2.0 153.0 -52.0 32.6 66.0 180.5 1.9 2.7 OWB 12 11.9640 308 2.5 146.0 -55.0 35.5 62.5 192.5 2.5 3.6 Ascutney [3] 2.9895 190 9.0 153.6 -55.6 36.1 68.5 186.6 9.2 12.8

Mount Pawtuckaway(43.10 ON,288.80 øE) T•VO 4 3.9656 87 9.9 341.8 60.3 41.3 76.4 197.1 11.4 15.0 JWP 6 5.9576 118 6.2 343.9 58.2 38.9 77.2 185.1 6.8 9.2 Pawtuckaway [2] 1.9996 -- -.- 342.9 59.3 40.0 76.9 191.2 5.8 7.7

Tripyramid,Ascutney and Pawtuckaway Reverse [6] 5.9730 185 (4.9) 69.4 186.7 Normal [2] 1.9994 -- -.- 76.9 191.3 Meanpole 3 2.9938 322 (6.9) 71.9 187.4

The number(n)of samplemean [site]directions; R, resultantvector length of total numberof sitemean vectors; k, Fisherdispersion parameter,ot95(A95), radius of 95% confidenceabout the mean direction (mean pole position); Dec., Inc., the declinationand inclination of the meanmagnetization; palco-A, palcolatitude; dp anddm, 95% confidenceellipse semiaxes parallel and perpendicular (respectively) to the site-to-polemeridian.

On the basis of wide geographicdistribution and variety in 1983; Crough et al., 1980; Duncan, 1984; O'Connor and tectonic settings of sampling localities, the different rock Duncan, 1990]. The Great Meteor tablemount (at 30 ø N types analyzed,the positive local and regional-scalereversal [Morgan, 198!, 1983]) or the approximatelocation of the cen- tests, and the presenceof reversalsconsistent with geomag- tral Atlantic geoid anomaly(at -27 ø N [O'Connor and Duncan, netic polarity time scales,we concludethat the good agreement 1990]) are two featureswhich havebeen proposed as candidate among paleomagneticpoles from these five studiesrecords a sitesof mostrecent New Englandhotspot activity. standstillduring the mid-Cretaceous(-124 Ma to 88 Ma) in While North Americawas apparentlystationary with respect North American apparentpolar wander at 71.2ø N, 194.1ø E to the palcomagneticreference frame from 124 Ma to 88 Ma, (A95 = 3.7ø, K = 421). This meanpole is only slightlyrevised datededifices along the New Englandhotspot track delineatea from that of Globerman and irving [1988] but is now better progressive apparent motion of more than 1500 km over a definedwith the additionof concordantresults from our studyof comparabletime interval, from the MonteregianHills at 124 the 122.5-Ma White Mountainsplutons. Ma to the Nashvilleseamcunt at 90 Ma. By representingthe mid-Cretaceousstandstill pole as a palcolatitudegrid for North COM•SON WrrH T•m NEW ENGLAND HOTSPOT America, it can be seen that the plutons and seamcuntsof the In addition to providing suitable material for palcomagnetic ,New Englandhotspot chain form a track from ~124 Ma to 90 study,the CretaceousWhite Mountainsplutons help document Ma that trendsalmost perpendicular to the lines of palcolati- the New England hotspottrack. Duncan [1984] has demon- tude(Figure 6). Thusthe New Englandhotspot plume appears strateda regular progression in K/At, 40Ar/39Ar, and m'mimum to have moved southrelative to the spin axis by 11.0ø :t:3.7 ø biostratigraphic ages along the New England (Kelvin) duringthe mid-Cretaceousand only arrivedat ,--30ø N, the pro- seamounts(~103-84 Ma). The ages and positions of the posedmodem latitude, by 90 Ma (Figure6). CretaceousWhite Mountains Magma Seriesplutons (122.5 + This southerlydrift of the New Englandhotspot plume at an 1.5 Ma [Hubacher and Foland, 1991])and the Monteregian averagerate of 0.27ø/m.y(Figure 74) maybe evidencefor a sys- Hills (124 :t: 1 Ma [Foland et al., 1986]) extendthis age pro- tematicshift of the mantlereference frame with respectto the gression on to the North American continent. The New rotationaxis; i.e., true polar wander. Alternatively,either the England chain is thus one of the oldest and longestherspot New Englandhotspot plume drifted independently and is not trackswith physicaldocumentation [Morgan, 1972, i981, representative of the mantle reference frame, or the mid- VAN FOSSENAND KENT:MID-CR•ACEOUS PALEOMAGNEIlCFIELD 19,657

280 ø 90 300 ø

Mid-Cretaceous Palcolatitudes for North America

Monteregian124Ma(118-136 HillsMa)

White Mountains 122.5 Ma (118-127 Ma)

L-DC, O seamount}...... 4oø 91Ma. j

"'i'" Bear103seamount Ma

280 ø 290 ø 300 ø

Fig. 6. New Englandhotspot track comparedto field of mid-Cretaceous(90-124 Ma) palcolatitudesdrawn over North Americausing the new standstillpole (71.2ø N, 194.1ø E), with 95% confidencelimit of :k3.7ø. The New Englandhotspot trackis nearlyorthogonal to thisfield, requitingabout 11 o of polewardmotion for North America.

Cretaceouspalcomagnetic field did not always correspondto a 1984]) which apparentlyrepresent the earliest surfaceexpres- geocentricaxial dipole and, for example, contained a large sion of the Tristan da Cunhaplume. nondipolefield component. To evaluatethese possible expla- Analysis of the Tristan da Cunha hotspot track in a mid- nations,we need to make an independentcomparison of the Cretaceouspalcomagnetic reference frame is not as straight- mid-Cretaceouspalcomagnetic field with hotspottracks on dif- forward as in the caseof North America, simply becauseAfrica ferent plates. not only moved relative to the hotspotsbut also with respect COMPARISONWITH THE TRISTAN DA CUNHA HOTSPOT to the palcomagneticpole [Hargraves, 1989]. One can, how- ever, calculatepalcolatitudes for positionsalong the Tristan da The Tristan da Cunha hotspot track in the South Atlantic Cunha track based on the Africa apparentpolar wander path. definesthe post-125 Ma motion of the African plate relative to Six of 10 Cretaceouspoles listed by Hargraves [1989] have a mantle hotspot presently located beneath the island of been selectedfor this analysis(poles with mid-Cretaceousages Tristanda Curt_haat 38ø S, 011ø W [Duncan, 1981; Morgan, and A95 <10ø). The selectedAfrican poles are (1) Madagascar 1981, 1983]; it is perhapsthe only hotspottrack other than volcanics(69.1 ø N, 240.0ø E; age = -90 Ma [McElhinny and the New England chain with good documentationback through Cowley, 1978]), (2) South African kdmberlites(64.1 ø N, themid-Cretaceous. New 40Ar/39Ar radiometric dates have fur- 226.1ø E; age = 83-101 Ma [Hargraves, 1989]), (3a) Wadi thor documentedthe age progressionin WaNis Ridge sea- Natashpole A, age • 90 Ma (69.3ø N, 258.1ø E [Schultetal., mounts to about 82 Ma [O'Connor and Duncan, 1990], 1981]), (3b) Wadi Natashpole B (64.6ø N, 251.8ø E [Ressetar Locationsof the hotspotbetween about 90 Ma and 120 Ma are et al., 1981]), (4) Lupata lavas (61.8ø N, 259.0ø E; age = 109- inferred from seamountson the easternWalvis Ridge, con- 113 Ma [Gough and Opdyke, 1963]), and (5) Namibia lavas strainedby recovery of lower Aptian (113-119 Ma) sediments (48.3ø N, 266.6ø E; age = 113-131 Ma [Gidskehaugetal., above basement (DSDP site 363 [Bolli etal., 1978]). These 1975). seamountsformed shortly after eruptionof the Entendekaflood These African palcomagneticdata indicate a southerlydrift of basaltsin western Namibia (ages 120-130 Ma [Erlank etal., the Tristan da Cunha plume relative to the spin axis during the 19,658 VAN FOSSENAND gENT: MID-CRETACEOUSPALEOMAGNETIC •

50 , ,, ,t ,• ,, I, ,• ,I •t •, I•, ,, I .... I .... I .... I arrived at their most recently occupiedlatitudes. Therefore in A MonteregianHills independent comparisons with palcomagneticreference frames, the widelyseparated New Englandand Tristan da Cunhahotspot tracks show a remarkable consistency during the mid- 4540 WhiteMountain• Cretaceous interval. • AtlantisII gcarT o,_•.• The consistentrelative motion betweenthe Atlantic hotspots (New Englandand Tristan cla Cunha) and the palcomagneticref- • 35 erence frame can be illustrated in a series of reconstructions of the southernAfrican plate in North American palcomagnetic • -•...... GreatMeteorTablemount- - coordinatesat 124 Ma, 105 Ma, and 90 Ma (Figures 8a-8c). These "paleomagnetic/paleogeographic"reconstructions are achievedby superimposingthe North Americanmid-Cretaceous palcomagnetic field onto relative restorations of western 201,,,,, .... , .... , .... , .... ,.,, Africa, southern Africa, South America, and North America 85 90 95 100 105 110 115 120 125 basedon the publishedseafloor spreading models of Pindell et age [Ma] al. [1988] and Rabinowitz and LaBrecque [1979]. Shown on -20 , thesemid-Cretaceous reconstructions are the African pole posi- tion data (convertedto palcolatitudeerror bars), alongwith two South American results at 124 Ma [Schult and Guerreiro, 1979; Ernesto et al., 1990]. The reconslrucfionsshow the internalconsistency of African, South American, and North American palcomagneticdata as .. ..• ß o 10.5' I , well as the generalconsistency between the New England and _•-35, (1,• Tristan da Cunha hotspots. At 124 Ma thesehotspots begin drifting southwardrelative to the spin axis from locations T which are -12 ø too far north than what would be predictedby a :{ /(2> / o -- { dipole field (Figures8a-8c). The reconstructedage-equivalent positionsalong the New Englandand Tristan da Cunhahotspot .... ,,,,',, chainsare in fact separatedby a consistentarc distanceof 69ø :t: 85 90 95 100 105 110 115 120 125 1ø duringthe interval124 - 90 Ma. This consistencysuggests that inter-hotspotmotion was small, nominally of the order of age [Ma] the lø-2 ø of post-Late Cretaceousinter-hotspot motion sug- Fig. 7. Mid-Cretaceouspaleolatitudinal drift of Ariantic hotspots. (a) Paleolatitudes with 95% confidence limits for selected dated edifices of gestedby Molnar and Stock [1987], and much lessthan the uni- the New Englandhotspot relative to the North American standstill form southerly12 ø shift of the New England and Tristan da pole. Errorsfor the New Englandseamounts (Bear, Atlantis I!, and Cunha hotspotswith respectto the spin axis during the mid- Goshold)are +4ø as inferredfrom the meanstandstill pole. Co)Tristan Cretaceous. da Cunhahotspot relative to (open symbols)selected African poles listed by Hargraves [1989]: (1) Madagascarvolcanics, -90 Ma DISCUSSION [McElhinnyand Cowley, 1978], (2) SouthAfrican kimbedites,83-101 Ma [Hargraves, 1989], (3) Wadi Natash, 86-100 Ma (3a dam from Schult et al. [1981]; 3b data from Re•etar et al. [1981]), (4) Lupata The consistentdiscrepancy of theNew Englandand Tristan da lavas, 109-113 Ma [Gough and Opdyke,1963], and (5) Namibia lavas, Cunhahotspots with respectto the mid-Cretaceouspalcomag- 113-131 Ma [Gidskehauget al., 1975]. Closed symbols,Tristan da netic field allowsthe possibilityof true polar wander;the uni- Cunha hotspot relative to the North American standstillpole (with form shift of the Earth (representedby the hotspots)with 95% confidencelimits --d:4ø) using centraland South Atlantic recon- respectto the spin axis (representedby the palcomagnetic structions. Locationsalong the Tristan hotspottrack are calculated from model B of O'Connorand Duncan [1990]. Linear fits to the data field). However,variable nondipole fields andlarge-scale dif- suggestsoutherly drift rams of 0.27 ø/m.y. (= 3 cm/yr) for the New ferential mantle motion in the mid-Cretaceous are still two Englandhotspot, and a higherbut lesswell-defined rate of 0.38ø/m.y. alternativeinterpretations given the available information. (= 4.2 cm/yr) for the Tristan da Cunha hotspotusing African paleo- At 124 Ma, a 25-30% quadrupolefield contributionat the magnetic data. If the North American standstillpole transferredto MonteregianHills and Entendekaobservation localities alone African coordinatesis used,the drift rate for the Tristanhotspot is 0.32 ø/m.y (= 3.5 cm/yr). could accountfor the -12 ø discrepancybetween the assumed constanthotspot palcolatitudes and palcomagneticpalcolati- tudes(i.e., thosecalculated according to the dipoleformula). mid-Cretaceous(Figure 7b). At 124 Ma the African polesplace This nondipolefield contributionwould need to decreasesys- the Tristan da Cunha hotspot at 25øS (beneath the Entendeka tematicallyover 36 m.y. becauseby 90 Ma, hotspotand dipole flood basaltprovince), whereas by 90 Ma, the plume is located palcolatitudesare in good agreement.Livermore et al. [1984] at 38øS (beneaththe easternWalvis Ridge). The 124-Ma posi- have suggestedCretaceous quadrupole field contributionsof tion is therefore13ø+4 ø north of 38øS, the calculatedpalcolati- nominally 10% or less, with a maximum of 15% between 80 tude of the hotspotat 90 Ma arid its location today. This dis- and90 Ma at whichtime we would,in fact, deema nondipole agreementat 124 Ma and subsequentsoutherly drift until -90 effectunnecessary. In an independentanalysis, Schneider and Ma at an averagerate of 0.38ø/m.y. (Figure 7b) is similar to Kent [1990] have estimatedquadrupolar contributions of-10% that demonstratedindependently for the New Englandplume or lessin the earliestTertiary. In fact, the high overallconsis- (11ø+4ø at 124 Ma and rate of 0.27ø/m.y., Figure 7a). tencyof mid-Cretaceouspalcomagnetic data for North America Furthermore,for both hotspotsthe discrepancywith the palc- andAfrica, asdiscussed by liargraves[ 1989] andpresented here omagneticfield is terminatedat about 90 Ma when the plumes (Figure 8a-c), arguesstrongly for a closeapproximation to an VAN FO$SBNAND KENT: MID-CREtACEOUS PALEOMAGNETICFIELD 19,6•9

C. 90Ma A. 124 Ma • 45ø 0o 125-1357)-3ø• Ma3 ø New90England - 12• Machain 30ø _15o

ß /-.. 15 ø

/ ß ,' ,..:58 ø , •' 5) ñ 3ø / / ...... 113-131Ma • • • •5 o ... 3a) 8.• / 3b)3.6 ø /

.... -.... o ...... 0ø _15•

• •'.. 45 .• :• • • -3oo'•

_60ø-

Fig. 8. Mid-Cretaceouspalcogeographic reconstructions based on relativeseafloor spreading models of Pindellet al. [1988] andRabinowitz and LaBrecque[1979], and the North American standstillpole representedas a palco-latitudinalgrid (applicablefor -90 to 124 Ma). Africanpalcomagnetic data (listed by Hargraves[1989], with symbolsas in Figure7b), shownhere as palcolatitudinalbars confirm the palcogeography.(a) Reconstructionat 124 Ma with additionalpalcolatitu- dinal datafrom SouthAmerica (6 data from Ernestoet al. [ 1990]; 7 datafrom Schultand Guerreiro [ 1979]), (b) reconstruction at 105 Ma, and (c) reconstructionat 90 Ma. The Tristanda Cunhahotspot arrives at its present-daylatitude of 38øS(dashed line) by 90 Ma; the present-daylocation of the New Englandhotspot plume is not known,but by 90 Ma it hasarrived at 30ø S, nearthe present latitude of the GreatMeteor tablemount [see O'Connor and Duncan,!990],

axial geocentric dipole specifically over this time interval. central Atlantic accommodatednortherly motion of the North Furthermore, Besse and Courtillot [1988] have found excellent Americanplate so that it maintainedconstant paleolatitude as overall agreement among global palcomagnetic poles which indicated by the mid-Cretaceousstandstill in North American was interpretedas evidenceagainst large nondipolefield con- apparentpolar wander. This senseof true polar wanderwould tributions in the Mesozoic, certainly none large enough to also predict that a mid-Cretaceoushotspot antipodal to the account for the 12ø discrepancybetween hotspot and paleo- Atlantic hotspots should show a northerly migration relative magnetic paleolatitudesat 124 Ma. Consequently,the evi- to the spin axis (or paleomagneticreference frame) and specifi- dencesuggests that a large, time-varyingnondipole field con- cally at about 124 Ma, the dipole paleolatitudefor a western tribution can be regarded as an extraneous hypothesis to Pacifichotspot should be 12ø southof what would be expected accountfor the observeddisagreement between hotspotsand for a stationarymantle plume. The Louisville hotspotin the palcomagneticreference frames. eastern Pacific, to our knowledge the lone mid-Cretaceous We therefore supposethat the 12ø difference documented hotspotdocumented on the Pacific plate, may have formed the between the Atlantic hotspots and paleomagneticreference Ontong Java Plateau approximately 120 m.y. ago frame at 124 Ma is real and reflects motion of the mantle, at [Engebretsonet al., 1985; Duncan and Clague, 1985]. Based least in the Atlantic hemisphere,with respectto the spin axis. on the Pacific apparentpolar wanderpath of Gordon [1990], If the whole mantle was involved, this would constitute evi- Tarduno et al. [199!] suggest that at about 120 Ma, the dence for true polar wander [Goldreich and Toorare, 1969]. Louisville hotspotwas located at paleolatitudesabout 10ø far- Under such an interpretation,the entire mid-CretaceousEarth ther north than its present location (50øS, 138øW), therefore rotated about an equatorial axis by ~12 ø such that the Atlantic implying southerlymotion of the Louisville plume relative to hemispheremigrated southwardwhile the Pacific hemisphere the spin axis since 120 Ma. Thus although not optimally migrated northwaxd. Southerly true polar wander along an located for a definitive test, the Louisville hotspot gives a Atlantic meridian would imply that seafloor spreadingin the senseof offset with respectto the paleomagneticfield that is 19,660 VAN FOSSENAND I•ENT: MID•TA•OUS PALEOMAONErIC•

ostensively in disagreementwith what is predicted by the Cretaceous,comparisons between the hotspotsand palcomag- Atlantic hotspotsif true polar wander occurredin the mid- netic referenceframes will have to await resolutionof pole Cretaceous. positioncontroversies [e.g., Van Fossenand Kent, 1992] and a clearerdefinition of older hotspottracks. CONCLUSIONS Acknowledgments.The authorswould like to thank two anony- New palcomagneticdata from 122.5-Ma plutons in New mous reviewersfor commentswhich improved the manuscript. England provide supportiveevidence for a standstillin the Funding for this study was provided by the National Science apparentpolar wanderpath for mid-CretaceousNorth America. Foundation,Earth SciencesDivision (researchgrant EAR88-03814). The meanstandstill pole is locatedat 71.2ø N, 194.1ø E (A95 = This is Lamont-DohertyGeological Observatory contribution 4971. 3.7ø for N = 5 studies)representing '[he palcomagnetic field in North America from about 124 Ma to 88 Ma. The standstill ]•EFERENCES pole constrainsmotion of the North Americanplate relative to the spin axis to within :t:4ø . Yet the mid-CretaceousNew Andrews,J., True polar wander:An analysisof Cenozoicand Mesozoic Englandhotspot track clearly shows 11ø+4 ø of northwardNorth palcomagneticpoles, J. Geophys.Res., 90, 7737-7750, 1985. American plate motion relative to the manfie. In an indepen- Besse,$., and V. Courtillot, Palcogeographicmaps of the Indian Oceanbordering continents since -the Upper Jdrassic, J. Geophys. dent analysis, a discrepancyof similar sense and magnitude Res., 93, 11,791-11,808, 1988. (-13 ø) betweenthe African apparentpolar wanderpath and the Besse,J., and V. Counillot, Revisedand syntheticapparent polar mantle was found in the mid-Cretaceousportion of the Tristan wanderpaths of the African, Eurasian,North Americanand Indian da Cunha hotspot track. By about 90 Ma, the discrepancy plates,and true polar wandersince 200 Ma, J. Geophys.Res., 96, 4029-4050, 1991. betweenreference frames appearsto end when both the New Billings, M.P., The Geology of New Hampshire, Part H, Bedrock England and Tristan da Cunhahotspots arrive at dipole palco- Geology, 203 pp., New Hampshire Planning Development latitudes compatible with present-day(or most recently occu- CommissionMining ResourceSurvey, Concord,1956. pied) locationsof the hotspotplumes. Bolli, H. M., et al., Walvis Ridge-Sites362 and363, Initial Rep. Deep If a globally coherentmantle hotspotreference frame existed Sea Drill. Proj., 40, !83-356, 1978. Crough,S. T., W. J. Morgan,and R. B. Hargraves,Kimberlites: Their during the mid-Cretaceous,then true polar wandercould explain relation to mantle hotspots,Earth Planet. Sci. Lett., 50, 260-274, the --12ø southerlydrift of the AtlanticNew Englandand Tristan 1980. da Cunha hotspotsrelative to the spin axis from about 124 to DieM, J. F., The Elkhorn Mountains revisited: New data for -the Late 88 Ma. The geometryof the macpolar wanderevent would pre- Cretaceouspalcomagnetic field of North America,J. Geophys. dict about 12ø of northerly hotspot motion in the western Res., 96, 9887-9894, 1991. Doheny, J. T., and J. B. Lyons, Mesozoic erosionrates in nonhem Pacific relative to the palcomagneticfield. The only available New England,Geol. Soc. Am. Bull., 91, 16-20, 1980. Pacific hotspot for comparison,the Louisville hotspot in the Duncan,R. A., Hot-spotsin the southernoceans-An absolute frame of eastern Pacific, is reported to show southerly motion with reference for the motion of the Gondwana continents, respectto the spin axis during the mid-Cretaceous[Tarduno et Tectonophysics,74, 29-42, 1981. Duncan,R. A., Age progressivevolcanism in the New England al., 1991]. This would favor large-scale, but differential, Seamountsand the opening of the central Atlantic Ocean, J. mantle motion although being only ~90 ø away from the Geophys.Res., 89, 9980-9990, 1984. Atlantic hotspots,the Louisville plume may not constitutea Duncan,R. A., and D. A. Clague,Pacific plate motion recorded by lin- definitive test for true polar wander. ear volcanic chains, inThe Ocean Basins and Margins, vol. 7A, editedby A. E. M. Naim, et al., pp. 89-121, Plenum,New York, Whether the data are interpretedin terms of mid-Cretaceous 1985. true polar wander or large-scale differential motion of the Eby, G. N., Geochronologyof the MonteregianHills alkalineigneous mantle, there remains an interestingcorrelation with proposed province,Quebec, Geology, 12, 468-470, 1984. global changesin productionof oceanic lithosphere. Larson Engebretson,D.C., A. Cox, and R. G. Gordon, Relative motions [1991] has suggesteda suddenincrease in the rate of oceanic betweenoceanic and continentalplates in the PacificBasin, Geol. Soc. Am. Spec.Pap. 206, 1-59, 1985. lithosphereproduction at about 125 Ma, a large portion of Erlank, A. J., J. S. Marsh, A. R. Duncan, R. M. Miller, C. J. which was manifestedas the so-calledPacific "superplume"; Hawkesworth,P. J. Betton,and D.C. Rex, Geochemistryand pet- this correspondsto when there is a large discrepancybetween rogenesisof the Entendeka volcanic rocks from SWA/Namimbia, the hotspot and palcomagneticlatitudes. Following this cli- Spec.Publ. Geol. Soc. S. Afr., 13, 195-245, 1984. max, oceanic lithosphereproduction steadily declinedover the Emesto, M., I. G. Pacca, F. Y. Hiodo, and A. J. R. Nardy, Palcomagnetismof the Mesozoic Serra Gera! Formation,southern mid-Cretaceousand has not changedmuch since about 85 Ma; Brazil, Phys. Earth Planet. Inter., 64, 153-175, 1990. this correspondsto about the time when for at least the Ariantic Foland,K. A., and H. Faul, Ages of the White Mountainintrusives- hotspots(New England and Tristan da Curtha) the latitudinal New Hampshire,Vermont and Maine, USA, Am. J. Sci., 277, 888- discrepancyis no longer apparent. To the extent that the 904, 1977. Foland, K. A., A. W. Quinn, and B. J. Giletti, K-At and Rb-Sr Jurassic superplumemechanism can triggerrelative motionbetween the and Cretaceousages for intrusivesof the White MountainMagma hotspot and palcomagneticframeworks, then consideringthe Series, Am. J. Sci., 270, 321-330, 1971. more modest rates of oceanic lithosphereproduction for 150- Foland,K. A., L. A. Gilbert, C. A. Sebring, and C. Jiang-Feng, 125 Ma andfor 85 Ma-present[Larson, 1991], we wouldexpect 40Ar/39Arages for plutons of theMonteregian Hills, Quebec: evi- dencefor a single episodeof Cretaceousmagmatism, Geol. Soc. to f'md little or no relative motion between the hotspotsand Am. Bull., 97, 966-974, 1986. palcomagneticreference frames for theseintervals preceding Foster,J., and D. T. A. Symons, Definh•ga palcomagneticpolarity and following the mid-Cretaceous. For the period 90 Ma- patternin the Monteregianintrusives, Can. Y. Earth Sci., 16, 1716- present, it is interesting to note that Livermore et al. [1984] 1725, !979. have suggested only minor discrepancies (<5 ø) between Gidskehaug,A., K. M. Creer, and J. G. Mitchell, Palcomagnetismand K-At agesof the southwestAfrican basalts and theirbearing on the hotspotsand global pole positiondata, even thoughBesse and -timeof initial rifting in the SouthAtlantic Ocean, Geoœhys. J. R. Courtillot [1991] appearto favor continuoustrue polar wander Astron. Soc., 42, 1-20, 1975. since the Late Cretaceous. As for times prior to the mid- Globerman,B. R., andE. Irving, Mid-Cretaceouspalcomagnetic field VAN FOSSENAND KENT: MID-CRETACEOUSPALEOMAGNET!C FIELD 19,661

for North America: Restudy of 100-Ma intrusive rocks from test in palaeomagneusm,Geop,•s. J. lnt., 103, 725-729, I990. Arkansas,J. Geophys.Res., 93, 11,721-11,733, 1988. Molnar, P. and J. Stock, Relative motionsof hotspots.in the Pacific, Goldreich,P., and A. Toorare,Some remarks on polar wandering,J. Atlantic and Indian Oceans since late Cretaceoustime, Nature, 327, Geophys.Res., 74, 2555-2567, 1969. 587-591, 1987. Gotdoll,R. G., Testfor biasin palcomagneticallydetermined palco- Morgan, W. J., Deep mantle convectionplumes and plate motions, latitudesfrom Pacificplate Deep Sea Drilling Projectsediments, J. AAPG Bull., 56, 203-213, 1972. Geophys.Res., 95, 8397-8404, 1990. Morgan, W. J., Hotspot tracks and the openingof the Ariantic and Gordon,R. G., and R. A. Livermore,Apparent polar wanderof the Indian Oceans,in The Sea, 7, editedby C. Emiliani,pp. 443-487, mean-lithospherereference frame, Geophys. J. R. Astron.Soc., 91, John Wiley, New York, 1981. 1049-1057, 1987. Morgan, W. J., Hotspottracks and the early rifting of the Atlantic, Gordon,R. G., A. Cox,and S. O'Hare,Palcomagnetic Euler poles and Tectonophysics,94, 123-139, 1983. the apparentpolar wanderand absolutemotion of North America O'Connor,J. M., and R. A. Duncan, Evolution of the Walvis Ridge- sincethe Carboniferous,Tectonics, 3, 499-537, 1984. Rio Grande Rise hot spot system:Implications for African and Gough,D. I. andN. D. Opdyke,The palcomagnetism of the Lupata South American plate motions over plumes,J. Geophys.Res., 95, alkaline volcanics, Geophys.J. R. Astron. Soc., 7, 457-468, 17,475-17,502, 1990. 1963. Opdyke, N. D., and H. Wensink, Palcomagnetismof rocks from the Hargraves,R. B., Palcomagnetismof Mesozoickimberlites in south- White Mountain plutonic-volcanicseries in New Hampshire and ernAfrica and the Cretaceous apparent polar wander curve for Africa, Vermont, J. Geophys.Res., 71, 3045-3051, 1966. J. Geophys.Res., 94, 1851-1866 1989 Pindell, J. L., S.C. Cande, W. C. Pitman !II, D. B. Rowley, J. F. Hadand, W. B., R. L. Armstrong,A. V. Cox, L. E. Craig, A. G. Smith Dewey, J. LaBrecque,and W. Haxby, A plate-kinematicframework and D. G. Smith, A Geologic Timescale 1989, Cambridge for modelsof Caribbeanevolution, Tectonophysics, 155, 121-138, University Press,New York, 1990. 1988. Hatch, N. L., Jr., and R. H. Moench, Bedrockgeologic map of the Rabinowitz, P. D., and J. LaBrecque, The Mesozoic South Atlantic wilderness and roadless areas of the White Mountain National Ocean and evolutionof its continentalmargin, J. Geophys.Res., Forest,Coos, Carroll and Graftoncounties, New Hampshire,U.S. 84, 5973-6002, 1979. Geol. Surv.Misc. Field Stud.Map, 1594-A, 1984. Ressetar,R., A. E. M. Nayin, and J. R. Monrad, Two phasesof Hubacher,F. A., andK. A. Foland,40Ar/39Ar ages for Cretaceous Cretaceous-Tertiarymagmatism in the eastern desert of Egypt: intrusionsof the White Mountain Magma Series, northern New Palcomagnetic,chemical and K-Ar evidence,Tectonophysics, 73, England,and their tectonic implications,Geol. Soc. Am. Abstr. 169-193, 1981. Programs,23(I), 47, 1991. Schneider,D. A., and D. V. Kent, Testingmodels of the Tertiarypalc- Irving, E., Palcomagnetismand Its Application to Geological and omagneticfield, Earth Planet. Sci. Lett., 101, 260-271, 1990. GeophysicalProblems, 399 pp., JohnWiley, New York, 1964. Schult, A., and S. D.C. Guerreiro, Palcomagnetismof Mesozoic Irving, E., and G. A. Irving, Apparent polar wander paths igneous rocks from the Maranh•o Basin, Brazil, and the time of Carboniferousthrough Cenozoic and the assemblyof Gondwana, opening of the south Atlantic, Earth Planet. Sci. Left., 42, 427- Geophys.Surv., 5, 141-188, 1982. 436, 1979. Kent, D. V. and F. M. Gradstein, A Jurassicto recentchronology, in Schult,A., A. G. Hussain,and H. C. Soffel, Palaeomagnetismof' Upper The Geologyof North America,vol. M, The WesternNorth Atlantic Cretaceous volcanics and Nubian sandstones of Wadi Natash, SE Region, edited by P. R. Vogt and B. E. Tucholke, pp. 45-50, Egypt and implicationsfor the polar wanderpath for Africa in the GeologicalSociety of America, Boulder, Colo., 1986. Mesozoic, J. Geophys.,50, 16-22, 1981. Kingsley,L., Cauldron-subsidenceof the OssippeeMountains, Am. J. Shire, P. N., and W. E. Frerichs, Palcomagnetismof the Niobrara Sci., 22, 139-168, 1931. Formation in Wyoming, Colorado, and Kansas,J. Geophys.Res., LaPointe, P. L., Palcomagnetismof the Notre Dame lamprophyre 79, 3001-3007, 1974. dikes,Newfoundland, and the openingof the North AtlanticOcean, Tarduno,J. A., W. V. Sliter, L. Kroenke, M. Leckie, H. Mayer, J. J. Can. J. Earth Sci., 16, 1823-1831, 1979. Mahoney, R. Musgrave,M. Storey, and E. L. Winterer, Rapid for- Larson,R. L., Latest pulse of Earth: Evidencefor a mid-Cretaceous mationof OntongJava plateau by Aptian mantleplume volcanism, superplume,Geology, 19, 547-550, 1991. Science, 254, 399-403, 1991. Larson, R. L. and W. C. Pitman III, Word-wide correlation of Van Fossen,M. C., and D. V. Kent, High-latitudepalcomagnetic poles Mesozoicmagnetic anomalies and its implications,Geol. Soc.Am. from Middle Jurassicplutons and Moat Volcanicsin New England Bull., 83, 3645-3662, 1972. and the controversyregarding Jurassic apparent polar wander for Livermore, R. A., F. I. Vine, and A. G. Smith, Plate motions and the North America, J. Geophys.Res., 95, 17,503-17,516, 1990. geomagneticfield, II Jurassicto Tertiary, Geophys.J. R. Astron. Van Fossen,M. C., and D. V. Kent, Palcomagnetismof the Front Soc., 79, 939-961, 1984. Range (Colorado) Morrison Formation and alternative model of Mankinen, E. A., Palcomagneticevidence for a Late Cretaceousdefor- Late JurassicNorth Americanapparent polar wander,Geology, 20, mation of the Great Valley sequence, Sacramento Valley, 223-226, 1992. California, J. Res. U.S. Geol. Surv., 6, 383-390, 1978. Wu, F., and R. Van der Voo, Palcomagnetismof Middle-LateTriassic May, S. R., and R. F. Buffer, North AmericanJurassic apparent polar plutonsin southernMaine, Tectonophysics,156, 51-58, 1988. wander, implications for plate motion, palcogeographyand Zimmerman, R. A., G. M. Reimer, K. A. Foland, and H. Faul, Cordilleran tectonics,J. Geophys.Res., 91, 11,519-11,544, 1986. Cretaceousfission track dates of apatites from northern New McElhinny,M. W., Palcomagnetismand Plate Tectonics,386 pp., England,Earth Planet. Sci. Lett., 28, 181-188, 1975. CambridgeUniversity Press, New York, 1973. McElhinny, M. W. and J. A. Cowley, Palcomagneticdirections and D. V. Kent and M. C. Van Fossen,Lamont-Doherty Geological pole positions,XV, pole numbers15/1 to 15/232, Geophys.J. R. Observatory Palisades,NY !0964. Astron. Soc., 52, 259-276, 1978. McFadden, P. L., A new fold test for palaeomagneticstudies, (Received June 24, 1991; Geophys.J. lnt., 103, 163-169, 1990. revised June 8 1992; McFadden,P. L. and M. W. McElhinny, Classificationof the reversal acceptedJune 23, 1992.)